Pomegranate seed oil (PSO), which is the major source of conjugated linolenic acids such as punicic acid (PuA), exhibits strong anti-inflammatory properties. Necrotizing enterocolitis (NEC) is a devastating disease associated with severe and excessive intestinal inflammation. The aim of this study was to evaluate the effects of orally administered PSO on the development of NEC, intestinal epithelial proliferation, and cytokine regulation in a rat model of NEC. Premature rats were divided into three groups: dam fed (DF), formula-fed rats (FF), or rats fed with formula supplemented with 1.5% of PSO (FF + PSO). All groups were exposed to asphyxia/cold stress to induce NEC. Intestinal injury, epithelial cell proliferation, cytokine production, and trefoil factor 3 (Tff3) production were evaluated in the terminal ileum. Oral administration of PSO (FF+PSO) decreased the incidence of NEC from 61 to 26%. Feeding formula with PSO improved enterocyte proliferation in the site of injury. Increased levels of proinflammatory IL-6, IL-8, IL-12, IL-23, and TNF-α in the ileum of FF rats were normalized in PSO-treated animals. Tff3 production in the FF rats was reduced compared with DF but not further affected by the PSO. In conclusion, administration of PSO protects against NEC in the neonatal rat model. This protective effect is associated with an improvement of intestinal epithelial homeostasis and a strong anti-inflammatory effect of PSO on the developing intestinal mucosa.
- enteral nutrition
- intestinal injury
- mucosal inflammation
the pomegranate (punica granatum) is one of the oldest edible fruits with a renowned medical history and a symbol of life, longevity, and health (40). Native to the region of Persia and the Himalayas, P. granatum has been widely cultivated and consumed for centuries by many different cultures starting from the Mediterranean region extending to China, India, and southeast Asia (36). Introduced into Latin America in 18th century, P. granatum is now commercially cultivated in dry Southwest of the United States. The fruit gives rise to three parts: the seeds, the juice, and the peels. Despite the growing popularity of juice products, the importance of the oily phase of the seed has been largely overlooked until recent years.
Pomegranate seed oil (PSO) is predominantly composed of triglycerides containing unsaturated fatty acids including high levels of conjugated linolenic acids (CLnA; 78–86%), followed by oleic acid (4–7%, 18:1n-9), linoleic acid (LA; 3–6%, 18:2n-6), palmitic acid (2–3%, 16:0), and stearic acid (2%, 18:0) (52). PSO CLnA consist of up to seven geometric isomers of 9,11,13-octadecatrienoic acid (18:3), separated by a single carbon-carbon linkage (Fig. 1) as opposed to being separated by methylene groups in α-linolenic acid (18:3n-3) (52). The major CLnA in PSO is punicic acid (PuA; 57–71%) (52). Structurally, PuA is a C18 carbon fatty acid containing cis-9, trans-11, cis-13 double bonds (cis-9,trans-11,cis-13-octadecatrienoic acid) (42).
It has been suggested that the higher intake of conjugated fatty acids is associated with the lower incidence of inflammatory diseases in the Middle East and Asian populations compared with Western countries (8). Experimental studies showed that PSO is able to reduce tumor occurrence in both ex vivo and in vivo rodent models (29, 35, 43). In addition, recent studies indicate that PuA ameliorates TNSB- and DSS-induced colitis (2, 8).
Necrotizing enterocolitis (NEC) is a major cause of morbidity and mortality in premature infants. Despite an increasing occurrence of NEC in the United States, the etiology is unknown. The combination of a genetic predisposition, intestinal immaturity, and high immunoreactivity of the intestinal mucosa leads to the development of NEC (48). The type and amount of enteral feeds is critical in this process (22). Unfortunately, no predictive diagnostic tests or effective treatments are available at this time (13).
Beneficial effects of infant formula supplemented with polyunsaturated fatty acids (PUFA) on the development of the central nervous system and visual acuity of neonates are well established (23). However, only limited information is available regarding the effect of PUFA on neonatal intestines. In a clinical study, Carlson et al. (14) found a lower incidence of NEC in infants fed formula containing higher amounts of PUFA compared with controls. In the rat NEC model, supplementation of PUFA into formula reduces the degree of intestinal injury (1, 12). The mechanism of PUFA-induced protection against NEC is not clear, but the decreased expression of Toll-like receptor 4 and platelet-activating factor receptor was observed (38). However, nothing is known about the effect of the CLnA on NEC pathogenesis.
