Comparison of the effects of normothermic machine perfusion and cold storage preservation on porcine intestinal allograft regenerative potential and viability

Abstract

Intestinal transplantation (IT) is the final treatment option for intestinal failure. Static cold storage (CS) is the standard preservation method used for intestinal allografts. However, CS and subsequent transplantation induce ischemia-reperfusion injury (IRI). Severe IRI impairs epithelial barrier function, including loss of intestinal stem cells (ISC), critical to epithelial regeneration. Normothermic machine perfusion (NMP) preservation of kidney and liver allografts minimizes CS-associated IRI; however, it has not been used clinically for IT. We hypothesized that intestine NMP would induce less epithelial injury and better protect the intestine’s regenerative ability when compared with CS. Full-length porcine jejunum and ileum were procured, stored at 4 °C, or perfused at 34 °C for 6 hours (T6), and transplanted. Histology was assessed following procurement (T0), T6, and 1 hour after reperfusion. Real-time quantitative reverse transcription polymerase chain reaction, immunofluorescence, and crypt culture measured ISC viability and proliferative potential. A greater number of NMP-preserved intestine recipients survived posttransplant, which correlated with significantly decreased tissue injury following 1-hour reperfusion in NMP compared with CS samples. Additionally, ISC gene expression, spheroid area, and cellular proliferation were significantly increased in NMP-T6 compared with CS-T6 intestine. NMP appears to reduce IRI and improve graft regeneration with improved ISC viability and proliferation.

1. Introduction

For patients suffering from intestinal failure (IF), nutritional autonomy is the ultimate goal given the improved patient survival demonstrated by enteral feeding and the significant influence that eating food has on improved quality of life.^1–3 Despite this, total parenteral nutrition (TPN) remains the gold standard treatment to meet IF patients’ nutritional, electrolyte, and fluid requirements.^3,4 Intestinal transplantation (IT) has been reserved for patients with irreversible IF who are unable to tolerate TPN.^5–7 While IT provides nutritional autonomy, allograft injury attributed to preservation may contribute to immune activation, higher rates of rejection and graft loss than observed in other organ transplants, and slightly worse patient survival than continued TPN in patients who lack life-threatening complications.^3,8–16

Preservation-induced ischemia and reperfusion injury (IRI) damage the intestinal epithelium, which compromises barrier function, resulting in bacterial translocation, graft dysfunction/loss of nutrient absorption, rejection, and death.^17–22 Furthermore, IRI to the epithelial crypt region diminishes the ability of the intestine to heal by inducing loss of the highly proliferative intestinal stem cells (ISCs) that are responsible for maintenance, regeneration, and repair of the intestinal epithelium, critical to graft health.^18,23,24

Acute graft rejection has been identified in 10.3% to 72% of IT recipients within 30 days of transplantation, with organ preservation injury thought to be one of the significant contributors.^8–11 Prior experiments demonstrate that the loss of barrier function correlates with the duration of cold storage (CS).^22,25 This explains the clinical finding that bacterial translocation after IT increases with length of CS.^22,26 One clinical study reported that while 14% of patients developed bacterial translocation with <7 hours of CS, bacterial translocation occurred in 76% of IT recipients with 9 hours or longer of allograft CS.²⁷ Thus, the investigation and development of alternative organ preservation and storage techniques are warranted to improve organ viability prior to transplantation, maintain graft metabolism, reduce graft injury, and overall improve transplantation success.^17,28

One such promising organ preservation method is normothermic machine perfusion (NMP), which circulates a body temperature, blood-based, oxygenated perfusate through an organ, attempting to maintain the normal metabolism of the graft.^29–31 NMP appears to reduce graft dysfunction, inhibit inflammation, and promote graft regeneration by preventing the hypothermic IRI associated with traditional CS preservation.^32–34 In clinical studies of liver transplantation, NMP preservation has been shown to be superior to CS; however, there has been very limited research focused on preservation of intestines using NMP.^14,32,33,35 In the present study, we used a porcine model of small intestinal procurement, preservation, and transplantation to evaluate the impact of CS or NMP preservation on mucosal integrity and ISC function. We hypothesized that NMP would induce less epithelial injury and better protect the intestine’s regenerative ability compared with CS.

2. Materials and methods

2.1. Experimental animals

Eight to 10-week-old Yorkshire crossbred pigs of either sex weighing between 15 and 30 kilograms (kg) were used (n = 42). Animal studies were approved by the Institutional Animal Care and Use Committee of North Carolina State University (IACUC Number 21–489) and the United States Army Medical Research and Development Command Animal Care and Use Review Office. Twenty-four pigs served as intestinal donors. The intestine from the initial 6 donor animals was used to compare tissues preserved for 6 hours of each storage method (CS and NMP), not for transplantation. The remaining 18 grafts underwent either CS (n = 10) or NMP (n = 8), followed by transplantation into recipient pigs.

