Intestinal Mucosal Perfusion and Integrity Are Maintained in Hypotensive Brain Dead Mice

Brain-dead individuals represent the main source of organs for transplantation. The period of brain death (BD) is characterized by a cytokine surge and a systemic inflammatory response syndrome as well as frequent periods of hemodynamic instability [1, 2]. Following a brief, initial period of hypertension with bradycardia immediately after brain herniation (the Cushing reflex), the loss of sympathetic activity results in vasodilatation, impaired cardiac output and potential hypoperfusion [3]. This commonly seen hemodynamic instability may result in morphologic and functional damage in various organs [4, 5]. Additionally, the cytokine surge may lead to inflammatory-activated organs [6, 7]. The complex alterations occurring between the time of brain herniation and organ procurement and its negative influence on organ quality prior to graft recovery explain, in part, the inferior results obtained after transplantation of organs from deceased donors compared with organs from living donors.

The small intestine is very sensitive to hypoperfusion and ischemia, and has been identified as the motor of multiple organ dysfunction syndrome during critical illness and shock [8]. The intestinal mucosa receives about 65% of the total intestinal blood flow and numerous animal and human studies have shown an impaired gastrointestinal mucosal blood flow as well as regional redistribution of blood during shock [9, 10]. Splanchnic ischemia and reperfusion injury may compromise the barrier function of the intestinal mucosa, leading to bacterial translocation, immune activation and subsequent development of systemic inflammatory response syndrome. Interestingly, several findings such as increased polymorphonuclear cell count and upregulated vascular adhesion molecules and interleukin-6 have been reported both in the intestines of brain-dead rats [11] as well as in rats with hemorrhagic [12] or septic shock [13, 14]. Hence, it is tempting to extrapolate that BD and sepsis share numerous common features, and altered mucosal perfusion and significant epithelial injury may be some of them.

Despite the critical role of an intact intestinal mucosa for organ donors in general and intestinal donors in particular, data on intestinal mucosal perfusion and an in-depth molecular assessment of the integrity of the intestinal mucosa following BD is lacking. The absolute majority of multi-organ donors largely shows macroscopically and microscopically healthy intestines at organ retrieval [15]. However, experimental observations allude that brain dead rats may have impaired microvascular perfusion in the mesentery [16]. Given the insufficient and conflicting knowledge about the intestine with regard to mucosal perfusion and the molecular structure of the mucosal barrier, we elucidated these aspects in a murine model of BD.

The study protocol was reviewed and approved by the animal experiments Ethics Committee of the Gothenburg University (Dnr. 238-2004) and conducted in accordance with European and National Institutes of Health guidelines for use of experimental animals. Adult BalbC male mice, aged 6–8 weeks and weighing 25–30 g (M&B A/S, Ry, Denmark) were housed in a room with access to food and water ad libitum, controlled temperature and 12-h light-dark cycles.

After an acclimatization period for at least 1 week before the experiments, animals were anesthetized and analgesia was induced with a mixture of hypnorm and diazepam (0.2 mL of each) injected intraperitoneally. Left carotid artery was cannulated for hemodynamic monitoring. In animals assigned for the brain-dead group (n = 10), a frontolateral trepanation (Ø 1 mm) was performed using a dental drill and a 3 F embolectomy catheter (Edwards Lifesciences, Irvine, CA, USA) was inserted into the extradural space with the tip pointing caudally. The balloon was then inflated over 1 min to increase the intracranial pressure, thereby inducing rapidly progressive brain injury and immediate BD. Initiation of BD was defined by a sharp rise and then a subsequent drop of blood pressure and heart rate. The state of BD was confirmed by the absence of corneal reflexes and by an apnea test. After BD induction, animals were ventilated with air through the tracheotomy (MiniVent 845; Hugo Sachs Elektronik, March-Hugstetten, Germany) with a tidal volume 150 μL at 200 breaths/min. Throughout the experiment animals were kept normothermic (37°C) by means of a heating pad. Mice in the control group (n = 10) underwent tracheotomy, carotid cannulation, and scalp incision as well as laparotomy and mucosal perfusion measurements. All animals were monitored for 2 h, followed by euthanasia and tissue sampling. Additionally, three unoperated mice euthanized under isoflurane anesthesia provided normal values.

