AJP - GI Journal of Neurophysiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Am J Physiol Gastrointest Liver Physiol 291: G1041-G1050, 2006. First published July 6, 2006; doi:10.1152/ajpgi.00139.2006
0193-1857/06 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Figures
Right arrow All Versions of this Article:
291/6/G1041    most recent
00139.2006v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Yaylaoglu, M. B.
Right arrow Articles by Henning, S. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Yaylaoglu, M. B.
Right arrow Articles by Henning, S. J.

MUCOSAL BIOLOGY

Diverse patterns of cell-specific gene expression in response to glucocorticoid in the developing small intestine

Murat B. Yaylaoglu,1 Barbara M. Agbemafle,2 Thomas J. Oesterreicher,2 Milton J. Finegold,3 Christina Thaller,1 and Susan J. Henning2,4

1Verna and Marrs McLean Department of Biochemistry and Molecular Biology, and Departments of 2Pediatrics, 3Pathology, and 4Molecular and Cellular Biology, Baylor College of Medicine, Houston, Texas

Submitted 27 March 2006 ; accepted in final form 26 June 2006


   ABSTRACT
TOP
 ABSTRACT
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Although glucocorticoids are known to elicit functional maturation of the gastrointestinal tract, the molecular mechanisms of glucocorticoid action on the developing intestine have not been fully elucidated. Our previous microarray studies identified 66 transcripts as being rapidly induced in the jejunum following dexamethasone (Dex) administration to suckling mice. Now we report the specific cellular location of a subset of these transcripts. Mouse pups at P8 received Dex or vehicle and intestinal segments were collected 3–4 h later. Robotic-based in situ hybridization (ISH) was performed with digoxygenin-labeled riboprobes. Transcripts studied included Ndrg1, Sgk1, Fos, and two unknown genes (Gene 9 and Gene 36). As predicted, ISH revealed marked diversity of cellular expression. In small intestinal segments, Sgk1 mRNA was in all epithelial cells; Fos mRNA was confined to epithelial cells at the villus tip; and Ndrg1 and Gene 36 mRNAs were localized to epithelial cells of the upper crypt and villus base. The remaining transcript (Gene 9) was induced modestly in villus stroma and strongly in the muscle layers. In the colon, Ndrg1, Sgk1, and Gene 36 were induced in all epithelial cells; Gene 9 was in muscle layers only; and Fos was not detectable. For jejunal segments, quantitation of ISH signals in tissue from Dex-treated and vehicle-treated mice demonstrated mRNA increases very similar to those measured by Northern blotting. We conclude that glucocorticoid action in the intestine reflects diverse molecular mechanisms operating in different cell types and that quantitative ISH is a valuable tool for studying hormone action in this tissue.

dexamethasone; primary response genes; quantitative in situ hybridization

THE EPITHELIUM OF THE SMALL intestine matures in three distinct phases: 1) morphogenesis; 2) completion of morphogenesis coincident with onset of functional maturation; and 3) terminal maturation of functional properties (28, 32, 37). In rats and mice, which are by far the most extensively studied species, morphogenesis is essentially completed during the fetal period so that at birth there are well formed villi and rudimentary crypts (19, 32). Likewise, initial functional maturation occurs during fetal life so that by the time of normal delivery the intestine is specifically equipped to handle digestion and absorption of all components of the milk diet. Early in the third postnatal week, as weaning begins, terminal maturation occurs such that by the fourth postnatal week the epithelium displays the full array of adult functions (14, 19, 32). These ontogenic changes are best illustrated by consideration of carbohydrate digestion. During the suckling period when lactose is essentially the only dietary carbohydrate, hydrolytic capacity for lactose is high as a result of strong expression of lactase-phlorizin hydrolase throughout the small intestine. Enzymes such as sucrase-isomaltase, maltase-glucoamylase, and trehalase, required for digestion of carbohydrates found in solid food, are expressed at low or undetectable levels during the suckling period (14, 19). At the time of weaning these enzymes (i.e., the {alpha}-glucosidases) display marked increase of their activities secondary to increased levels of the respective mRNAs (19, 32, 35).

