Journal Logo

Original Basic Science—General

17β-Estradiol as a New Therapy to Preserve Microcirculatory Perfusion in Small Bowel Donors

Vieira, Roberta Figueiredo MD, PhD1; Breithaupt-Faloppa, Ana Cristina PhD1; Correia, Cristiano Jesus PhD1; Armstrong, Roberto Jr BSc1; Coutinho-e-Silva, Raphael dos Santos BSc1; Ferreira, Sueli Gomes PhD1; Moreira, Luiz Felipe Pinho MD, PhD1; Sannomiya, Paulina PhD1

Author Information
doi: 10.1097/TP.0000000000003280
  • Free
  • Visual Abstract



In the past years, intestinal transplantation has evolved from being an experimental procedure toward becoming a viable treatment option for patients with intestinal failure and complications of total parenteral nutrition.1 However, intestinal transplantation continues to be an arduous procedure associated with worse results than solid organ transplantation.2

Extensive data support that brain death (BD) elicits hemodynamic and hormonal disarrangements that might impair organ function.3 Indeed, despite all clinical efforts, organ dysfunction still compromises retrieval in up 25% cases,4 that may be even worse with intestinal grafts. Because intestinal mucosa is profoundly sensitive to ischemia, grafts derive solely from brain-dead donors, and those from deceased donors are rejected.5,6

Previous studies have shown that 17β-estradiol (E2) confers beneficial effects on the endothelium, including production of nitric oxide (NO) and prostacyclin, mediating vasodilation, and promoting antiinflammatory and antioxidant effects.7 The hormone acts as a protective vasoactive agent against deleterious microcirculatory conditions, such as BD and ischemia/reperfusion injury.8,9 Indeed, E2 was effective in improving survival of human pancreatic islet in a mice transplant model.10 In addition, brain-dead rats treated with E2 exhibit limited lung injury, mainly due to its actions on endothelial and inducible NO synthases. To determine whether intestinal tissue might be affected similarly, this study explored the intestinal microcirculatory actions of E2 in a fast-onset BD model.


Animals were treated with humane care in compliance with ethical standards for animal research embraced by the Brazilian College of Animal Experimentation and with the 2011 “Guide for the Care and Use of Laboratory Animals,” endorsed by the U.S. National Institutes of Health. The University of São Paulo Medical School Ethics Committee for Research Projects Analysis accepted this protocol under number 3989/13/112.

Groups and Treatments

Male Wistar rats (aged 8 wk; 250–300 g) were maintained under standard conditions with free access to water and food. Rats were randomly distributed to 3 groups: sham-operated rats that were only trepanned (SH, n = 10); brain-dead rats subjected to intracranial catheter inflation (BD, n = 10); and BD rats treated with E2 1 hour after BD induction (BD-E2, n = 10). The BD-E2 group received a single intravenous injection of 280 μg/kg E2 (β-Estradiol-Water Soluble, Sigma-Aldrich, St Louis, MO). The SH and BD groups received an equal dose of estradiol vehicle (2-hydroxypropyl-β-cyclodextrin, Sigma-Aldrich) diluted in saline solution. Experiments were performed in 2 steps: (a) for intravital microscopy analyses (small vessels perfusion, blood flow, and leukocyte-endothelial interactions), 5 rats per group were used; (b) for serum and tissue analyses (cytokines/hormones measurements, immunohistochemistry, real-time polymerase chain reaction (PCR), and histopathology), another 5 animals per group were used, to avoid any bias related to manipulation.

Brain Death Model

General anesthesia was induced with 5% isoflurane, followed by intubation and mechanical ventilation (10 mL/kg tidal volume, 70 cycles/min respiratory rate, and 100% oxygen fraction [Harvard Apparatus, Inc., Holliston, MA]). Arterial pressure monitoring was obtained with carotid artery cannulation, and the internal jugular vein was used for blood sampling and solution infusion. BD induction was accomplished by rapidly inflating Fogarty-4F catheter (Baxter Health Care, Deerfield, IL) through a parietal hole with 0.5 mL saline solution, as previously described.11 Anesthesia was discontinued, and BD was confirmed by mydriasis, apnea, and absence of corneal reflexes. Sham animals were submitted to all procedures described above except for the intracranial catheter insertion and the discontinuation of general anesthesia.

