Comparative blood chemistry and histopathology of tilapia infected with Vibrio vulnificus or Streptococcus iniae or exposed to carbon tetrachloride, gentamicin, or copper sulfate

Abstract

A comparative study was conducted of blood chemistry and histopathology of Nile tilapia ( Oreochromis niloticus) after infection with Vibrio vulnificus or Streptococcus iniae or exposure to carbon tetrachloride (CT), gentamicin (Gm), or copper sulfate (Cu). Treatments were chosen as models targeting different organs and from which different types of lesions were expected. Moderate histological changes were observed in the gill, trunk kidney, liver, spleen, and head kidney in both bacterial infections. Gentamicin induced significant changes in the trunk kidney, liver, and intestine, while carbon tetrachloride treatment caused changes in the liver and trunk kidney. Copper treatment only caused mild proliferation of the gill epithelium. Among the blood chemistry parameters measured, hematocrit, iron, glucose, sodium, and chloride were good indicators of deteriorating health in tilapia, with significant changes in those parameters being observed in most treatment groups. Severity of histopathology in the liver, intestine, and trunk kidney was associated positively with alanine aminotransferase (ALT) and aspartate aminotransferase (AST) but negatively with sodium, chloride, cholesterol, and total protein. Anion gap was the only blood chemistry parameter correlating to histopathology in the spleen and head kidney. Standardized principal component analysis, discriminant analysis, and canonical correlation analysis were applied to the blood chemistry and histopathology data. A consensus set of important parameters that included sodium, chloride, bicarbonate, total protein, ALT, AST, cholesterol, total iron binding capacity, and iron was obtained. Using a 7-parameter model that included potassium, ALT, iron, total protein, bicarbonate, creatine kinase, and AST, 70% of the fish could be correctly classified to the appropriate treatment group, while a 17-parameter model allowed 81% proper classification.