The gut mucosal immune system is capable of producing an array of cytokines important in the development and control of the inflammatory response (32). We (26, 27) and others (20, 41, 53) have shown the dysregulation of inflammatory cytokines in the intestinal mucosa during NEC pathogenesis. Particularly, interleukin (IL)-6, IL-8, IL-12, IL-18, and tumor necrosis factor-α (TNF-α) have an important role and a diagnostic value in sepsis and in NEC (9, 26, 45). IL-23 is critical in the etiology of inflammatory bowel disease (51), but its role in NEC pathogenesis has not been studied yet.
The intestinal epithelium protects the underlying lamina propria against damage and microbe invasion through the production of mucins (Muc) and trefoil factors (Tff). In the rat small intestine, Muc2 is the predominant secretory mucin produced by goblet cells. Mucins are cosecreted with Tff, small peptides exerting multiple biological effects on epithelium. Impaired production of Muc2 and Tff3 has been reported in clinical and experimental NEC (15, 33).
The aim of this study was to determine whether oral administration of PSO protects against experimental NEC and to elucidate the mechanisms underlying the protective actions of PSO. In a rat model of NEC, we investigated the efficacy of PSO on disease development and severity, the effect on intestinal morphology, enterocyte proliferation, mucin expression, and proinflammatory cytokine production.
MATERIALS AND METHODS
Animal model of NEC and diets.
This protocol was approved by the Animal Care and Use Committee of the University of Arizona (A-324801-95081). Sixty neonatal Sprague-Dawley rats (Charles River Laboratory, Pontage, MI) were collected by caesarian section 24 h before their scheduled birth, and the first feeding started 2 h after delivery. Animals were hand fed six times daily with a total volume of 850 μl of rat milk formula per day (19). Experimental NEC was induced by asphyxia (breathing 100% nitrogen gas for 60 s) and cold stress (4°C for 10 min) twice daily (18). Caesarian section-delivered pups were divided into the following experimental groups: neonatal rats formula fed (FF; n = 20), neonatal rats fed with formula containing 1.5% of PSO (FF + PSO; n = 20), and dam-fed littermates fed by surrogate mothers as a baseline control (DF; n = 20). After 96 h, all surviving animals were terminated via decapitation. Animals that developed signs of distress or imminent death before 96 h were terminated and included in the study.
The rat milk formula used for feeding of neonatal rats was prepared primarily from evaporated milk as described previously (19). In the FF group, the major source of lipids in the formula was the mixture of Intralipid (Fresenius Kabi, Uppsala, Sweden) and almond oil. In the FF + PSO diet, almond oil was replaced with 1.5% wt/wt of PSO. The FA composition of the FF and FF + PSO diets expressed as a percent of total FA is shown in Table 1. The FF + PSO diet contained 2.7% of CLnA of the total FA, whereas the CLnA were absent in the FF diet. Concentration of arachidonic acid (AA; 20:4n-6) and docosahexaenoic acid (DHA; 22:6n-3) was similar in both diets.
All measurements were performed in the ileum from the same set of rat pups that were used for NEC evaluation.
After termination, a 2-cm piece of distal ileum was removed and fixed in 70% ethanol, paraffin embedded, sectioned at 4–6 μm, and stained with hematoxylin and eosin (H&E) for histological evaluation of NEC. Pathological changes in intestinal architecture were evaluated using our previously published NEC scoring system (18, 33, 50). Histological changes in the ileum were scored by a blinded evaluator and graded as follows: 0 (normal), no damage; 1 (mild), slight submucosal and/or lamina propria separation; 2 (moderate), moderate separation of submucosa and/or lamina propria, and/or edema in submucosal and muscular layers; 3 (severe), severe separation of submucosa and/or lamina propria, and/or severe edema in submucosa and muscular layers, region villous sloughing; 4 (necrosis), loss of villi and necrosis. Intermediate scores of 0.5, 1.5, 2.5 and 3.5 were also utilized to more accurately assess levels of ileal damage when necessary (18, 33, 50). To determine incidence of NEC, animals with histological scores of <2 have not developed NEC; animals with histological scores of ≥2 have developed NEC (33, 50).
Morphometrical measurements and epithelial cell proliferation in the ileum.