Intravital Microscopy of Mesenteric Microcirculation

Through midline abdominal incision, the small bowel was exposed and mesentery microcirculation analyzed by intravital microscopy, as previously described.12 The mesentery was constantly flushed with warmed Krebs–Henseleit solution saturated with 95% N2/5% CO2. The image analyzer system, including an AxioCam HSc camera (Carl Zeiss Co., München-Hallbergmoos, Germany), connected to an Axioplan 2 microscope (Carl Zeiss), and Axiovision 4.1 software (Carl Zeiss), was employed to evaluate leukocyte-endothelial interactions and small vessels perfusion. Perfused small vessel number was quantified by applying a 10× light microscopic objective in 1 mm2 mesenteric area, and 5–7 areas were selected for each animal. The perfusion of the third-order vessels (<30 µm diameter) was classified based on continuous or intermittent/absent flow. Leukocyte–endothelial interactions were evaluated applying a ×40 light microscope objective in post capillary venules (15–30 µm diameter), to evaluate the number of rolling leukocytes per 3 minutes, adhered leukocytes/100 µm venule length, and migrated cells/5000 µm2. These were measured in 5 fields for each animal. The mesenteric microcirculation blood flow was determined in situ and in vivo using a Fine Needle Probe coupled to a laser flowmeter (IN191 Laser Doppler Flowmetry, AD Instruments, Colorado Springs, CO). Mesenteric blood flow was calculated by multiplying the average blood cell number and the speed in the tissue under the laser beam. The flow was assessed by applying a laser flowmeter probe that analyzed the microcirculatory area comprising vessels with varying diameters (first- to third-order vessels).


The animals were exsanguinated from the abdominal aorta, and the mesentery was procured. Hexane solution in nitrogen was used for fast freezing of the tissue. The mesentery was sectioned (8-µm), 2 sections per slide, and fixed in cold acetone (10 min). The direct immunohistochemical assays started by washing the samples in Tris-buffered saline Tween-20 (TBST) and then permeabilized with TBST and Triton X-100 (excepted for adhesion molecules). TBST containing 1% BSA was used for nonspecific sites blocking, followed by endogenous peroxidase blockade (2% H2O2). Primary antibodies, antirat intercellular adhesion molecule (ICAM)-1 (1:50, CD54, Santa Cruz Biotechnology, Dallas, TX), antirat vascular cell adhesion molecule (VCAM)-1 (1:100, CD106, BD Pharmingen, San Jose), antihuman P-selectin (1:40, CD62P, R&D Systems, Minneapolis, MN), antiendothelial nitric oxide synthase (eNOS) (1:100, Abcam, Cambridge, MA), and antiendothelin-1 (1:100, Santa Cruz Biotechnology) were applied diluted in TBST-BSA. The sections were incubated for 2 hours at 37°C, rinsed with TBST, and incubated with 1:200 antimouse or antirabbit secondary antibodies, IgGs linked to horseradish peroxidase (Millipore, Billerica, MA) for 1 hour at 37°C. The substrate solution of 3-amino-9-ethylcarbazole (Vector Laboratories, Burlingame, CA) was applied for 5 to 10 minutes and counterstained with Mayer’s hematoxylin. For negative control sections without primary antibodies, incubation were used. The evaluation was done employing a system consisting of a digital camera (DS-Ri1, Nikon, Tokyo, Japan) connected to a microscope (Nikon) and evaluated applying the NIS-Elements BR-software (Nikon).

Real-time Polymerase Chain Reaction

Mesentery RNA extraction (mirVana kit; Ambion, Carlsbad, CA) and cDNA transcription (High Capacity Reverse transcriptase kit; Applied Biosystems) were evaluated by real-time PCR (Step One Plus; Applied Biosystems, CA), applying TaqMan primers (Applied Biosystems): GAPDH (Rn01775763_g1), β-actin (Rn00667869_m1*), eNOS (Rn02132634_s1*), and endothelin-1 (Rn00561129_m1*).