Numerous studies over the past 30 years have pointed to two primary mechanisms underlying terminal maturation of the rodent intestine during the third postnatal week. The normal onset of the changes appears to be triggered by an intrinsic timing mechanism, which remains poorly understood. After the onset of changes, the rate of maturation appears to be controlled by glucocorticoid hormones whose circulating concentrations rise dramatically during the third postnatal week (14, 19, 32). In rodents it is clear that this hormone surge also serves to coordinate terminal maturation throughout the entire gastrointestinal tract (18). Moreover, in humans the analogous rise of circulating cortisol at the end of term is believed to be an important stimulus to maturation of the human intestinal tract (19). Beyond these effects of endogenous glucocorticoid on intestinal maturation, early administration of exogenous glucocorticoid is capable of eliciting precocious maturation in both rodents and humans (14, 19). The latter in turn has important implications in the management of preterm infants in whom the immaturity of the intestine constitutes a major cause of morbidity and mortality. Thus elucidating the cellular and molecular mechanisms of glucocorticoid action on the developing intestine has significance both to our basic understanding of intestinal development and for the therapeutic use of these hormones.

Prior studies designed to understand the molecular mechanisms of glucocorticoid action on the developing rodent intestine have typically used either sucrase-isomaltase or trehalase as functional markers. For both of these enzymes, the dramatic increases in activity seen following glucocorticoid administration during the suckling period have been shown to be secondary to increased levels of the respective mRNA (25, 35) due in turn to induction of gene transcription (15, 44). Although the classical action of glucocorticoid hormones is to stimulate transcription of target genes following binding of the hormone receptor complex to response elements in promoter or enhancer regions, such effects are typically quite rapid, yielding peak levels of mRNA 2–6 h following hormone administration (2, 12). In contrast, studies of sucrase-isomaltase mRNA following administration of dexamethasone (Dex) to suckling mice (1, 6) have shown a lag of at least 6–8 h with significant levels of the mRNA being detected only at 12–24 h. Thus we hypothesized that, in the developing intestine, glucocorticoid induction of genes such as sucrase-isomaltase is a secondary effect, elicited in turn by a protein product from one or more primary response genes. To identify such genes we performed microarray analysis of intestinal tissue just 2 h after administration of Dex. This revealed 66 mRNAs that were increased by the hormone treatment (1). As with all such studies, the challenge following microarray profiling of gene expression is to determine which mRNA changes are indeed relevant to the question at hand.

In a tissue of great cellular complexity such as the intestine, the further analysis of microarray data can be facilitated by elucidating the location of the changes in gene expression. Based on work by numerous investigators over the last 30 years, the ability of glucocorticoid hormones to elicit precocious maturation of the small intestine is known to be mediated by proliferating cells in the upper third of the intestinal crypts (14, 19, 32). For example, increased expression of sucrase-isomaltase is observed only gradually, beginning at the base of the villi as newly programmed cells migrate from the crypts. Thus in our case the relevant primary responses to glucocorticoid would be predicted to be observed in either the epithelium or the underlying mesenchyme of the upper crypts. Thus the goal of the present study was to take five candidate primary response genes from the previous microarray study (chosen as described in RESULTS) and determine the cellular location of their expression by in situ hybridization (ISH) at early times following glucocorticoid administration to suckling mice. Since most of the mRNAs identified by microarray displayed relatively modest increases (2- to 16-fold), we needed a method that was capable of detecting quantitative changes of gene expression in a cell-specific manner. Because standard ISH methods are not quantitative, a secondary goal of the present study was to demonstrate that robotic-based ISH allows quantitation comparable to that achieved by Northern blotting of RNA from equivalently treated tissue. The data that we generated also show that, just as previously demonstrated in the mouse brain (9), this ISH approach further allows determination of different quantitative responses in different cells within the intestinal tissue.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Animals, Treatments, and Tissue Collection