2.2. Donor intestine procurement

One day prior to surgery, donor pigs were administered ceftiofur (5 mg/kg intramuscularly [IM]) and fasted for 16 to 18 hours. Anesthesia induction, intraoperative medications, intestinal procurement procedure, and graft vascular flush with the University of Wisconsin solution were performed as previously described.³⁶ All grafts underwent an initial 1-hour period of CS. The infra-renal abdominal aorta and caudal vena cava were procured for use as vascular extensions for graft implantation, and pentobarbital (100 mg/kg intravenously [IV]) was administered to complete euthanasia.

2.3. Storage protocols

2.3.1. Cold storage

After the initial 1-hour CS noted above, the allografts undergoing CS were maintained on ice and refrigerated at 4 °C in the University of Wisconsin solution for 6 additional hours (CS-T6).

2.3.2. Normothermic machine perfusion

Following the initial 1-hour CS, allografts undergoing NMP were placed onto the proprietary perfusion machine (Functional Circulation), which circulated a 34 °C, oxygenated, blood-based perfusate through the cranial mesenteric artery and a hemodiafiltration circuit for an additional preservation period of 6 hours (NMP-T6), as described.³⁶ In the current study, 2 NMP protocol generations (Gen 2 and 3) were used for jejunal samples, whereas ileal samples were obtained only from Gen 3 grafts.³⁶

2.4. Tissue collection

Jejunal and ileal intestinal biopsies were collected during the initial 1-hour period of standard CS to serve as controls (T0). Further biopsies were obtained after CS-T6 or NMP-T6 preservation and at 1 hour of reperfusion following transplantation (Fig. 1).

Figure 1.

Experimental outline. Jejunal and ileal biopsies were collected after graft procurement, after 6 hours of CS or NMP, 1 hour after in vivo reperfusion posttransplant, and after a 48-hour postoperative period. CS, cold storage; NMP, normothermic machine perfusion.

2.5. Tissue histomorphometric evaluation and immunofluorescence

Tissue for histomorphometric evaluations was fixed in 10% neutral buffered formalin, embedded in paraffin, sectioned (approximately 5 to 7 μm), mounted, and stained with hematoxylin and eosin. A board certified pathologist with experience assessing intestinal transplant biopsies evaluated slides blinded to time and storage conditions using the Chiu/Park injury scale.³⁷

Immunofluorescence staining was performed for KI67, a cellular proliferation protein; sex-determining region Y-box 9 (SOX9), a protein associated with ISCs, Paneth cells, and transit amplifying cells; and cleaved caspase 3 (CC3), a marker for cellular apoptosis, as previously described.³⁸ All positive cells were counted per crypt or crypt-villus axis, as appropriate.

2.6. Intestinal crypt isolation, real-time quantitative polymerase chain reaction, culture, spheroid/enteroid area measurements, and whole mount immunofluorescence

Crypt isolation was performed as previously described for RNA extraction or cell culture.^38,39 Crypts for RNA extraction were snap-frozen in liquid nitrogen and stored at −80 °C.

Total RNA was extracted (Qiagen RNeasy Minikit; Qiagen, Valencia, CA), yield and quality were determined (NanoDrop Technologies, Thermo Fisher Scientific, Waltham, MA; 260/280 nm = 2.03 to 2.07), and complementary DNA conversion was performed (iScript cDNA synthesis kit, Bio-Rad, Hercules, CA). Quantitative real-time polymerase chain reaction was used to determine changes in gene biomarker expression of ISC populations (leucine-rich repeat-containing G-protein coupled receptor 5 [LGR5], olfactomedin 4 [OLFM4], SOX9, and homeodomain-only protein homeobox [HOPX]) and cellular proliferation genes (proliferating cell nuclear antigen [PCNA], KI67, and yes-associated protein 1 [YAP1]) (iTaq Universal SYBR green Supermix; Bio-Rad, Hercules, CA; QuantStudio 6 Flex; Applied Biosystems, Waltham, MA). GAPDH was used as the housekeeping gene. Primer sequences are described.^38,40,41 The ΔΔC_t method was used to measure relative gene expression changes, testing samples in triplicate.

Crypt culture was performed as previously described.^38,40 Area measurements were obtained from 10 to 20 randomly selected spheroids/enteroids daily for 48 hours (Freehand Polygon Selection Tool and Measurement Function; NIH ImageJ, Bethesda, MD). Area measurements were averaged.

To identify proliferating cells, cultures were incubated with 5-ethynyl-2’-deoxyuridine (EdU Click-iT, Thermo Fisher Scientific, Waltham, MA) as previously described.⁴² The number of EdU-positive cells from at least 10 spheroids/enteroids was counted.

2.7. Image acquisition

Images were captured using an inverted fluorescence microscope (Olympus IX83, Tokyo, Japan) fitted with a monochrome digital camera (ORCA-flash 4.0, Hamamatsu, Shizuoka, Japan) and color camera (DP26, Olympus, Tokyo, Japan). The objective lenses used were X10, X20, and X40 (numerical apertures of 0.3, 0.45, and 0.6, respectively) (LUC Plan FLN, Olympus, Tokyo, Japan).