A 2-cm section of distal ileum stained with H&E was used for morphometric measurements as previously described (16). Briefly, twenty villi were measured in each histological sample, and 8–10 animals were evaluated per experimental group. Sections from animal with a NEC score of ≥3 were not included in analyses because of the lack of intact tissue to evaluate. Villi were measured from the tip to the crypt base and using an image analysis system (Image-Pro Plus; Media Cybernetics, Silver Spring, MD). Statistical analysis of all morphological data was performed in a blind manner to prevent observer bias.
Expression of proliferating cell nuclear antigen (PCNA) was evaluated in serial sections of the paraffin-embedded samples from the ileum prepared as we previously described (16). After deparaffinization, rehydration and incubation in hydrogen peroxide, sections were blocked with 1.5% appropriate serum (Vector Laboratories, Burlingame, CA), then incubated with 2.0 μg/ml goat polyclonal anti-rat PCNA (Santa Cruz Biotechnology, Santa Cruz, CA), followed by biotinylated secondary antibody (Vector Laboratories), Vectastain Elite ABC reagent (Vector Laboratories), diaminobenzidine activated by hydrogen peroxide, and counterstained with hematoxylin. Control sections were treated with the same procedure except they were incubated with 2.0 mg/ml rabbit Ig (for Casp-3) or 2.0 μg/ml goat Ig (for PCNA) (Sigma, St. Louis, MO). No immunostaining was observed in the controls. The enumeration of positively stained cells was accomplished by counting 15 villi per animal and evaluating six rats per experimental group.
Fatty acid analysis.
Total lipid from rat milk formula and ileal samples were extracted using a modification of the Bligh and Dyer method (7). Ileum was sampled for fatty acid analysis, with care taken to insure there was no obvious contamination from residual luminal contents. Fatty acid methyl esters (FAME) were prepared using 14% BF3 in methanol (Sigma Chemical). FAME analyses were performed using a Hewlett Packard 5890 Gas Chromatography-Flame Ionization Detector (GC-FID). A BPX-70 column (25 m × 0.22 mm × 0.25 μm; SGE, Austin, TX) was used for the analysis with H2 as the carrier gas. FAME identities were determined using a Varian Star 3400 GC coupled to a Varian Saturn 2000 ion trap MS. FAME identities and double-bond position were determined by chemical ionization mass spectrometry using acetonitrile as reagent gas (37). FA levels are expressed as percentage, weight for weight (%wt/wt).
RNA preparation and real-time PCR.
Total RNA was isolated from ileal tissue (snap frozen in liquid N2) using the RNeasy Plus Mini Kit (Qiagen, Santa Clarita, CA) as described in the manufacturer's protocol. RNA concentration was quantified by ultraviolet spectrophotometry at 260 nm, and the purity and integrity were determined using a NanoDrop (Thermo Fisher Scientific, Wilmington, DE) (34).
RT real-time PCR assays were performed to quantify steady-state mRNA levels of selected cytokines (IL-6, IL-8, IL-12, IL-18, IL-23, and TNF-α). cDNA was synthesized from 0.2 μg of total RNA. Real-time PCR amplification was performed using Primer Express Software (Applied Biosystems, Foster City, CA). Target probe was labeled with fluorescent reporter dye. PreDeveloped TaqMan primers and probes were used for the detection of IL-6, IL-8, IL-12, IL-18, and IL-23. The following sequences was used for TNF-α (GenBank X66539): sense primer- 5′-gtgatcggtcccaacaagga-3′; anti-sense primer- 5′-gggccatggaactgatgaga-3′; and probe- 5′-cccatttgggaacttc-3′.
Reporter dye emission was detected by an automated sequence detector combined with ABI Prism 7700 Sequence Detection System software (Applied Biosystems). Real-time PCR quantification was then performed using TaqMan 18S controls.
Immunohistology and enumeration of Muc2- and Tff3-positive cells.
Serial sections of the ileum were stained for either Muc2 or Tff3. Briefly, after deparaffinization and rehydration, sections were incubated with either rabbit anti-Muc2 polyclonal antibody (Santa Cruz Biotechnology) or rabbit polyclonal Tff3 antibody (54) for 30 min, washed with PBS three times, and incubated with a biotinylated goat anti-rabbit secondary antibody (Vector Laboratories) for 30 min. Vectastain Elite ABC reagent (Vector Laboratories) was then applied, followed by diaminobenzidine as a substrate. Sections were counterstained with hematoxylin, dehydrated, and mounted on coverslips. Muc2- and Tff3-positive cells were counted from nine animals per experimental group, and the total number of epithelial cells per crypt-villus unit was also enumerated. Fifteen crypt-villus units were counted for each animal.
