Sixteen Large White pigs (25.5 ± 2.5 kg) were used for ITx (eight donors, eight recipients). They were all female, non-related, outbred, and aged 8–10 weeks to ensure uniformity in veterinary management. Prior to use, all animals underwent strict veterinary control that determined their good health and condition. The protocol was approved by the Animal Welfare Ethics Committee (PROEX 58.7/20) and complied with the EU and Spanish Directives on experimental animals (63/2010 EU, RD 53/2013). The entire donation process under DCD conditions was strictly controlled by the veterinary staff of our institution.

The animals were starved for 24 h before surgery. Premedication was performed directly in the box (12 mg/kg of ketamine intramuscularly + 0.5 mg/kg of midazolam). With the animals under inhalational anesthesia (sevoflurane, 4%), we performed catheterization of the ear veins and arteries and endotracheal intubation, which was followed by a propofol (11 mg/kg) and fentanyl (5.4 μg/kg/h) infusion. A 14-gauge angiocath was inserted in the fifth intercostal space in the mid-axillary line. Withdrawal of life support therapy was performed, and a bolus of unfractionated heparin (300 IU/kg) was injected through a central venous catheter. Manual air thoracic insufflation (100 mL/2′) was performed to induce progressive tension pneumothorax and subsequent lethal cardiovascular collapse. After 200 ± 50 mL, a persistent decline in oxygen saturation level was observed. Functional warm ischemic time (fWIT) was initiated when oxygen saturation was sustained at <80% and/or mean blood pressure was <50 mmHg, as established by the Spanish legislation for DCD in clinical practice [15]. Death was declared after cardiac arrest and a 5 min “no-touch period.”

A rapid laparotomy was performed, and the aorta and inferior vena cava were cannulated (14/16 French [Fr] and 20/22/24 Fr, respectively; DLP™ Medtronic, Minneapolis, MN, United States). Thoracotomy was performed, and the descending thoracic aorta was cross-clamped to prevent perfusion of the brain and coronary arteries. Subsequently, NRP was established using a compact custom closed extracorporeal circulation circuit (Rotaflow™ RF-32 centrifugal pump; Maquet, Hirrlingen, Germany). The detailed procedure for NRP control and determinations during this 1 h period are provided as Supplementary Material.

During NRP, the small bowel was prepared for procurement. After 1 h of NRP, the circuit was stopped, and the same cannulas were used for cold perfusion with cold-preservation solution (Celsior^R, IGL). The graft comprised the small bowel from the duodenum to the terminal ileum, with a vascular pedicle formed by the superior mesenteric artery and portal vein. During this cold ischemia time (CIT), the graft was kept in cold storage using the same solution until engraftment.

Heterotopic ITx was performed. The native bowel was shifted to the left side of the abdomen, and the infrarenal aorta and cava were exposed. Portocaval anastomosis, followed by anastomosis between the superior mesenteric artery and aorta, was completed, and reperfusion of the bowel was established. The proximal end of the bowel remained closed, and a terminal ileostomy was performed for graft sampling. A gastrostomy was performed to facilitate daily administration of immunosuppressive drugs, antibiotics and analgesics during the post-transplant monitoring of the animals.

After ITx, the recipients were individually isolated with controlled feed and water ad libitum. The animals’ welfare was supervised twice a day. Analgesia was provided with a fentanyl patch every 3 days for 1 week and with ibuprofen (10 mg/kg/12 h for 7 days, by tube). Cefixime 20 mg/kg was administered every 12 h. Tacrolimus was administered daily, and the doses (0.2–1 mg/kg/24 h) were adjusted according to serum levels. All drugs were administered through gastrostomy.

Ileoscopy was performed through the ostomy using a 9 mm pediatric endoscope on postoperative days (PODs) 1, 2, 7, and 14. Four biopsies were obtained during the examination. The animals were sacrificed and sampled on POD 14 by intravenous injection of 5M KCl under general anesthesia.