2.8. Porcine transplantation—graft implantation, postoperative care, and euthanasia

Recipient pigs received the same preoperative medications and management as donor pigs, with the addition of tacrolimus (0.2 mg/kg subcutaneously) once a day starting 1 day before surgery.³⁶ Subsequent tacrolimus dosing was determined based on daily measurements of whole blood concentrations, with a target trough level of 8 to 12 ng/mL.⁴³ General anesthesia and surgical approach were the same as for donor pigs. Intestinal resection was performed (ENSEAL X1, Ethicon, Raritan, NJ) across the length of the mesentery, discarding the native intestine. The donor-derived intestine allograft vasculature was flushed with 500 mL of cold 5% albumin immediately before transplantation.

Vascular extensions were anastomosed end-to-side to the recipient’s abdominal aorta and/or caudal vena cava, when necessary. Vascular end-to-end anastomoses were then performed, and the transplanted allograft underwent reperfusion in vivo. Five minutes after reperfusion, 20 mg/kg methylprednisolone was administered IV.³⁶ Donor-to-recipient end-to-end small bowel anastomoses and routine abdominal closure were performed.

Postoperative pain management included 1 to 5 mg/kg of bupivacaine subcutaneously at skin closure, lidocaine-ketamine constant rate infusion, a fentanyl transdermal patch, and/or buprenorphine (0.01 mg/kg IV every 12 hours). Additional medications included maropitant (1.0 mg/kg IV every 24 hours) and omeprazole (20 mg orally every 24 hours). Heparin or enoxaparin (heparin, 33 IU/kg every 12 hours or 80 IU/kg bolus plus 18 IU/kg/h constant rate infusion; enoxaparin, 1 to 2 mg/kg subcutaneously every 12 hours) was administered for thrombosis prophylaxis with dosage adjusted based on measurement of activated clotting times (target of approximately 180 to 400 seconds). Blood samples were drawn daily to measure activated clotting times, lactate, complete blood count, and blood serum chemistry, as described previously.³⁶ In addition to continuing the tacrolimus therapy, methylprednisolone was administered in a taper fashion. Physical examinations, enteral feeding, and fecal production were closely monitored. Hill’s Prescription Diet a/d Urgent Care wet food (Hill’s Pet Nutrition) was offered to the recipient pigs every 6 hours following recovery from anesthesia. At the 48-hour postoperative endpoint (or when noted to be in distress), animals were anesthetized with xylazine (1.5 mg/kg IM) and ketamine (11 to 20 mg/kg IM) and euthanized using 100 mg/kg pentobarbital.

2.9. Statistics

Statistical analyses were performed using Prism (GraphPad Software, La Jolla, CA) software. Outliers were identified using the robust regression and outlier removal method. A Shapiro-Wilk normality test was performed on raw data, followed by either a Kruskal-Wallis test with Dunn’s multiple comparisons, 2-way analysis of variance with Tukey’s multiple comparisons, or 2-way analysis of variance with Dunnett’s multiple comparisons. Statistical significance was defined as P <.05.

3. Results

3.1. Histologic assessment of CS and NMP intestines immediately following storage and 1-hour postreperfusion

Histologic comparison of the intestinal biopsies obtained at T0, NMP-T6 or CS-T6, and after reperfusion in transplanted allografts was performed to determine differences in intestinal injury induced by each storage method (Figs. 2 and 3). Jejunal Chiu/Park injury grades (median, range) were increased in NMP-T6 treated allografts (median = 2, range = 0–3; P = .0003) compared with T0 (median = 0, range = 0–2), consistent with mild epithelial loss from villi tips (Fig. 2B). There was no difference between CS-T6 (median = 1, range = 0–3) and NMP-T6 injury grades (P =.1966).

Figure 2.

Histologic images and pathologic injury scoring of the jejunum after 6 hours of cold storage (CS) or normothermic machine perfusion (NMP) and after 1 hour of reperfusion posttransplant. (A) Jejunal hematoxylin and eosin-stained sections of T0 (control), after 6 hours of CS or NMP (CS-T6, NMP-T6), and after 1 hour of reperfusion posttransplant were evaluated for differences in intestinal injury following each storage method. (B) NMP-T6 jejunum had elevated Chiu/Park injury grades compared with control (T0), consistent with mild to moderate villous epithelial loss. After 1 hour of reperfusion, transplanted CS jejunum had a significantly greater injury than that of control (T0) or after 6 hours of CS (CS-T6). Gen 2 and Gen 3 are the second and third generations of NMP protocols. T0 = 20 pigs, CS-T6 = 11 pigs, and NMP-T6 = 9 pigs (Gen 2 = 4 pigs, Gen 3 = 5 pigs). After 1 hour of reperfusion: CS = 7 pigs, NMP = 7 pigs (Gen 2 = 1 pig, Gen 3 = 6 pigs). Original magnification: 10×, scale bar: 100 μm. *P ≤.05, **P ≤ 0.01, ***P ≤.001, and ****P ≤.0001.