A segment of jejunum (3 cm) was procured, immersed in 10% paraformaldehyde for 24 h/day, and then embedded in paraffin. Two slices (4 µm) per animal were stained with hematoxylin and eosin and analyzed by light microscopy for intestinal edema, hemorrhage, and inflammatory cells. The tissue area was captured, digitalized (Panoramic Viewer; 3DHISTECH, Budapest, Hungary), and exported to NIS Elements Software Basic Research (NIS-elements, Nikon, Tokyo, Japan). Histological evaluation of edema and hemorrhage was performed by 2 independent observers, applying an intensity score: 0 (absent), 1 (slight), 2 (moderate), and 3 (intense). Inflammatory cells were counted and divided by villus height (µm). The results were verified by a correlation test.

Measurement of Cytokines and Hormones

Measurements of serum levels of interleukin (IL)-10, cytokine-induced neutrophil chemoattractant (CINC)-1, tumor necrosis factor (TNF)-α, and vascular endothelial growth factor (VEGF) were determined applying a Milliplex kit (Millipore). ELISAs were used to quantify serum levels of corticosterone (Cayman, Ann Arbor, MI) and E2 (USCN Life Sciences, Cloude-Clone Co., TX).

Statistical Analysis

Graph Pad Prism Software v.6 (Software Inc., La Jolla, CA) was employed for statistical analysis. Kruskal-Wallis test followed by Dunn’s test for multiple comparisons were applied for nonparametric data, and ANOVA followed by Tukey’s test for multiple comparisons was used for parametric data. P values < 0.05 were considered as statistically significant. Results are presented as mean values with the scanning electron microscope.


Hemodynamic Behavior

There were no fatal events in any of the groups. Both BD groups displayed an abrupt rise in blood pressure, followed by a 1-hour hypotensive period. At 3 hours, mean arterial pressure values were similar to the SH group (data not shown).


Brain-dead groups exhibited significantly reduced levels of serum corticosterone in comparison with SH group, while E2 levels were significantly higher in BD-E2 animals compared with those in the other groups (Figure 1).

Measurement of serum corticosterone and E2 levels in SH (n = 7), rats subjected to BD (n = 7), and BD-E2 (n = 7). Reference value for corticosterone was 1356 ± 196 pg/mL. Naïve male rats presented E2 levels of 10.81 ± 2.78 pg/mL (n = 5). Values represent the mean ± SEM. General P value obtained with Kruskal-Wallis, followed by Dunn’s test. λ P = 0.0007 SH vs BD and *P = 0.0384 BD-E2 vs BD. BD, brain death; BD-E2, brain-dead rats treated with E2; E2, 17β-estradiol; SEM, standard error of the mean; SH, sham-operated rats.

Histopathological Data

BD-E2 rats exhibited decreased mucosal edema and submucosal hemorrhage in relation to BD group. The number of inflammatory cells did not vary between groups as demonstrated in Table 1.

Histomorphometry of intestinal edema and hemorrhage

Leukocyte–endothelial interaction data are described in Table 2. The number of rolling, adhered, and migrated leukocytes did not vary among groups at the end of the procedure. The diameter of the analyzed small vessels was similar between groups.

Leukocyte-endothelial interactions at mesenteric microcirculation

Mesenteric Small Vessel Perfusion and Blood Flow

BD group displayed a significant reduction in the proportion of perfused small vessels compared with SH group (Figure 2A), which improved after treatment with E2. Mesenteric blood flow did not differ among groups (Figure 2B).

Determination of microcirculation blood flow and mesenteric perfused small vessels in SH (n = 5), rats subjected to BD (n = 5), and BD-E2 (n = 5). Values represent the mean ± SEM. General P value obtained with ANOVA, followed by Tukey’s test. *P = 0.0093 SH vs BD; φ P = 0.0404 BD-E2 vs BD. ANOVA, analysis of variance; BD, brain death; BD-E2, brain-dead rats treated with E2; E2, 17β-estradiol; SEM, standard error of the mean; SH, sham-operated rats.

eNOS and Endothelin-1 Protein and Gene Expression

Animals treated with E2 showed a significant increase in eNOS gene and protein expression compared with BD rats (Figure 3). Endothelin-1 expression did not vary between groups (Figure 3). eNOS immunostaining photomicrographs are presented in Figure 4.