Intestinal blood flow at capillary level was assessed using laser-Doppler flowmetry, as described earlier [17]. After a midline laparotomy, a flexible, customized fiberoptic probe (Ø 1 mm) connected to a Periflux^® 4001 Laser-Doppler unit (PeriMed, Stockholm, Sweden) was gently applied on the jejunal serosa, with minimal manipulation of the small intestine. Mucosal perfusion was assessed every 15 min for 2 h after BD induction/sham operation and expressed as arbitrary perfusion units. The signal was checked for artifacts caused by inappropriate contact of the probe with the tissue or by bowel movements. A measurement was considered adequate when the signal was stable for at least 10 s and without artifacts, and when the total backscattered light was constant. Between 3 and 5 satisfactory measurements were obtained at different bowel levels and a mean perfusion value was calculated. The abdomen was temporarily covered with a plastic sheet between measurements and warm saline was dropped in the abdominal cavity to prevent desiccation.

Formalin-fixed tissue was paraffinized, embedded, and cut in five-micron sections. Sections were then stained with hematoxylin and eosin. Intestinal morphology was assessed blindly by one experienced observer using the Chiu-Park score [18] on seven fields selected randomly from three different sections. Morphometric analysis (villus length, surface, volume) was performed using the Axioscan software as described earlier [19].

Five-micron thick tissue sections were deparaffinized and rehydrated; then, antigen retrieval was performed using pressure cooking in citrate buffer as previously described [20]. In brief, slides were incubated overnight at 4°C with polyclonal antibody against claudin-3 (1:100; Abcam, Amsterdam, The Netherlands). Thereafter, slides were incubated with a secondary antibody conjugated with Alexa 594 (1:500; Invitrogen), counterstained with 4′6′-diamidino-2-phenylindole (DAPI), mounted with aqueous mounting medium (Vector Laboratories, Burlingame, CA, USA) and scanned using the ZEISS Axioscan 7 (Carl Zeiss AG, Jena, Germany) slide scanner. Image acquisition and processing were performed using the Axioscan software.

Gene expression was studied in intestinal samples from both experimental groups. Briefly, frozen tissues were cut into 25–30 mg pieces, and the total RNA was isolated using the RNeasy mini kit (#74134; Qiagen, Hilden, Germany) and quantified using NanoDrop One (Thermo Scientific). After cDNA conversion (#1708891; iScript cDNA synthesis kit; Bio-Rad Laboratories, Hercules, CA, USA), the digital droplet PCR reactions were partitioned into 12,000–21,000 droplets/reaction using the QX2000 droplet generator (BioRad), amplified, and then, fluorescence was measured with the QX2000 droplet reader (BioRad). Finally, data were analyzed with QuantaSoft software (Bio-Rad), and the number of copies/μL was normalized with the reference gene and transformed to fold change.

Hydrolysis probes and primers were synthetized by Bio-Rad: Vcam (dMmuCPE5098832), Icam (dMmuCPE5100680), Nos2 (dMmuCPE5121694), IL-6 (dMmuCPE5095532), Tjp3 (dMmuCPE5120522), Ocel1 (dMmuCPE5090990), Cldn3 (dMmuCPE5088998), and Ppih (dRnoCPE5177931) was used as a reference gene. All procedures described herein followed the manufacturer’s instructions and dMIQE guidelines [21].

Frozen whole tissue specimens were homogenized in PE buffer (10 mm potassium phosphate buffer, pH 6.8 and 1 mm EDTA) containing 10 mm 3-[(3-cholamidopropyl) dimethylammonio]-1-propane sulphonate (CHAPS; Boehringer Mannheim, Mannheim, Germany) and protease inhibitor cocktail tablet Complete (Roche Diagnostics AB, Stockholm, Sweden). Following centrifugation, the supernatant was extracted, and protein concentration was determined by the Bradford method. Samples were diluted and denatured using Laemmli sample buffer and β-mercaptoethanol (Bio-Rad) before being loaded onto Bio-Rad Criterion^™ TGX Stain-free 4–15% gel (Tris-Glycine eXtended Stain-Free^™ precast gel) together with i-Bright^™ protein prestained ladder (LC5615; Invitrogen, Darmstadt, Germany). The stain-free gel was then visualized in Chemi-doc^™ MP Imager before being transferred onto a PVDF membrane using Bio-Rad’s Trans-Blot Turbo with high molecular weight settings. Once transferred, the stain-free image was obtained, followed by repeated wash (PBS, Tween 20) and block cycles (PBS, I-block, and Tween 20). Membranes were incubated with primary antibodies against claudin-1 (71-780, Invitrogen), claudin-3 (34-1700, Invitrogen), occludin (71-1500, Invitrogen), and tricellulin (48-8400, Invitrogen) overnight at 4°C. After repeated washings, the membranes were incubated for 1 h room temperature with an HRP-linked secondary antibody and visualized using chemiluminescent Clarity Western ECL (1705062; Bio-Rad). Membranes were stripped using Re-blot Plus Mild Antibody stripping solution (2502; Merck, Darmstadt, Germany) between antibodies. Data were then obtained using Image Laboratory software (Bio-Rad) using a stain-free normalization method.

