BackgroundCurrent systems for assessing protein quality such as the Digestible Indispensable Amino Acid Score (DIAAS) correct apparent amino acid digestibility for basal endogenous protein losses (bEPL), ignoring the potential influence of the diet on these losses. However, the quantification of total endogenous protein losses (tEPL) poses a challenge. ObjectiveTo evaluate different methods for quantifying tEPL and bEPL, and to assess their potential in discriminating between tEPL originating from bacteria and host. MethodsUsing an incomplete Youden square design, twelve ileal cannulated pigs received ten different protein sources, and a nitrogen-free (NF) diet. Ileal digesta were collected on days 6 and 7 of each 1-week feeding period, to quantify endogenous protein losses (EPL) and analyze apparent ileal digestibility. Ileal EPL were estimated based on 1) 16S-+18S gene copy qPCR, 2) diaminopimelic acid (DAPA)+18S, 3) differential AA profiles in digesta, EPL and bacteria, equaling tEPL, and 4) an NF diet and 5) whey protein isolate (WPI), equaling bEPL. ResultsIleal bEPL based on the NF and WPI method correlated moderately-highly (r=0.69, P<0.05), but the NF method probably underestimated bEPL. In pigs fed the WPI diet, EPL based on the WPI and AA profile method were highly correlated (r=0.88, P<0.01). Overall, tEPL based on the AA profile method were moderately correlated with the 16S+18S method (r=0.58, P<0.001), and DAPA+18S (r=0.57, P<0.001). Low correlations were observed between bacterial tEPL based on the AA profile method and 16S or DAPA. Host tEPL based on the AA profile method and 18S were weakly correlated (r=0.39, P<0.001). ConclusionsThe AA profile method seems the most appropriate method for tEPL quantification, while the WPI method is preferred for bEPL quantification. Despite challenges in distinguishing between bacterial and host EPL, it is evident that bacterial proteins substantially (on average 37-83%, depending on method) contribute to the EPL.