Timed pregnant C57BL/6J mice were received from Jackson Laboratories (Bar Harbor, ME) on days 11–13 of gestation. They were housed individually and provided food (5001 Rodent Diet; PMI Nutrition International, Brentwood, MO) and acidified tap water ad libitum. Our Institutional Animal Care and Use Committee approved all animal housing and protocol details. On postnatal day 8 (P8), ten pups (2 from each of 5 litters) were injected subcutaneously with vehicle (0.8% ethanol in 0.15 M NaCl) and ten pups (2 from each of the same 5 litters) were injected subcutaneously with 0.4 µg/g body wt of Dex. Duodenum, jejunum, ileum, and ascending colon were collected from vehicle-treated and Dex-treated littermates at 3–4 h after injection. For Northern blotting (3 litters), all tissues were collected directly into RNAlater (Ambion, Austin, TX) and stored at 4°C until RNA isolation. For in situ hybridization analysis (2 litters), intestinal segments were flushed with ice-cold 0.9% saline solution. Portions (~1 cm in length) from the mid region of each segment were aligned in parallel in a Cryomold Intermediate (Tissue-Tek, Santa Cruz, CA) containing optimum cutting temperature compound (Tissue-Tek) and then frozen on dry ice. The frozen specimen block was removed and reoriented in optimum cutting temperature in a larger freezing chamber that would allow easier sectioning. Cryostat sectioning was performed at –10°C as described by Yaylaoglu et al. (42). In preliminary studies we experimented with sections ranging from 10- to 16-µm thickness and concluded that with this fragile tissue from postnatal animals the 16-µm sections gave the best results.

Robotic In Situ Hybridization

The ISH probes were generated from cDNA clones (in the vector pSPORT) obtained from the National Institute of Aging 15K cDNA clone set (36). The cDNA templates were amplified by PCR using a GCG clamp containing T7 (GCGTAATACGACTCACTATAG), Sp6 (GCGATTTAGGTGACACTATAG) or T3 (GCCAATTAACCCTCACT) primer. The PCR product was purified by using QIAquick spin columns (Qiagen, Germantown, MD) and used as the template for in vitro transcription of digoxygenin-labeled riboprobes as described by Yaylaoglu et al. (42). The ISH was performed using a robotic platform as reported previously (8, 42). The concentration of proteinase K found to be optimal for this tissue was 3 µg/ml. The hybridized complementary probe was detected by catalyzed reporter deposition using biotinylated tyramide followed by colorimetric detection of biotin with an avidin-alkaline phosphatase conjugate (8, 42). The latter yields a dark blue precipitate whose abundance is proportional to the number of detected transcripts (8).

Quantitative analysis of ISH data was performed on all jejunal samples using the Celldetekt protocol to determine cellular gene expression strengths (8). Using digital images, this method first identifies pixels representing precipitate and then applies a sliding-window technique to classify the expression level in each detected cell as strong, moderate, or weak. After the software detects the three levels of expression, it will pseudo-color the area in red, blue, or yellow, respectively. The software also detects cells without precipitate and sums these with the expressing cells to quantitate the total cell covered area that was included in the scan. The strong, moderate, and weak signals then can be counted with a second protocol called Cellcount that allows determination of the relative RNA abundance (8). This software assigns a value of 9 for strong signals, 4 for moderate signals, and 1 for weak signals and then normalizes the composite value on the basis of the total tissue area scanned. For each probe the normalized composite value for each hybridized section was expressed as a percentage of the maximum observed with that probe for all sections of each experimental group (i.e., vehicle and Dex).