Figure 3.

Histologic images and pathologic injury scoring of the ileum after 6 hours of cold storage (CS) or normothermic machine perfusion (NMP) and after 1 hour of reperfusion posttransplant. (A) Ileal hematoxylin and eosin-stained sections of T0 (control), after 6 hours of CS or NMP (CS-T6, NMP-T6), and after 1 hour of reperfusion posttransplant were evaluated for differences in intestinal injury following each storage method. (B) The NMP-T6 ileum had elevated Chiu/Park injury grades compared with the control (T0) and ileum that underwent 6 hours of CS (CS-T6), consistent with mild villous epithelial loss. After 1 hour of reperfusion, transplanted CS ileum had significantly greater injury than that of the control (T0). Gen 3 is the third generation of NMP protocols. T0 = 15 pigs, CS-T6 = 7 pigs, and NMP-T6 = 4 pigs (all Gen 3). After 1 hour of reperfusion: CS = 7 pigs, NMP = 5 pigs (all Gen 3). Original magnification: 10×, scale bar: 100 μm. *P ≤.05, **P ≤.01, ***P ≤.001, and ****P ≤.0001.

After in vivo reperfusion following transplantation, a progression of moderate-to-severe histologic injury was observed in CS jejunum (median = 3, range = 2–4), which was significantly worse than T0 (P <.0001) and CS-T6 (P =.0009). In contrast, no histologic evidence of reperfusion injury was observed in NMP transplants (median = 2, range = 0–4). Similar findings were observed in ileal-derived biopsies (Table 1; Fig. 3).

Table 1.

Chiu/Park injury scoring of ileum (median and range) after 6 hours of CS or NMP and after 1 hour of reperfusion posttransplant. Data from control (T0), cold storage (CS-T6), and normothermic machine perfusion (NMP-T6) ileum.

Ileal Chiu/Park injury grades
	Control	Cold storage		Normothermic machine perfusion
	T0	CS-T6	Reperfusion	NMP-T6	Reperfusion
Median	0	0	1	1.5	1
Range	0–1	0–0	0–3	0–3	0–4

CS-T6, 6 hours of cold storage; NMP-T6, 6 hours of normothermic machine perfusion; T0, control tissue.

3.2. Gene expression analysis of ISC and proliferating cell biomarkers

To evaluate the impact of each storage method on tissue regenerative potential, gene expression analyses for stem/progenitor and proliferating cell biomarkers were performed on T0, NMP-T6, and CS-T6 jejunal and ileal-derived crypts (Fig. 4). In the jejunum at CS-T6, LGR5, SOX9, and PCNA were significantly increased (P =.0449; P <.0001; P =.0406, respectively) compared with T0 (Fig. 4A). In NMP-T6 jejunum, OLFM4, HOPX, PCNA, KI67, and YAP1 were increased (P = .0122; P = .0398; P =.0179; P <.0001; P =.0014, respectively) compared with T0. Jejunal CS-T6 SOX9 was increased compared with NMP-T6 (P < .0001), whereas HOPX and KI67 were increased in NMP-T6 compared with CS-T6 (P = .0441; P < .0001, respectively). Similar increases were measured in reserve ISC and cellular proliferation gene biomarkers in NMP-treated, ileal-derived crypts compared with T0 and CS-T6 (Fig. 4B).

Figure 4.

Gene expression analysis of intestinal stem cell and proliferating cell biomarkers after 6 hours of cold storage (CS) or normothermic machine perfusion (NMP). (A) At CS-T6, LGR5, SOX9, and PCNA were significantly increased in jejunal crypts compared with T0. In jejunum, at NMP-T6, OLFM4, HOPX, PCNA, KI67, and YAP1 all increased compared with T0. In jejunum, at CS-T6, SOX9 was increased compared with NMP-T6, whereas jejunal HOPX and KI67 were increased in NMP-T6 compared with CS-T6. (B) In the NMP-T6 ileum, KI67 and YAP1 were increased compared with T0. Ileal KI67, YAP1, and HOPX were increased in NMP-T6 compared with CS-T6. Jejunum and ileum: T0 = 12 pigs, CS-T6 = 7 pigs, NMP-T6 = 9 pigs. *P ≤.05, **P ≤.01, ***P ≤.001, and ****P .0001. CS-T6, 6 hours of cold storage; NMP-T6, 6 hours of normothermic machine perfusion.