Statistical analyses between DF, FF, and FF + PSO groups were performed using ANOVA followed by Fisher paired least-significant difference and by the Student's t-test at the 95% confidence level. Analysis of NEC score between groups was accomplished using the Kruskal-Wallis test for nonparametric values followed by pairwise comparison using the Mann-Whitney test. The Pearson's chi-squared (χ2) test was used to analyze differences in incidence of NEC between groups. All statistical analyses were conducted using the statistical program StatPlus:mac LE for Macintosh computers (AnalystSoft, Alexandria, VA). All numerical data are expressed as means ± SE.
Oral administration of PSO reduces the severity and incidence of NEC.
There was no difference between experimental groups in the survival of C-section-delivered premature rats. The survival rates for these studies were as follows: DF 90% (18/20); FF 90% (18/20); and FF + PSO 95% (19/20). The degree of intestinal injury and the incidence of NEC were evaluated in prematurely born rats fed formula with or without PSO. Damage to the villi and submucosa was evident in the FF group and supplementation of PSO into formula significantly reduced ileal damage (Fig. 2) to a median histological NEC score of 1.5 compared with 2.0 in the FF group (P ≤ 0.01). The incidence of NEC was markedly decreased from 61% (11/18) in the FF group to 26% (5/19) in the FF + PSO group. The null hypothesis of no difference in NEC incidence between formula-fed groups was rejected (P = 0.033); PSO reduced the incidence of NEC by 57% (61% vs. 26%). In DF rats, the median histological score was 0.75 and incidence of NEC was 0% (0/18).
CLnAs are incorporated into the ileum.
The FF diet had no detectable PuA or other cis-trans isomers of 9,11,13 C18:3 such as the eleostearic acid isomers (Table 1). In contrast, FF + PSO contained 1.7% and 1.0% PuA and 9,11,13–18:3 isomers; these CLnA proportions are within the range of values reported very recently for several cultivars of pomegranate seeds (21). A conjugated linoleic acid (CLA), rumenic acid (cis-9, trans-11 18:2) was found at 0.2% in both groups.
PuA and other isomers were found in ileal lipids of the FF + PSO and, unexpectedly, at low levels in the FF group (Table 2). The PuA concentration was found to be only 28% of the total CLnA isomers detected (1.10 of 3.86 total) even though it was >60% in the feed. Rumenic acid was more than fourfold greater in the FF + PSO group (0.33) compared with the FF group.
PSO improves intestinal morphology.
Intestinal morphology was determined by light microscopy (Fig. 3A). The FF group had significantly shorter villi compared with healthy controls (171.2 ± 6.7 μm vs. 194.3 ± 5.9 μm). In the formula-fed pups, supplementation of PSO into formula significantly increased villi length (Fig. 3B), reaching values similar to those in the DF group (194.4 ± 8.6 μm; P ≤ 0.01 vs. DF or FF + PSO). To determine whether the increase in villus height in the FF + PSO group was a result of hypertrophy or hyperplasia, the total number of epithelial cells in each measured villus was counted, and results from these measurements were expressed as number of epithelial cells per micrometer of villus. There were no statistically significant differences seen between two groups, indicating that PSO caused hyperplasia of the epithelial cells and not hypertrophy (Fig. 3C).
PSO induces epithelial proliferation in the ileum.
To determine whether the PSO induces mitogenic effects on epithelial cells in the ileum, the cell proliferation marker PCNA was evaluated by immunohistochemistry, and PCNA-positive cells were enumerated (16). PCNA-positive cells were detected in the crypt epithelium and the lower half of villi (Fig. 4A). The number of PCNA-positive cells was significantly reduced in the FF group (13.6%) compared with healthy controls, indicating a significant decrease in the intestinal epithelial cell proliferation in formula-fed pups (Fig. 4B; *P ≤ 0.01 vs. DF or FF + PSO). Supplementation of PSO into the formula normalized this effect, and the number of PCNA-positive cells in the terminal ileum was similar to values seen in the DF group (21.5% in the FF + PSO vs. 21.7% in the DF group).
