The winter ulcer disease caused by the G ram-negative bacteria Moritella viscosa, is currently one of the major threats to the health and survival of farmed Atlantic sal mon. Given the escalating prevalence of M. viscosa , we sought to compare three experimental models, the commonly used bath immersion with M. viscosa, a co-habitation model with shedder and contact fish, and an in-vitro system of fish scale explants. F ish and cell cultures were pre-expos ed to commensal Atlantic salmon skin seawater bacteria prior to bacteria challenge .
Mortality was higher (6%) in the bath immersion trial compared to the co-habitation model (2.6%) (Fig. 1 A) . At the end of the bath trial the fish were mainly without clinical symptoms. In the co-habitation trial ulcer development was more pronounced, with 25% of the fish having superficial lesions, and 15% had moderate to severe wound development, accompanied by reduced weight. F ish exposed to M. viscosa displayed less scale loss as compared to controls (Fig. 1 B) , but with no effect of treatment with commensal bacteria prior to challenge . Further, the commensal Atlantic skin bacteria induced morphological changed to skin cells (Fig. 1 C) and induced unspecific immune responses (Fig. 1 D) in the in-vitro scale cultures, with large difference towards dead and live bacteria.
Overall , the co -habitation model represented progression of winter ulcer disease as typically observed in the field . In addition, In-vitro scale explants represented an excellent model to study transcriptional and morphological responses in the skin towards bacteria.