Graft samples were obtained at 30′ and 60′ after NRP initiation. An additional sample was obtained after flushing with the cold-preservation solution. Three intraoperative intestinal graft samples were obtained as follows: pre-reperfusion, immediately after reperfusion, and 60′ later. Samples were always taken from the ileum for uniformity. Additional intestinal samples were obtained on PODs 1, 2, 7, and 14 through the ileostomy. Blood samples were obtained at 5′, 30′, and 60′ after the beginning of NRP (donor) and on PODs 1, 2, 7, and 14 (recipient). Baseline samples were collected from each animal.

Histological analysis was performed on 5 μm hematoxylin–eosin-stained tissue sections. Ischemia-reperfusion injury (IRI) was evaluated using the Park-Chiu score (PCS), whereas the Wu score was used to evaluate rejection.

Mucosal intestinal biopsy specimens were immediately submerged in RNAlater solution (Invitrogen, Waltham, MA, United States) and stored at 4°C for 24 h and then at −80°C. RNA was isolated using the RNeasy Mini Kit (Qiagen, Hilden, Germany) and retrotranscribed into cDNA using a High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Waltham, MA, United States).

Expressions of genes for tight junction protein 1 (TJP1), epithelial cell adhesion molecule (EpCAM), mucine 2 (MUC2), lysozyme (LYZ), interleukin (IL)-6, and tumor necrosis factor alpha (TNF-α) were measured using predesigned TaqMan™ Gene Expression assays (Ss03373514_m1, Ss03384752_u1, Ss03377386_u1, Ss03394856_m1, Ss03384604_u1, Ss03391318_g1) with TaqMan™ Fast Advanced Master Mix on a 7500 Fast Real-Time PCR System (Applied Biosystems). Threshold cycle (Ct) scores were calculated as the mean of the duplicates and normalized against the Ct scores of the endogenous control GAPDH (Ss03375629_u1). Relative expression was determined as 2^−ΔCt, where ΔCt = Ct gene of interest–Ct endogenous control [16]. 2^−ΔCt values were normalized by logarithmic transformation, and fold-change values were calculated using the median value of native intestines as a reference.

Citrulline levels were measured sequentially (baseline, PODs 1, 2, 7, and 14) as a biomarker of small bowel mass and function. The absorptive function of the grafts was evaluated on POD 14, as previously reported [17]. A solution containing glucose (2 g/kg) was intraluminally administered to the transplanted intestines. The glucose level of peripheral blood (tongue) and draining veins from both the intestinal graft and native small bowel was measured (Accu-Chek blood glucose meter, Roche) immediately before and 15′, 30′, and 60′ after glucose administration to verify whether the increase was due exclusively to the absorption of the solution and not to a pre-existing blood glucose level from its native intestine.

Differences in gene expression values between grafts and native bowels were assessed using the unpaired t-test (with the Welch correction for significantly different variances). In contrast, the paired t-test was used to compare the graft samples at different time points. Spearman rho and linear regression values were calculated for correlation analysis between gene expression values and ischemia times (fWIT or CIT). Citrulline levels were compared among the different time-points using the paired t-test. Statistical significance was set at p < 0.05, and GraphPad Prism (version 8.0.2) software was used for all the tests.

Pneumothorax-induced cardiac arrest within <5′, and fWIT was 20′ ± 5′. NRP achieved good reperfusion of the abdominal organs and lactate clearance in all cases. The laboratory variables analyzed are summarized in Table 1 and Figure 1. After 1 h of NRP, 8/8 intestines were successfully obtained from DCD since all eight showed a good macroscopic appearance during graft dissection and intravascular washing. After 147.8′ ± 12′ of CIT, the grafts were implanted, achieving good graft reperfusion in 7/8 procedures (87.5%). In one case, the transplanted small bowel was poorly reperfused because of venous stenosis; hence, this animal was excluded. The remaining seven recipients recovered adequately. One animal with a gastrostomy leak required an anticipated endpoint on POD 10, while the remaining six reached the scheduled POD 14.