Protein and gene expression of eNOS and endothelin-1 in the mesenteric microcirculation in SH (n = 5), rats subjected to BD (n = 5), and BD-E2 (n = 5). Values represent the mean ± SEM. Gene expression levels are relative to those in naïve rats. General P value obtained with Kruskal-Wallis, followed by Dunn’s test *P < 0.0001 BD-E2 vs BD; λ P = 0.0083 BD-E2 vs SH. BD, brain death; BD-E2, brain-dead rats treated with E2; E2, 17β-estradiol; eNOS, endothelial nitric oxide synthase; SEM, standard error of the mean; SH, sham-operated rats.
Immunostaining for eNOS in mesenteric microcirculation in SH, rats subjected to BD, and BD-E2. Note the higher eNOS protein expression in the BD-E2. Original magnification, ×400 for all images. BD, brain death; BD-E2, brain-dead rats treated with E2; E2, 17β-estradiol; eNOS, endothelial nitric oxide synthase; SH, sham-operated rats.

Serum Cytokines

As summarized in Figure 5, there was an increase in TNF-α levels and IL-10 and a decrease in VEGF levels in BD groups in comparison with SH group. Treatment with E2 was effective in decreasing IL-10 and CINC-1 serum levels.

Measurement of serum cytokines: IL-10, TNF-α, VEGF, and CINC-1 in SH (n = 5), rats subjected to BD (n = 5), and BD-E2 (n = 5). Values represent the mean ± SEM. General P value obtained with Kruskal-Wallis, followed by Dunn’s test. γ P = 0.0029; φ P = 0.0285 SH vs BD; *P = 0.0236 BD-E2 vs BD. BD, brain death; BD-E2, brain-dead rats treated with E2; CINC, cytokine-induced neutrophil chemoattractant; IL-10, interleukina-10; SEM, standard error of the mean; SH, sham-operated rats; TNF-α, tumor necrosis factor α; VEGF, vascular endothelial growth factor.

Adhesion Molecules

Both BD groups exhibited increased ICAM-1 expression compared with SH group. 17β-Estradiol treatment reduced VCAM-1 and P-selectin expression (Figure 6).

Expression of ICAM-1, VCAM-1, and P-selectin in mesenteric microcirculation in SH (n = 5), rats subjected to BD (n = 5), and BD-E2 (n = 5). Values represent the mean ± SEM. General P value obtained with Kruskal-Wallis, followed by Dunn’s test. *P = 0.0004 SH vs BD; φ P = 0.0034 BD-E2 vs BD. BD, brain death; BD-E2, brain-dead rats treated with E2; ICAM, intercellular adhesion molecule; SEM, standard error of the mean; SH, sham-operated rats; VCAM, vascular cell adhesion molecule.


The data presented suggest that E2 treatment improved mesenteric perfusion by increasing protein and gene eNOS expression. In addition, the hormone favorably influenced microcirculatory actions, reducing intestinal edema and hemorrhage, chemokine levels, and adhesion molecule expression.

Those results are relevant, as most intestinal grafts are procured from brain-dead donors, who have endured catastrophic cerebral injury followed by hemodynamic and metabolic instability that further harms intestinal perfusion.6,13 As the intestinal mucosa demands high blood flow to sustain oxygen consumption, even brief periods of hypoperfusion can cause significant tissue damage that might result in bacterial translocation or intestinal perforation.14,15 Moreover, it is crucial to preserve crypts perfusion in order to maintain the intestinal regenerative capacity.16 Therefore, the recommendation is to only accept donors with stable hemodynamics, which limits the graft pool.5 This might account for the low percentage of intestinal grafts considered suitable for transplantation.13

BD is associated with microcirculatory dysfunction, despite maintenance of systemic hemodynamics. As demonstrated in this study, BD rats exhibited a reduction in mesenteric-perfused small vessels (<30 µm), despite normal mesenteric blood flow (comprising first, second, and third-order vessels). Similarly, a reduction in conjunctival and sublingual microcirculation has been observed in brain-dead patients.17 In addition, impairment of mesenteric, pancreatic, and hepatic microcirculation has been previously described in brain-dead rats.11,18