PSO reduces inflammatory response in the ileum.
Proinflammatory and effector cytokines IL-6, IL-8, IL-12, IL-18, IL-23, and TNF-α are the major cytokines associated with NEC pathogenesis and neonatal sepsis (26, 53). Thus gene expression of these selected cytokines was determined in the terminal ileum using RT-PCR. Expression of IL-6, IL-8, and TNF-α (Fig. 5A) was significantly increased in animals with NEC (FF) and normalized in the FF + PSO group to levels found in DF animals (*P ≤ 0.01, #P ≤ 0.05). Gene expression of IL-12 and IL-23 (Fig. 5B) was markedly increased in NEC animals compare to DF or FF + PSO (*P ≤ 0.01), but IL-23 mRNA levels in the FF + PSO group was significantly higher compared with DF rats (‡P ≤ 0.05). There was no effect on ileal gene expression of IL-18 (results not shown) between experimental groups.
Evaluation of Muc2 and Tff3 production.
Ileal Muc2 production was evaluated using immunohistochemistry, and enumeration of Muc2-positive cells in the ileum was compared among all experimental groups (Table 3). The signal of Muc2 was similar in all experimental groups. Ileal Tff3 was evaluated by enumeration of Tff3 positively stained cells. There was a significant decrease of Tff3-positive cells in the ileal tissue from the FF and FF + PSO groups compared with DF controls (P ≤ 0.01).
In this study, for the first time, we report that supplementation of PSO into milk formula reduces the incidence and severity of NEC in a neonatal rat model of NEC. The beneficial effects of PSO against NEC are associated with the improved enterocyte proliferation and reduced expression of proinflammatory cytokines in the site of injury. Oral administration of PSO does not affect Muc2 and Tff3 secretion in the terminal ileum.
Neonatal NEC is a multifactorial disease, which primarily affects prematurely born infants. Decreasing gestational age of newborns clearly correlates with increasing incidence of NEC (24). Another critical factor in NEC pathogenesis is the type and composition of infant diet. Breastfeeding has an advantage over infant formula feeding in the reduction of both human (39) and experimental NEC (17). Arachidonic acid (AA) and docosahexaenoic acid (DHA) have been shown to reduce NEC (1, 11), whereas the effect of conjugated FAs on the developing intestine and NEC pathogenesis have not been studied yet.
There are two types of structurally related conjugated octadecaenoic acids available in human diets, CLA (18:2) and CLnA (18:3). The CLA have been intensively studied for their beneficial biological activities in a variety of disease models (6), including mouse and pig models of colitis (3, 5), and a clinical trial in patients with Crohn's disease (4). In contrast, information concerning the biological function of CLnA is still scarce (55). Human milk contains a significant amount of the CLA (44), but the presence of the CLnA has yet to be established. In the present study, we showed that our rat milk formula contains only traces of the CLA (0.2%) and that the CLnA are absent. Supplementation of formula with 1.5% of PSO significantly increased CLnA levels but had no effect on the concentration of CLA, AA, and DHA in our formula (Table 1). Thus we were able to evaluate the direct effect of orally administered CLnA in the rat NEC model.
A clinical study by Carlson et al. (14) revealed that preterm infants fed a formula supplemented with AA and DHA had a lower incidence of NEC (14). In studies with rodent NEC models, first Akisu et al. (1) and then Caplan et al. (12) demonstrated that supplementation of PUFA into formula reduces the incidence of experimental NEC. Comparison of three different preparations of PUFA (AA + DHA, egg phospholipids, and DHA alone) showed a similar efficacy of each treatment (38). Interestingly, very low levels of either AA or DHA (0.7% or 0.5% of the total fatty acids) were sufficient to protect intestinal tissue against injury (38). The presence of either CLA or CLnA was not detected in the above-mentioned studies. Results from our study showed that supplementation of PSO into formula markedly reduced the incidence of NEC (from 61% to 26%) and severity of ileal damage in the rat NEC model. Because concentrations of CLA, AA, and DHA were the same in both FF and FF + PSO diets, we conclude that CLnA is responsible for the protective effect of the PSO.