3.3. Immunofluorescence analysis of ISC, cellular proliferation, and apoptosis protein biomarkers

Due to the increased gene expression associated with ISC and proliferation observed predominantly within the jejunum, further analysis of protein expression (mean ± standard deviation) associated with cellular processes of regeneration and apoptosis was performed using immunofluorescence on associated tissue sections (Fig. 5). At NMP-T6, there was no difference in the number of KI67⁺ proliferating cells compared with T0 (NMP-T6, 40.57 ± 10.61; T0, 38.15 ± 8.856) (Fig. 5B); however, the number of KI67⁺ cells was significantly lower at CS-T6 (33.26 ± 7.686) compared with T0 (P =.0293). The number of SOX9⁺ cells was significantly increased in NMP-T6 (58.34 ± 18.16) compared with T0 (47.22 ± 22.67, P = .0269), but no significant difference was identified between CS-T6 and T0 (P > .9999). Both the number of KI67⁺ and SOX9⁺ cells in NMP-T6 were significantly increased compared with CS-T6 (KI67⁺, P = .0065; SOX9⁺ CS-T6: 42.44 ± 16.64, P = .0078) (Fig. 5A, B). Cellular apoptosis was significantly increased at CS-T6 (7.545 ± 4.751) compared with T0 (2.339 ± 2.065) and NMP-T6 (4.323 ± 2.914) (P <.0001 and P =.0493, respectively), while the number of CC3⁺ cells in NMP jejunum was only mildly elevated compared with T0 (P =.0089) (Fig. 5C, D).

Figure 5.

Immunofluorescence analysis of the jejunum after 6 hours of storage and after 1 hour of in vivo reperfusion posttransplant. (A) The expression of KI67 (green) and SOX9 (red) was evaluated between T0, CS-T6, and NMP-T6. (B) At NMP-T6, there was no significant difference in the number of KI67⁺ cells compared with T0, whereas the number of KI67⁺ cells was significantly lower at CS-T6 compared with T0. The number of SOX9⁺ cells was increased in NMP-T6 compared with T0. Both the number of NMP-T6 KI67⁺ and SOX9⁺ cells were significantly increased compared with CS-T6. (C) Expression of CC3 (red) was evaluated between T0, CS-T6, NMP-T6, and after 1 hour of in vivo reperfusion posttransplant. (D) The number of CC3⁺ cells was significantly increased at CS-T6 compared with T0 and NMP-T6, while the number of NMP-T6 CC3⁺ cells was only mildly elevated compared with T0. After 1 hour of reperfusion, there was a significantly increased number of CC3⁺ cells in the CS-T6 jejunum compared with T0 and NMP-T6. NMP-T6 was decreased in CC3⁺ cells compared with T0. All sampling timepoints KI67, SOX9: n = 7 pigs. T0 and T6 CC3: n = 7 pigs. After 1 hour of reperfusion CC3: n = 14 pigs. KI67, SOX9 original magnification: 40×, scale bar: 20 μm. CC3 original magnification: 10×, scale bar: 100 μm. *P ≤ .05, **P ≤ .01, ***P ≤.001, and ****P ≤ .0001. CC3, cleaved caspase 3; CS, cold storage; CS-T6, 6 hours of cold storage; NMP, normothermic machine perfusion; NMP-T6, 6 hours of normothermic machine perfusion; SOX9, sex-determining region Y-box 9; T0, control tissue.

Following transplantation and in vivo reperfusion, KI67⁺ and SOX9⁺ cells at NMP-T6 (KI67⁺, 52.52 ± 15.42; SOX9⁺, 83.68 ± 25.84) were significantly elevated compared with T0 (P = .0003 and P <.0001, respectively), whereas the number of KI67⁺ cells in CS jejunum (29.17 ± 11.85) was significantly decreased (P = .005). Both KI67⁺ and SOX9⁺cells at NMP-T6 were significantly elevated compared with CS-T6 (P <.0001). There was an increase in the number of CC3⁺ cells in CS jejunum (19.74 ± 9.454) after 1 hour of in vivo reperfusion compared with T0 and NMP-T6 (P < .0001, both), while there was a decrease (P = .0004) in apoptotic cells in NMP jejunum (0.6162 ± 0.752) compared with T0 (2.339 ± 2.065) (Fig. 5C, D).

3.4. Three-dimensional crypt culture as a functional analysis of regenerative potential

Although compelling, gene and protein expression are not functional analyses of ISC proliferative potential. Crypt cultures were therefore utilized to evaluate spheroid/enteroid growth. At 24 and 48 hours of cell culture, jejunal spheroid area measurements (μm²; mean ± standard deviation) were significantly larger from NMP-T6-derived crypts (24 hours: 6433 ± 5110; 48 hours: 13,077 ± 9683) compared with those derived from T0 (24 hours: 1210 ± 1816; 48 hours: 2650 ± 4583) or CS-T6 (24 hours: 2067 ± 3473; 48 hours: 3322 ± 4774) (all, P <.001) (Fig. 6).

Figure 6.