In this study, E2 treatment was effective in improving perfusion of mesenteric small vessels, primarily by increasing eNOS protein and gene expression. Estradiol vascular effects are attributed to regulation of eNOS expression and activity, as demonstrated in experimental models of ischemia/reperfusion injury.9,19 This suggests that E2 enhances NO availability by eNOS upregulation, which leads to vasodilation, improving microcirculatory perfusion. The beneficial endothelial actions of E2 might account for reductions in intestinal edema and hemorrhage associated with BD. This is consistent with prior studies, which have found similar results in the lung tissue of brain-dead rats treated with E2.8

Aiming to evaluate endothelial cell activation and inflammation related to BD-induced microvascular dysfunction, expression of endothelial adhesion molecules, cytokines, and hormones were quantified. As previously demonstrated, a reduction in serum corticosterone levels is associated with increased ICAM-1 expression, a marker of endothelial cell activation in BD rats.11 In this study, increased expression of ICAM-1 was accompanied by increased serum levels of TNF-α, a proinflammatory cytokine that upregulates ICAM-1, and by increased levels of IL-10, a cytokine that counteracts proinflammatory mediators.20,21 In addition, reduced levels of VEGF in both BD groups suggest that this growth factor upregulates cytoprotective molecules.22 Indeed, brain-dead patients exhibit upregulation of proinflammatory cytokines, like IL-6, IL-8, and TNF-α, as long with antiinflammatory cytokines, like IL-10.23 In this study, treatment with E2 displayed an early protective effect in reducing VCAM-1 expression, a possible biomarker of primary graft dysfunction, and in decreasing CINC-1 (homolog to human IL-8) levels. This treatment also resulted in nondifferential reductions in both TNF-α and IL-10. Indeed, previous publications establish the antiinflammatory properties of estrogen in reducing CINC-1 levels.24,25

This study has limitations that must be acknowledged. First, mesenteric microcirculation was assessed 3 hours after induction of BD, and, at that point, the inflammatory process was not fully developed, as indicated by equivalent leukocyte-endothelial interactions and inflammatory cell counts in intestinal villous among groups. Second, the sham group behaved as a mild injured control due to surgical manipulation. In addition, for a complete understanding of the impact of estradiol treatment on intestinal tissue, a transplant might have been performed and the graft analyzed.

In conclusion, treatment with E2 was effective in improving mesenteric perfusion and reducing intestinal edema and hemorrhage associated with BD.

The suggestion is that E2 might be considered a therapy to mitigate, at least in part, the deleterious effects of BD in small bowel donors.


The authors wish to thank Assistant Professor Dr Luiz Fernando Ferraz da Silva for his valuable contribution to the histological analyses.