Graph Pad Prism Software 9 (Software Inc., La Jolla, CA, USA) was employed for statistical analysis. ANOVA, Kruskal-Wallis, and Mann-Whitney U test were applied, and p values <0.05 were considered as statistically significant.

At baseline, there was no significant difference between the groups in the mean arterial pressure (MAP), heart rate (HR), and intestinal microvascular perfusion. In control animals, MAP remained constant during the observation period of 2 h. In the BD group, however, MAP rose steeply (the Cushing reflex) within the first 3 min after the induction of BD. This was followed by a significant fall, which persisted throughout the 2 h of the experiment, compared with both the baseline (p < 0.01) and the measurements in the control group (p < 0.01) (Fig. 1). In BD mice, HR increased significantly during the Cushing reflex but returned to normal values within 5 min after the induction of BD. Thereafter, BD mice had a tendency towards bradycardia and HR was found significantly lower in the BD group compared to the control group at several time points. Laser-Doppler measurements of intestinal microvascular perfusion showed a decreasing trend during the first 60 min, more pronounced in the BD group. The blood flow appeared to improve spontaneously after 60 min. There was no significant difference between intestinal microvascular perfusion between groups at any time point.

Microscopical analysis of the small intestines found no significant difference between the brain-dead mice and the control group. Sections of small intestine evaluated using the Chiu-Park grading system exhibited normal or near normal histology or incipient ischemic injury (discrete epithelial detachment at the tip of the villus), similar between the two groups (Fig. 2). Mucosal thickness, villus length, villus area, and villus volume did not differ between groups.

Only the Tjp3 expression showed to be significantly different among the study groups (p = 0.0496), being downregulated in the BD group. The rest of the studied genes showed comparable expression levels between the groups, but with a trend toward an upregulation for IL-6, Ocel1, and Clnd3 (Fig. 3).

The semiquantitative assessment of several intestinal junctional proteins using Western blot did not discern any significant differences in tissue expression of four different tight junction proteins in the intestines of BD mice compared with the control mice (Fig. 4). As a marker for tight junction integrity, claudin-3 expression and localization in the intestinal mucosa was assessed by immunofluorescence histology. Immunofluorescence showed a normal staining pattern and expression levels for the tight junction protein claudin-3 with a fine reticular pattern along the basolateral membrane from the crypts to the tip of the villi, without differences between the control group and the BD mice.

The current results show that, in spite of prolonged mild hypotension, intestinal mucosal perfusion remains well maintained and several key cellular and molecular features are preserved in brain-dead mice. Maintaining the intestinal mucosa integrity until procurement is paramount for prospective intestinal grafts, as the cold storage will likely cause various degrees of epithelial detachment, which is the hallmark of intestinal preservation injury [22].

Intestinal microcirculation is structurally complex and regulated by extrinsic neural input, circulating catecholamines as well as local metabolic factors (mostly vasodilator action) [23]. However, this particular setting, i.e., BD, is unique in the sense that much of the afferent neural signaling is abolished due to the injury to the brain stem. The intestine also has an intrinsic autoregulation of the blood flow through the enteric nerves, which induce vasodilation within the two parallel-coupled capillary networks for the mucosa and the muscular layer. The current findings indicate that intestinal perfusion appeared well preserved despite altered systemic hemodynamics and mild hypotension. Whereas the detailed compensatory mechanisms behind this finding remain unclear, our study did not identify any increased local sympathetic activity or effects, as a consequence the abrogated parasympathetic (vagal) innervation subsequent to the BD.

The intestine is notoriously vulnerable to episodes of hemodynamic instability and hypotension as well as to cardiac arrest. Neurogenic and hormonally driven splanchnic vasoconstriction are frequently observed following trauma or during critical illness. Whereas this mechanism is intended to preserve coronary and cerebral blood flow, it also raises concerns of ischemic damage to abdominal organs potentially usable for transplantation, including the intestine. However, this concern has been tempered by a German analysis on 39 intestinal donors indicated that although 31% of the donors were hemodynamically unstable upon hospital admission, they recovered and the following intestinal transplantations were eventually successful [24]. Likewise, in a series of 12 intestinal donors with cardiac arrest and where cardiopulmonary resuscitation was performed for a mean of 19 ± 12 min, results of intestinal transplantation were similar with the transplants from donors not requiring cardiopulmonary resuscitation [25]. Thus, the current results showed herein, together with the limited clinical evidence suggests that a certain degree of hemodynamic instability in potential intestinal brain-dead donors, either due to trauma, cardiocirculatory dysfunction or cardiac arrest can be tolerated and compensated. Whether this depends to an intestinal resilience higher than previously anticipated or due to other mechanisms secondary to BD, i.e., vasodilation, remains unclear.