Jejunal intestinal stem cell culture spheroids and area measurements. (A) Images of 24- and 48-hour cell culture for spheroids from control (T0) jejunum, and jejunum stored for 6 hours (CS-T6 and NMP-T6). Spheroids cultured from transplanted jejunum that underwent 1 hour of in vivo reperfusion are not pictured. (B) At 24- and 48-hours of cell culture, jejunal spheroid area measurements were significantly larger from NMP-T6-derived crypts compared with crypts from T0 and CS-T6 jejunum. After 1 hour of reperfusion posttransplant, NMP-stored jejunal spheroid area measurements were significantly larger than T0, NMP-T6, and 1 hour reperfused CS jejunum. Additionally, after 1 hour of reperfusion posttransplant, CS jejunal spheroid area measurements were significantly larger than T0 and CS-T6. T0 = 15 pigs, CS-T6 = 11 pigs, and NMP-T6 = 10 pigs. After 1 hour of reperfusion: CS = 6 pigs, NMP = 4 pigs at 24 hours, and 5 pigs at 48 hours of cell culture. Original magnification: 20×, scale bar: 50 μm. *P ≤.05, **P ≤.01, ***P ≤.001, and ****P ≤.0001. CS, cold storage; CS-T6, 6 hours of cold storage; NMP, normothermic machine perfusion; NMP-T6, 6 hours of normothermic machine perfusion; T0, control tissue.

After 1 hour of reperfusion posttransplant, NMP-stored jejunal spheroid area measurements (24 hours: 15 347 ± 10 899; 48 hours: 20 433 ± 12 804) were significantly larger than T0 (24 and 48 hours: P < .001), NMP-T6 (24 hours: P = .0492), and transplanted CS (24 hours: 4003 ± 3147; P <.001. 48 hours: 4676 ± 3967; P <.001) (Fig. 6B). Additionally, CS jejunal spheroid area measurements were significantly larger than T0 and CS-T6 (24 and 48 hours: P <.001). Similar increases in spheroid sizes were appreciated in crypts derived from ileal biopsies (Fig. 7 and Table 2).

Figure 7.

Ileal intestinal stem cell culture spheroids and area measurements. (A) Images of 24- and 48-hour cell culture for spheroids from control (T0) ileum, and ileum stored for 6 hours (CS-T6 and NMP-T6). Spheroids cultured from transplanted ileum that underwent 1 hour of in vivo reperfusion are not pictured. (B) At 24- and 48-hours of cell culture, ileal spheroid area measurements were significantly larger from NMP-T6 derived crypts compared with crypts from T0 and CS-T6 ileum. After 1 hour of reperfusion posttransplant, NMP-stored ileal spheroid area measurements were significantly larger than T0 and reperfused CS jejunum. T0 = 14 pigs, CS-T6 = 11 pigs, and NMP-T6 = 9 pigs. After 1 hour of reperfusion: CS = 6 pigs and NMP = 5 pigs. Original magnification: 20×, scale bar: 50 μm. *P ≤.05, **P ≤.01, ***P ≤.001, and ****P ≤.0001. CS, cold storage; CS-T6, 6 hours of cold storage; NMP, normothermic machine perfusion; NMP-T6, 6 hours of normothermic machine perfusion; T0, control tissue.

Table 2.

Three-dimensional crypt culture. Data from control (T0), cold storage (CS-T6), and normothermic machine perfusion (NMP-T6) ileum. Ileal intestinal stem cell culture spheroids area measurements (μm²; mean ± standard deviation).

Ileal three-dimensional crypt culture
Spheroid area measurements (μm2; mean ± standard deviation)
Cell culture duration	Control	Cold storage		Normothermic machine perfusion
	T0	CS-T6	Reperfusion	NMP-T6	Reperfusion
24 h	4506 ± 3505	5392 ± 4157	4676 ± 2949	7626 ± 6,317	13 757 ± 8445
48 h	9895 ± 7203	11 042 ± 6755	6653 ± 4341	15 899 ± 11 872	21 545 ± 11 517

CS-T6, 6 hours of cold storage; NMP-T6, 6 hours of normothermic machine perfusion; T0, control tissue.

To confirm that the increase in spheroid area was due to increased cellular proliferation vs increased cell size, EdU proliferation assays were performed. A greater number of EdU⁺ cells (mean ± standard deviation) were quantified in NMP-T6 jejunal enteroids (38.14 ± 32.48) compared with T0 (6.1 ± 6.027, P <.0001) or CS-T6 enteroids (4.745 ± 5.735, P <.0001) (Fig. 8), with similar results identified in ileal cell culture (Fig. 8 and Table 3).

Figure 8.

5-ethynyl-2’-deoxyuridine (EdU) proliferation assay images and the number of EdU⁺ cells after 6 hours of jejunal and ileal CS or NMP. (A) EdU (purple) whole mount immunofluorescence on jejunal and ileal T0, CS-T6, and NMP-T6 spheroids. (B) NMP-T6 jejunal and ileal spheroids contained a great number of EdU⁺ cells compared with T0 or CS-T6. T0 = 10 pigs, CS-T6 = 7 pigs, and NMP-T6 = 8 pigs. Original magnification: 20×, scale bar: 50 μm. *P ≤.05, **P ≤.01, ***P ≤.001, and ****P ≤.0001. CS, cold storage; CS-T6, 6 hours of cold storage; NMP, normothermic machine perfusion; NMP-T6, 6 hours of normothermic machine perfusion; T0, control tissue.