1. Koudstaal LG, ‘t Hart NA, Ottens PJ, et al. Brain death induces inflammation in the donor intestine. Transplantation. 2008; 86:148–154
2. Roskott AM, Nieuwenhuijs VB, Dijkstra G, et al. Small bowel preservation for intestinal transplantation: a review. Transpl Int. 2011; 24:107–131
3. Chudoba P, Krajewski W, Wojciechowska J, et al. Brain death-associated pathological events and therapeutic options. Adv Clin Exp Med. 2017; 26:1457–1464
4. McKeown DW, Bonser RS, Kellum JA. Management of the heartbeating brain-dead organ donor. Br J Anaesth. 2012; 108(Suppl 1):i96–107
5. Kotloff RM, Blosser S, Fulda GJ, et al.; Society of Critical Care Medicine/American College of Chest Physicians/Association of Organ Procurement Organizations Donor Management Task Force. Management of the potential organ donor in the ICU: Society of Critical Care Medicine/American College of Chest Physicians/Association of Organ Procurement Organizations Consensus Statement. Crit Care Med. 2015; 43:1291–1325
6. Roskott AM, van Haaften WT, Leuvenink HG, et al. Histopathologic and molecular evaluation of the organ procurement and transplantation network selection criteria for intestinal graft donation. J Surg Res. 2014; 189:143–151
7. Chakrabarti S, Morton JS, Davidge ST. Mechanisms of estrogen effects on the endothelium: an overview. Can J Cardiol. 2014; 30:705–712
8. Vieira RF, Breithaupt-Faloppa AC, Matsubara BC, et al. 17β-Estradiol protects against lung injuries after brain death in male rats. J Heart Lung Transplant. 2018; 37:1381–1387
9. Rocha de Sousa PT, Breithaupt-Faloppa AC, Correia C, et al. 17β-estradiol prevents mesenteric injury induced by occlusion of the proximal descending aorta in male rats J Vasc Surg. 2018; 67:597–606
10. Contreras JL, Smyth CA, Bilbao G, et al. 17beta-estradiol protects isolated human pancreatic islets against proinflammatory cytokine-induced cell death: molecular mechanisms and islet functionality Transplantation. 2002; 74:1252–1259
11. Simas R, Sannomiya P, Cruz JW, et al. Paradoxical effects of brain death and associated trauma on rat mesenteric microcirculation: an intravital microscopic study. Clinics (Sao Paulo). 2012; 67:69–75
12. Nakagawa NK, Nogueira RA, Correia CJ, et al. Leukocyte-endothelium interactions after hemorrhagic shock/reperfusion and cecal ligation/puncture: an intravital microscopic study in rat mesentery. Shock. 2006; 26:180–186
13. Fischer-Fröhlich CL, Königsrainer A, Schaffer R, et al. Organ donation: when should we consider intestinal donation Transpl Int. 2012; 25:1229–1240
14. Weaver JL, Matheson PJ, Matheson A, et al. Direct peritoneal resuscitation reduces intestinal permeability after brain death J Trauma Acute Care Surg. 2017; 84:265–272
15. Schildberg FA, Liu B, Afify M, et al. Cyclooxygenase inhibitors as a new therapeutic strategy in small bowel transplantation Transplantation. 2016; 100:2324–2331
16. Lautenschläger I, Pless-Petig G, Middel P, et al. Cold storage injury to rat small-bowel transplants-beneficial effect of a modified HTK solution Transplantation. 2018; 102:1666–1673
17. Tamosuitis T, Pranskunas A, Balciuniene N, et al. Conjunctival microcirculatory blood flow is altered but not abolished in brain dead patients: a prospective observational study. BMC Neurol. 2016; 16:95
18. Obermaier R, von Dobschuetz E, Keck T, et al. Brain death impairs pancreatic microcirculation. Am J Transplant. 2004; 4:210–215
19. Breithaupt-Faloppa AC, Breithaupt-Faloppa AC, Fantozzi ET, et al. Protective effect of estradiol on acute lung inflammation induced by an intestinal ischemic insult is dependent on nitric oxide. Shock. 2013; 40:203–209
20. Grütz G. New insights into the molecular mechanism of interleukin-10-mediated immunosuppression J Leukoc Biol. 2015; 77:3–15
21. Xue M, Qiqige C, Zhang Q, et al. Effects of tumor necrosis factor α (TNF-α) and interleukina 10 (IL-10) on intercellular cell adhesion molecule-1 (ICAM-1) and cluster of differentiation 31 (CD31) in human coronary artery endothelial cells Med Sci Monit. 2018; 24:4433–4439
22. Limani P, Linecker M, Oberkofler CE, et al. Remote ischemic preconditioning: a novel strategy in rescuing older livers from ischemia-reperfusion injury in a rodent model. Ann Surg. 2016; 264:797–803
23. Schwarz P, Custódio G, Rheinheimer J, et al. Brain death-induced inflammatory activity is similar to sepsis-induced cytokine release. Cell Transplant. 2018; 27:1417–1424
24. Hsu JT, Yeh HC, Chen TH, et al. Role of Akt/HO-1 pathway in estrogen-mediated attenuation of trauma-hemorrhage-induced lung injury. J Surg Res. 2013; 182:319–325
25. Shimizu T, Suzuki T, Yu HP, et al. The role of estrogen receptor subtypes on hepatic neutrophil accumulation following trauma-hemorrhage: direct modulation of CINC-1 production by Kupffer cells. Cytokine. 2008; 43:88–92
Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.