RNA Isolation and Northern Blot Analysis

Total RNA was isolated from tissues by homogenization in guanidinium isothiocyanate and pelleting through a cesium chloride cushion as described elsewhere (29). Northern blots were made by using 10 µg of total RNA per lane on a 1% formaldehyde agarose gel. The RNA was transferred onto uncharged nylon membrane by capillary transfer with 10 x SSC and then UV cross linked. The blots were sequentially probed with various cDNAs obtained from the National Institute of Aging 15K cDNA clone set (36) and mouse Actb (3). In all cases, clone inserts were 32P-labeled by random primed oligo-labeling. Prehybridization and hybridization were performed according to procedures previously published (1). No blot was probed more than four times, including the Actb probe. All blots were subjected to autoradiography, and the size of each transcript was interpolated from the 18S and 28S rRNA bands.

Signals from Northern blots were quantified by phosphorimaging. To correct for loading variation, these data were expressed as a ratio of the hybridization signal of the band of interest to that of the constitutive marker Actb on the same blot. These ratios were then expressed as a percentage of the peak value of each blot. For graphical presentation, values for individual animals from each experimental group were calculated as means ± SE.

Statistical Analyses of Quantitative Data

For each transcript, statistical significance was assessed by two-way ANOVA, with treatment (vehicle vs. Dex) as one variable and method (ISH vs. Northern blotting) as the other variable. In the case of Gene 9 an additional two-way ANOVA was performed to compare the effects of treatment (vehicle vs. Dex) on levels of transcript in different cellular locations (villus core vs. muscle layers).


   RESULTS
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 GRANTS
 REFERENCES
 
Choice of Transcripts for ISH

For this ISH study we chose to focus on 5 of the 66 transcripts originally identified by microarray as being rapidly induced in the jejunum of suckling mice following Dex administration (1). Among the 66 were 11 transcripts for which a complete time course of induction had been established (1). From these 11 we chose three known genes that were likely candidates for mediating glucocorticoid-induced functional maturation of the intestine, based on their known roles in other tissues. These were Ndrg1 (N-myc downstream regulated 1), which has been reported to be associated with development and/or differentiation in several tissues (1), including human colon (40); Sgk1 (serum/glucocorticoid regulated kinase 1), which is known to mediate glucocorticoid and mineralocorticoid effects in various epithelia (23); and Fos (FBJ osteosarcoma oncogene), which is a transcription factor that has previously been implicated in intestinal maturation (6, 10, 31). In addition, because the microarray analysis had shown that the rapidly induced transcripts included a large proportion of unknown genes (47%), we also chose two of these, namely Gene 9 and Gene 36 (using the numbering system from the prior study). All five transcripts were known to be detectable by Northern blotting of total jejunal RNA. Moreover, because the previous time course analysis showed these transcripts to display peak induction between 2 and 6 h after Dex administration (1), for the present study tissue was collected 3–4 h after either the vehicle or Dex injection. Finally, as glucocorticoid-induced precocious maturation results in activation of different secondary response genes in different regions of the intestinal tract (e.g., sucrase-isomaltase in jejunum compared with bile acid binding protein in ileum), although our prior study had focused only on jejunum, in the present study we have examined the five transcripts in four different regions: duodenum, jejunum, ileum, and colon.

Cellular Location of Dex-Induced Transcripts

Figure 1 shows that despite the fragility of the P8 intestinal tissue (particularly the small intestine), acceptable frozen cross sections were obtained. Moreover, these tightly controlled ISH procedures allow visualization of quantitative changes of gene expression as indicated by the stronger signal in all the specimens from Dex-treated pups compared with those from vehicle-treated pups. To control for specificity, in addition to hybridization with antisense riboprobes (as seen in Fig. 1), serial sections were also hybridized with the respective sense riboprobes. As can be seen in Supplemental Figs. S1–S5, which are included in the online version of this article, although there was some variation from probe to probe, in the intensity of the gray "ghost" images, these sense hybridizations showed a total absence of blue precipitate. This indicates that both endogenous alkaline phosphatase and nonspecific binding of riboprobes were effectively eliminated by the rigorous hybridization and washing procedures used in this robotic protocol. The data shown in all of the main figures of this paper represent specific binding of the antisense probes and thus of the respective transcripts.