Intestinal ischemia results in altered villus morphology, particularly villus shortening [26, 27]. Villus shortening is considered a mechanism of repair of intestinal epithelium after injury, and it is based on the active contraction of lamina propria muscles fibers directly after ischemia and a zipper-like epithelial contraction, ultimately resulting in the reduction and coverage of any denuded surface areas [28]. The morphometric analysis after 2 h of BD could not discern any morphometric differences between the intestinal mucosa of brain-dead mice and controls, excluding any significant ischemic episode occurring transiently during the experiment.

One integral part of intestinal integrity is the tight junctional proteins. Tight junctions are multiprotein complexes located at the apex of the basolateral plasma membrane and are major determinants in the regulation of paracellular permeability. Rapid degradation and redistribution of various intestinal tight junction proteins have been reported within a few hours following ischemia [29], hemorrhagic shock [30], osmotic [31], or heat stress [32]. Gene expression analysis identified several early and discrete changes in the expression of several tight junction-related genes, potentially indicating early remodeling in response to either hypotension or the alleged inflammatory activation following BD. However, protein analysis at the end of the experiment did not discern any differences in the expression and distribution of several key components of the junctional complex, usually affected during low flow states.

Vasopressors are frequently used to maintain the circulatory stability in the potential organ donor, although this may impair intestinal blood flow. Hypotensive organ donors are usually resuscitated using a combination of fluid and vasopressor medication. In the setting of intestinal donation, there is some concern that vasopressors can induce visceral vasoconstriction and intestinal hypoperfusion and high doses of vasopressors often preclude intestinal donation [24, 33]. Although the intestinal changes induced by these resuscitation strategies are quite well known in shock and trauma patients, the effect of these strategies on the intestinal perfusion and injury in brain-dead individuals have not been studied. Considering the rather unexpected findings presented herein, we speculate that brain-dead subjects may react differently to vasopressors in terms of intestinal perfusion and further research on this topic is mandated.

The present study has certain limitations, as tissue sampling was performed at only 1 time-point and the period of BD was relatively short. Our previous unpublished observations indicated that beyond 2 h, unresuscitated brain-dead mice gradually became profoundly hypotensive and died. However, despite of the short time frame, the charges associated with BD such as the typical hemodynamic alterations such as the Cushing reflex and bradycardia as well as the inflammatory activation were reproduced by our mouse model. In addition, some intestinal changes induced by hypotension should have been apparent during the 120 min of observation, as other murine models of hypotension reported significant tissue injury within 90 min from the onset of hypotension [30, 33, 34]. Among the strengths of the study are the iterative and simultaneous mucosal perfusion and blood pressure measurements and the presence of a proper control group.

The current results show that brain-dead mice have well maintained intestinal perfusion and that there is no evidence of ongoing intestinal ischemia despite the mild hypotension. These findings suggest the existence of intrinsic vasoregulatory mechanisms in brain-dead individuals, which may compensate for the mild hypotensive episodes even without aggressive vasoactive resuscitation.

The authors gratefully acknowledge the skillful assistance of Margareta Filipson in performing the animal experiments.

The study followed the regulations outlined by the European Union and was reviewed and approved by the Gothenburg Committee of the Swedish Animal Welfare Agency (Dnr. 238-2004).

The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

Funding for this study was provided by grants from the Swedish state under the agreement between the Swedish government and the country councils (Grants ALFGBG-518371 and ALFGBG-812881). Lucas Ferreira was supported by a scholarship from the Brazilian federal government (Coordenação de Aperfeiçoamento de Pessoal de Nível Superior).

Mihai Oltean, Jasmine Bagge, Anna Casselbrant, Lucia de Miguel Gomez, and Lucas Ferreira da Anunciação performed experiments and data analysis, interpretation of data, and drafted the manuscript. Andreas Lundgren, Tomas Lorant, and Michael Olausson designed the study, participated in its coordination, and reviewed the manuscript. Mats Hellström participated in the interpretation of data and review of the manuscript. All authors read and approved the final manuscript.

The data presented in this study are available on request from the corresponding author. The data are not publicly available due to further analysis ongoing.