Maternal milk affects the exposure to FA in the newborn intestine (31). Cao et al. (10) have shown that conjugated FA in maternal milk can be proportionally incorporated into liver phospholipids (PL) and change their FA composition. In our study, both CLA and CLnA were incorporated into ileum in a diet-dependent manner. Intestinal tissue from FF + PSO pups exhibited more then 10-fold greater CLnA compared with the control FF group. Moreover, our results show that there is active metabolism of PuA at the intestine of these immature rats. PSO feeding led to shift in CLnA profile from 60% PuA in the FF + PSO feeds to about 28% in the ileal lipids, whereas there was more than a fourfold increase in ileal rumenic acid (Table 2). These data are consistent and indicate that conversion at the level of the ileum is operative in very immature rats.
In the healthy intestine, epithelial homeostasis and tissue integrity is maintained by balancing the rate of cell proliferation and turnover. Milk-borne factors and nutrients can affect neonatal intestinal growth and intestinal cellular proliferation (16). In the FF group, a significant reduction of villus length was detected in the intact portions of the ileum compared with DF controls. Feeding of formula supplemented with PSO resulted in normalization of villus length. There was no significant difference in the number of enterocytes per micrometer of villus between experimental groups, suggesting that hyperplasia rather than hypertrophy is occurring. Intestinal epithelial proliferation was decreased in the FF group compared with the DF group, and this effect was completely reversed in rats fed with FF + PSO diet. These results indicate that PSO protects the epithelial barrier and preserves the intestinal architecture.
Recent studies show that CLnA may reduce inflammation and enhance the immune response (28). The small intestine is the largest immune organ in the body. Our and other laboratories showed that proinflammatory cytokines are important factors contributing to NEC pathogenesis (26, 27, 53, 56). Patients with NEC have increased serum levels of TNF-α and IL-6 (53). In addition, excessive IL-8 response is observed in the intestinal mucosa during NEC (47). Previously, we showed increased levels of IL-12, IL-18, and TNF-α in the rat NEC model (25, 26). In the present study, we found that gene expression of proinflammatory IL-6, IL-8, and TNF-α mRNA was markedly increased in the ileum of NEC rats compared with controls. Supplementation of the PSO into formula reduced these levels to values seen in controls.
IL-23 is a member of the IL-12 cytokine family secreted by activated dendritic cells and macrophages (49). Because of the shared p40 subunit between IL-12 (p40/p35) and IL-23 (p40/p19), original studies showing the critical role of IL-12 and T helper 1 (Th1) responses in intestinal inflammation (before IL-23 discovery) had to be reevaluated (46). It is clear now, that both IL-23 and IL-12 contribute to intestinal inflammation (30) by inducing Th17 and Th1 effector responses. However, the role of IL-23 in NEC pathogenesis has not been reported yet. Our results showed sevenfold increase in IL-23 and twofold increase in IL-12 gene expression in the ileum of rats with NEC compared with controls (DF). Feeding formula with PSO significantly reduced IL-23 and IL-12 mRNA levels, but IL-23 levels still did not completely normalize to values seen in the DF groups. These results suggest that PSO protects against NEC injury via downregulation of the inflammatory response in the developing intestinal mucosa.
In conclusion, this study demonstrates the orally administered PSO protects against NEC injury in the rat model. The protective effect is associated with improvement of enterocyte proliferation, protection of intestinal architecture, and reduction of inflammatory response in the site of injury, the terminal ileum. We hypothesize that PuA is important in the maintenance of structural and functional integrity of intestinal mucosa. This study introduces a novel nutritional factor, which should be considered for supplementation into formula for premature neonates.
This work was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development grant HD039657 (to B. Dvorak) and a gift from Mead Johnson Nutrition (to B. Dvorak).
B. Dvorak has grants with Mead Johnson, Meiji Dairies, and NIH. There are no conflicts of interest. J.T. Brenna has grants with NIH and Danone, and no conflict of interest.
Author contributions: C.F.C.-B., C.L.S., C.K.A.-R., and P.L. performed experiments; C.F.C.-B., C.L.S., P.L., J.T.B., and B.D. analyzed data; C.F.C.-B. and B.D. drafted manuscript; J.B.-R., R.H., J.T.B., Z.E.J., and B.D. edited and revised manuscript; J.T.B., Z.E.J., and B.D. interpreted results of experiments; Z.E.J. and B.D. conception and design of research; B.D. prepared figures; B.D. approved final version of manuscript.
Current affiliation for Christine F. Coursodon-Boyiddle: Ventana Medical Systems, Tucson, Arizona.
- Copyright © 2012 the American Physiological Society