Figure 1
View larger version (33K):
[in this window]
[in a new window]
  		Fig. 1. Representative low power cross-sectional in situ hybridization (ISH) images for duodenum (Duo), jejunum (Jej), ileum (Ile), and colon (Col), 3–4 h after injection of P8 mice with vehicle (V), or dexamethasone (D). Sections were hybridized with antisense riboprobes for Ndrg1 (A), Sgk1 (B), Fos (C), Gene 9 (D), and Gene 36 (E) (x10 magnification).

 
The ISH revealed exquisite specificity and remarkable diversity of cellular location of gene expression in response to glucocorticoid. This can be discerned in Fig. 1 but is more apparent in the high-power images shown in Fig. 2 and Supplemental Figs. S1–S5. The details are given below.


Figure 2
View larger version (62K):
[in this window]
[in a new window]
  		Fig. 2. Higher power images of jejunal sections from dexamethasone (Dex)-treated P8 mice hybridized with Ndrg1 (A), Sgk1 (B), Fos (C), Gene 9 (D), and Gene 36 (E) (x40 magnification).

 
Ndrg1. Jejunal sections from Dex-treated pups showed strong induction of Ndrg1 transcripts in epithelial cells of the lower villi and upper crypts (Fig. 1A and 2A). This pattern was also observed in the other two regions of the small intestine (Fig. 1A and Supplemental Fig. S1); however, in the colon Ndrg1 was induced in all epithelial cells. In all segments from vehicle-treated pups the Ndrg1 transcripts appeared to be in the same respective location but at much lower abundance than in those from Dex-treated pups.

Sgk1. The mRNA for Sgk1 was barely detectable in any of the sections from control pups, whereas very strong signals were observed throughout the epithelium of the duodenum, jejunum, ileum, and colon from Dex-treated pups (Fig. 1B). High-power images shown in Fig. 2B and Supplemental Fig. S2 confirmed that the Dex induced Sgk1 transcripts are expressed in all epithelial cells of both small and large intestines. There does appear to be some variation in the level of expression with individual scattered cells showing more intense signal than others.

Fos. In contrast to the previous two patterns, Fos transcripts were detected only in the small intestine and in all three regions (duodenum, jejunum, and ileum) were found primarily in the epithelia of the villus tips (Figs. 1C and 2C and Supplemental Fig. S3). This appeared to be true for sections from vehicle-treated pups as well as from Dex-treated pups. The latter sections showed both greater signal intensity and a greater proportion of villi being positive. Both low-power (Fig. 1C) and high-power (Supplemental Fig. S3) images of the colon showed no blue precipitate, indicating that Fos transcripts were either absent or below the level of detection in this region of the intestinal tract.

Gene 9. Whereas all three previous transcripts were confined to various regions of the epithelium, transcripts for Gene 9 presented a very different pattern. Low-power images showed strong expression in the external muscle layers of all four regions (Fig. 1D) from Dex-treated animals. In addition, in the duodenum, jejunum, and ileum, transcripts were also induced in the lamina propria of the villus core. High-power images (Fig. 2D and Supplemental Fig. S4) reveal that sections from vehicle-treated animals displayed Gene 9 transcripts in the same respective locations. In the duodenum, jejunum, and ileum the magnitude of Dex induction appeared to be greater in the muscle layers than in the villus core. This observation is addressed further in Fig. 5 (see Differential Expression of Gene 9 in the Dex-Induced Intestine).