The distal 40–50 cm of the graft was explored. All cases showed a mucosa with normal appearance, coloration and vascularization, without erosions, ulcers or bleeding, with a well-preserved villous pattern.

As shown in Figures 2A–D, in both NRP-30′ and NRP-60′ samples, the small bowel showed a well-preserved architecture. However, two NRP-60′ samples showed edema at the villus tip (PCS 1). As expected, the highest IRI was observed after 1 h of graft reperfusion, with denuded villi appearing in 3/7 evaluated samples. Interestingly, all grafts recovered 48 h after transplantation.

Mild rejection appeared on POD 7 in 2/7 (28.5%) cases; one showed full recovery, and the other showed severe acute cellular rejection on POD 14 (1/7, 14.2%). The remaining 5/7 (71.4%) recipients showed no signs of rejection at any point (Figures 3A–D).

Relative expressions of integrin EpCAM, TJP1 (zonulin), LYZ, MUC2, the proinflammatory cytokines IL-6, and TNF-α were assessed by quantitative polymerase chain reaction in serial samples (NRP-30′, NRP-60′, pre-reperfusion, and PODs 0, 2, 7, and 14) of five animals and three additional samples of independent native bowels (Figure 4 and Supplementary Figure S1). MUC2 was excluded because it was undetectable in most samples. NRP-30′ samples showed differences in the molecular graft signature of native intestines (Figure 4) and a dramatic 350–1600-fold increase in TNF-α levels (p < 0.001). IL-6 levels also increased 2–8-fold in four of the five animals (p = 0.17). Among genes related to epithelial integrity, zonulin showed an almost significant upregulation (p = 0.06), while EpCAM appeared to be downregulated in all samples (not significant because of a high dispersion). Generally, LYZ expressions were quite similar to those in the native intestines.

Next, we evaluated the possible correlation between fWIT, CIT, NRP times and gene expression values, were no significant association was observed (Supplementary Figure S1).

In samples taken 1 h after reperfusion, gene expression values seemed to normalize at the end of the transplantation procedure (POD 0), except TNF-α, which remained high (p = 0.002). Interestingly, no relevant changes in gene expression were observed when compared to the pre-reperfusion samples, and IL-6 expression levels showed only a slight increase (p = 0.055).

During the post-transplantation period, TNF-α expression persisted above the native intestinal levels, differing significantly at every time point studied (POD 2, p < 0.001; POD 7, p = 0.002; POD 14, p < 0.001). Additionally, upregulation of this gene became evident on POD 7, becoming significant on POD 14 (p = 0.003). The maximum expression values were also observed at this point, with 1600–6700-fold changes. A similar, albeit less prominent, dynamic was observed for IL-6, which showed stable expression until POD 7, followed by an abrupt increase (except in one animal) at the endpoint (p = 0.037 versus controls).

The epithelial-related genes EpCAM and zonulin showed significant upregulation on POD 2 compared to POD 0 samples (p = 0.008 and p = 0.005, respectively). On POD 7, these differences persisted (p = 0.019 for both) and remained until POD 14 for zonulin (p = 0.042). Moreover, zonulin expression remained significantly higher than that in native bowels throughout the post-transplantation period (POD 2: p = 0.003; POD 7: p = 0.055; POD 14: p = 0.004). LYZ expression varied among the animals, and it seemed to be lower than that in native intestines; however, it was not significantly different from the POD 0 samples.

Citrulline blood levels increased significantly in all cases in the first 2 days post-ITx (p = 0.0003). Despite differences between the animals after this point, only animal 3 showed levels below the baseline, which corresponded to severe acute cellular rejection in the histological examination (Figure 5A). Graft-absorption capacity on POD 14 was evident 15′ after glucose administration, reaching maximum glucose levels at 30′. Elevated graft blood glucose levels were maintained throughout the study period (Figure 5B). In addition, blood samples from the peripheral circulation and native small bowel showed a slow and gradual increase in glucose levels, confirming that the increase in the glucose level is due to the graft’s glucose absorption and its distribution to the general circulation.