Table 3.

EdU proliferation assay. EdU proliferation assay was used to determine the number of EdU⁺ cells (mean ± standard deviation) in spheroids cultured from T0, CS-T6, and NMP-T6 ileal crypts.

Ileal EdU proliferation assay
Number of EdU+ cells (mean ± standard deviation)
Cell Culture Duration	T0	CS-T6	NMP-T6
48 h	24.88 ± 21.32	10.81 ± 10.25	39.58 ± 27.19

CS-T6, 6 hours of cold storage; EdU, 5-ethynyl-2’-deoxyuridine; NMP-T6, 6 hours of normothermic machine perfusion; T0, control tissue.

3.5. Recipient recovery following transplantation of CS or NMP-stored intestine

All 18 recipient pigs successfully recovered from anesthesia following IT. Six of the 8 (75%) pigs that were transplanted with NMP-stored intestine, compared with only 3 of the 10 (30%) pigs transplanted with CS intestine, survived to the designated euthanasia time. Necropsy examination of the 9 recipient pigs that were euthanized 48 hours postoperatively revealed these pigs suffered from complications, including vascular thrombosis (n = 4 [1 NMP and 3 CS]), intestinal strangulation (n = 2 [1 NMP and 1 CS]), intra-abdominal hemorrhage (n = 1 CS), native jejunum mesenteric rent (n = 1 CS), and unknown cause (n = 1 CS).

4. Discussion

IRI appears to be a major contributor to short- and long-term intestine allograft injury.^20,44 In this study, we utilized a porcine model to examine the effects of NMP compared with CS for intestinal preservation, hypothesizing that NMP plays a role in mitigating IRI-associated allograft damage and may improve organ preservation. The superiority of NMP for the preservation of the liver has led to US Food and Drug Administration-approved, clinically used devices; however, limited research has focused on intestinal preservation.^14,32,33,35 Our research differs significantly from previous studies that predominantly used short, singular segments of either jejunum or ileum, nonblood-based perfusates, only histology to evaluate tissue injury, and did not perform transplantation.^36,45–47 We recently published the first protocol using NMP preservation for successful IT in a porcine model.³⁶ This study addressed the more complex metabolic and physiological considerations necessary for successful intestinal preservation compared with liver, which includes regional variability in the vasoactivity of mesenteric arterial supply and the absence of lactate clearance. Here, we expand on these initial findings to directly compare the clinical outcomes of transplantation after CS to NMP-preserved intestine allografts. We chose the total preservation time of 7 hours in this study as it is less than the previously established limits of 8 hours of CS for optimal outcomes in clinical reports of human IT.^27,48,49

This is the first study to demonstrate the clinical superiority of intestine NMP preservation compared with CS. Of the 18 pigs in this study that underwent IT, the majority of animals that did not survive to the study endpoint were those with CS allografts. Although our initial choice of preservation time was based on human clinical studies of IT suggesting 8 hours was the upper limit of safe preservation, in retrospect, the poor porcine post-transplant survival after intestinal CS may be explained by species differences.⁵⁰

Two different NMP protocols (Gen 2 and 3) were used during this study, which have been previously described.³⁶ During optimization of our NMP protocol, we found that while Gen 2 appropriately perfused jejunal tissue, the ileum had variable degrees of suspected hypoperfusion.³⁶ Vasodilatory modifications in the Gen 3 protocol resulted in a grossly normal and transplantable full-length intestine.³⁶ As Gen 2 had no obvious negative effects on the jejunum, jejunal data include both Gen 2 and 3, while ileal data were procured using only Gen 3.