Figure 5
View larger version (28K):
[in this window]
[in a new window]
  		Fig. 5. Differential induction of Gene 9 in duodenum. A: standard (top) and pseudo-color (bottom) images of duodenal sections from vehicle-treated and Dex-treated P8 mice hybridized with antisense riboprobes for Gene 9. As in Fig. 3, the pseudo-color images show strong (red), moderate (blue), and weak (yellow) signal intensities. B: quantitative analysis of relative abundance of Gene 9 transcripts in lamina propria of villus core compared with external muscle layers. Values are given as mean ± SE. Open bars show values for sections from vehicle-treated animals; solid bars show values from Dex-treated animals.

 
Gene 36. Transcripts for Gene 36 displayed patterns reminiscent of those seen with Ndrg1. In duodenum, jejunum, and ileum from both vehicle-treated and Dex-treated animals, low-power images showed Gene 36 transcripts to be confined to the epithelium of the crypts and lower villi. High-power images (Fig. 2E and Supplemental Fig. S5) showed the Dex-induced transcripts to extend further along the villi, but in a declining gradient toward the villus tip. Just as for Ndrg1, in the colon, Gene 36 transcripts were found throughout the epithelium. In all regions, the magnitude of Dex induction of Gene 36 appeared to be somewhat less than that for Ndrg1 primarily because of higher basal levels of expression in the tissues from vehicle-treated pups.

Quantitation of Relative mRNA Abundance

As explained in detail in METHODS, our ISH data for jejunal samples were subjected to quantitative analysis by a two-step process. In the first step, the digital images were scanned with specially designed software that is able to detect strong, moderate, and weak levels of expression depending on the signal strength and will accordingly color the area with red, blue, or yellow. Typical pseudo-color images are shown in Fig. 3. Even without further analysis these images are of value because, for all transcripts studied, the greater abundance in tissue from Dex-treated pups is far easier to observe than it was in the native images shown in Fig. 1. The data shown in Fig. 3 confirmed that for all five transcripts the cellular location of expression under control conditions (vehicle injection) was the same as following Dex treatment. However, the pseudo-color images also revealed subtle differences in the magnitude of the Dex-induction in different cells. For example, for Sgk1 mRNA (Fig. 3B), although present in all epithelial cells, appears to be more strongly induced on the lower halves of the villi. For Gene 9 there is an even more dramatic difference, this time between the induction in the villus core compared with the muscle layers. This is documented further in Fig. 5 and discussed under Differential Expression of Gene 9 in the Dex-Induced Intestine.


Figure 3
View larger version (24K):
[in this window]
[in a new window]
  		Fig. 3. Low-power pseudo-colored images of jejunal sections from vehicle-treated and Dex-treated P8 mice hybridized with the 5 antisense riboprobes as indicated. Signal intensities are shown as strong (red), moderate (blue), or weak (yellow).

 
In the second step of the quantitative analysis, values were assigned to the strong, moderate, and weak signals, giving a composite number that was then normalized to the area of tissue scanned. Figure 4 shows that, when the data thus generated are compared with those obtained by phosphorimaging of Northern blots, for each transcript the results from the two methods were remarkably comparable. Statistical analysis (Fig. 4) showed that for all transcripts the effect of treatment (i.e., Dex vs. vehicle) was highly significant (P < 0.001), whereas the effect of method (Northern vs. ISH) was not significant (P > 0.3). Most importantly, the interaction term in the two-way ANOVA was also, in all cases, not significant (P > 0.15), indicating that the method of analysis did not significantly affect the conclusion regarding treatment. The comparability of the two methods is further illustrated in Fig. 4 by the numerical values for the average fold induction for each transcript in tissue from Dex-treated animals compared with vehicle-treated animals.


Figure 4
View larger version (20K):
[in this window]
[in a new window]
  		Fig. 4. Quantitative data for relative mRNA abundance as determined by Northern blotting compared with ISH. Values are shown as means ± SE for tissue from vehicle-treated (open bars) and Dex-treated (solid bars) P8 mice. Numbers shown within the solid bars are fold increase. Results of 2-way ANOVAs are shown to the right, where "treatment" refers to V vs. Dex and "method" refers to Northern vs. ISH.

