**1. Introduction**

Liver machine perfusion (MP) was introduced in the clinical setting by Guarrera and colleagues [1] in 2009. Based on its ability to recondition, evaluate, and preserve liver grafts, MP showed a particular potential in reverting the detrimental impact of extended criteria donors (ECD) on post-liver transplant (LT) outcome and quality of life [2,3]. Among MPs, normothermic machine perfusion (NMP) represents the most promising technology due to its higher evaluation potential [4]. Further, NMP can be used in preclinical settings to evaluate other preservation or treatment techniques outside LT. In this last setting, NMP is defined as normothermic machine reperfusion (NMRP).

Since reactions elicited during ex-situ dynamic perfusion are largely unknown, technical and biological aspects of liver MP need to be extensively investigated through committed research. A number of animal models have been developed for this purpose. Because of their immediate translational value (e.g., appropriate human-like liver size), swine models have been broadly used in the start-up phase to rapidly translate preclinical results into the clinical setting. Conversely, small animal models can better be exploited to investigate subcellular mechanisms and changes associated with MP. To this purpose, based on low cost, reproducibility, and better understanding of subcellular events, rat models have been widely used [5].

Over the last ten years, 94 research papers were based on liver MP use in rodents. We reviewed all those reporting rat ex-situ perfusions with recirculating fluid and 39 of them described NMP/NMRP (Supplementary Materials: a brief review of the literature is provided in Tables S1 and S2). A large variability in NMP protocols among different research groups was evident. In particular, the potential usefulness of an oxygen carrier (OxC) during normothermic perfusion was not fully explored. While only few researchers used OxC perfusate for NMP or NMRP [6–16], most of them adopted a non-OxC model. The infrequent use of OxC could be related to the high number of rats needed to be used as blood donors (at least 3-4 animals/experiment) and unavailability of other oxygen carriers. However, while adequate oxygen delivery (DO2) was observed in the absence of an oxygen carrier, the increased oxygen consumption ( . *VO*2) needed to control the reperfusion injury suggests the opportunity to increase DO2 during reperfusion [8,17]. To this purpose, two main strategies may be adopted: use of non-cellular hemoglobin [18] or employment of other sources of blood [13]. While the first is limited by the restricted availability of these products, human blood cells can be easily procured for experimental use.

The aim of present research was optimization of procurement and perfusion procedures to obtain a safe and reproducible rat model of NMP. The study examined the potential protective role of improved DO2 during NMP and evaluated the safety and efficacy of human red blood cell use. Indeed, the use of non-murine red blood cells can allow for the reduction of the number of animals/experiments in full respect of the 3Rs principles (Refinement, Reduction, and Replacement) [19].

#### **2. Material and Methods**