Histology of the intestine allograft immediately following CS (CS-T6) in this study demonstrated minimal evidence of injury based on Chiu/Park grading, similar to biopsies obtained at graft procurement (T0). This was expected, as the method of CS preservation for transplantation is based on the physiological effect of decreased cellular metabolic rate, which serves to delay the mechanisms of ischemic injury.^28,51–53 Furthermore, CS solutions are specifically formulated to help mitigate IRI, with experimental studies using human intestines demonstrating initial signs of mild histologic injury after 6 hours and more severe injury with epithelial detachment extending down the villi and diffuse damage into the crypts not identified until 9 and 12 hours of CS, respectively.⁴⁸ However, despite the minimal evidence of histologic change noted using the Chiu/Park scoring system, the observation of increased cellular apoptosis in these tissues provided evidence of CS-associated damage. The most significant damage observed in the current study was in CS allografts after reperfusion. In these CS allografts, there was severe epithelial injury compared with NMP-preserved allografts. Furthermore, the subsequent increase in CC3⁺ cells in CS allografts following reperfusion indicated a progression of damage that was not observed in the NMP intestine. Cold storage has been found to increase caspase-3 activity and the number of apoptotic cells in the liver, kidney, and intestine, with a significant correlation between the degree of apoptosis and the duration of cold ischemia.^32,54–58 In contrast, normothermic regional perfusion is thought to have alleviated IRI in intestinal allografts by reducing caspase-3 expression in enterocytes posttransplant, diminishing cellular apoptosis and the loss of intestinal epithelial cells, and resulting in improved graft function.⁵⁷ We suspect that decreased epithelial cell apoptosis is one of the protective mechanisms of NMP that results in a healthier graft. It is possible that the more severe IRI in the CS intestine contributed to graft thrombosis and poorer survival of these transplant recipients because IRI is known to promote endothelial activation and increased leukocyte and platelet adhesion, leading to vascular thrombosis.^13,59–62 However, the other causes for early euthanasia in our pigs, such as intestinal strangulation, intra-abdominal hemorrhage, and native jejunum mesenteric rents, are unlikely to be a result of allograft IRI. Furthermore, no evidence of histologic graft rejection was observed in any samples obtained from recipient animals.

This study is the first to evaluate the effects of preservation method on ISC. The ISC are crypt-based columnar cells that are responsible for differentiation, proliferation, and migration of all intestinal epithelial cell types and are crucial for regeneration of the epithelium after injury. In this study, we found that NMP significantly increased the expression of genes and proteins associated with ISC and cellular proliferation. We attribute these findings to reparative functions likely activated following the initial procurement that was noted by the mild histologic injury observed at NMP-T6. The increase in HOPX expression from NMP-preserved jejunal crypts suggests activation of the reserve population of ISC, which did not occur in the CS-preserved intestine.⁴² Furthermore, increases in OLM4 have not only been associated with ISC activation but have also been associated with anti-inflammatory and antiapoptotic effects, aligning with the observed antiapoptotic effect of NMP in our findings.⁶³ The increase in some genes associated with ISC and proliferation in CS-T6 compared with T0 was unexpected. However, it is consistent with the notion that there is an inevitable degree of epithelial injury associated with procurement and therefore a potential for ISC activation. Yet, instead of culminating in active repair, as indicated by increases in ISC-associated protein expression in NMP allografts, further damage ensued following transplantation.

Ultimately, 3-dimensional crypt culture was used as a functional analysis to confirm ISC activation and cellular proliferation. It has been previously shown that crypts collected from in vivo ischemic-injured intestines can be placed in culture to assess the postinjury viability of ISC.⁴⁰ In this study, similar to the prior description of warm ischemic intestinal injury, we demonstrated that spheroid area measurements from CS-derived crypts were smaller at 24 hours than spheroids derived from NMP-preserved intestine. This suggests a more severe intestinal ischemic injury to the ISC from CS. This was further supported by the lower rate of EdU incorporation in the cultured crypts from the CS intestine, indicating decreased ISC proliferative potential. These findings suggest that NMP, by improving preservation of ISC, facilitates cellular regeneration and repair of epithelial tissues inevitably injured during procurement and the brief period of CS.

This is the first IT study to show the superiority of NMP to CS preservation, as most dramatically noted in molecular and cellular assessments. Survival after transplantation of the NMP intestine was notably better than CS, likely due to differences in the degree of IRI associated with different storage methods. However, further studies using more animals are required to confirm an association between NMP and improved survival posttransplant. Potential mechanisms of NMP protection identified here include suppression of cellular apoptosis and increased ISC proliferative potential, represented by the improved preservation of epithelial architecture, increased expression of ISC and cellular proliferation genes and proteins, and the robust size and number of proliferating cells in the NMP spheroids compared with CS. Our results indicate that, compared with CS, NMP may improve graft regenerative potential, resulting in transplantation of healthier bowels and superior recipient survival. Future research investigating the effects of intestinal NMP on the prolongation of safe storage durations and potential to repair ischemic intestine obtained via donation after hypotensive episodes or cardiac death is warranted to further advance intestinal preservation and transplantation.

Abbreviations:

CC3

cleaved caspase 3

CS

cold storage

CS-T6

6 hours of cold storage

EdU

5-ethynyl-2’-deoxyuridine

Gen

generation

HOPX

homeodomain-only protein homeobox

IACUC

Institutional Animal Care and Use Committee

IF

intestinal failure

IM

intramuscular

IRI

ischemia-reperfusion injury

ISC

intestinal stem cell

IT

intestinal transplantation

IV

intravenously

LGR5

leucine-rich repeat-containing G-protein coupled receptor 5

NMP

normothermic machine perfusion

NMP-T6

6 hours of normothermic machine perfusion

OLFM4

olfactomedin 4

PCNA

proliferating cell nuclear antigen

SOX9

sex-determining region Y-box 9

T0

time 0, control tissue

TPN

total parenteral nutrition

YAP1

yes-associated protein 1

Data Availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.