 
Differential Expression of Gene 9 in the Dex-Induced Intestine

To illustrate the power of these ISH procedures to detect quantitative differences in different cellular locations, Fig. 5A presents both standard and pseudo-color images for Gene 9 transcripts in duodenum. As can be seen, in tissue from vehicle-treated pups moderate- and weak-intensity signals were observed in both the lamina propria of the villus core and the external muscle layers. Tissue collected 3–4 h after Dex injection showed induction in both regions, but with a distinct difference in magnitude. In the villus core, although some strong (red) signals were detected, the majority were moderate (blue) and some remained weak (yellow). In contrast in the muscle from Dex-treated animals the Gene 9 transcripts showed almost exclusively strong (red) signals. Results of quantitative analysis from the two regions are shown in Fig. 5B. These data confirmed the impressions from the images in showing greater Dex induction in the muscle (12.6-fold) compared with the villus core (4.9-fold). Two-way ANOVA of these data showed that not only was the Dex effect highly significant (P < 0.001), but the difference between muscle and villus core was also significant (P < 0.015). In addition, the interaction term in the ANOVA was significant (P < 0.02), indicating that location affected the magnitude of the Dex induction. These data illustrate that this method for quantitative ISH is a valuable tool for analyzing differential gene expression within distinct areas of a tissue.


   DISCUSSION
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
GRANTS
 REFERENCES
 
Although the molecular mechanisms of glucocorticoid hormone action have been studied for many years, to date there have been relatively few attempts to identify the entire array of primary response genes. At the time of our microarray study on postnatal intestine (1) there were only seven other comparable studies, and of these five were performed on cell lines in vitro and only two (11, 20) examined immediate early effects of glucocorticoids in vivo (in lung and liver, respectively). Subsequently there has been one additional in vivo report (2), this time in kidney; however, there have been no studies to assess the cellular location of the induced transcripts. In complex tissues such as lung, kidney, brain, and intestine, information regarding spatial location is critical to the ultimate understanding of the functional role of the primary response genes. In our present study, given that a subset of only five upregulated genes was studied, the most striking finding is the diversity of the cellular location of these rapidly induced transcripts. Of the five, one (Sgk1) was found to be expressed in the entire intestinal epithelium; another (Fos) was confined to epithelial cells at the tips of the villi; another (Gene 9) was modestly induced in the lamina propria and more strongly induced in the external muscle layers. Two genes (Ndrg1 and Gene 36) were induced specifically in the epithelium of the upper crypt and villus base. Thus from five genes we observed four distinct patterns of expression following glucocorticoid treatment. Notably, none of the genes was expressed in all cell types. For the three known genes this was not surprising since we avoided genes from the "metabolism" category of the Gene Ontology analysis (1). For the two unknown genes we made no a priori prediction as to their function and thus the location of their transcripts.

In terms of our goal of identifying genes that may mediate glucocorticoid induction of functional maturation of the intestinal epithelium, what emerges from this study is that Ndrg1 and Gene 36 are plausible candidates. Interestingly, in the Gene Ontology analysis of our previous study (1), Ndrg1 was found in the "development" category and was considered a potential candidate in view of this together with the report that it is associated with differentiation of colonic epithelium and can be found in the nucleus (40). Thus for Ndrg1 the cellular location of expression confirmed the earlier prediction that this would be a gene worthy of further study. We are also planning to pursue Gene 36, which has no homology to any other known genes and may in fact be a noncoding transcript. Sgk1, which was found to be induced in the entire epithelium, is unlikely to be the trigger for glucocorticoid-induced maturation. However, this kinase is known to shuttle in and out of the nucleus and to be capable of phosphorylating transcription factors (23). Thus Sgk1 could be an important accessory enzyme in glucocorticoid action via a role in phosphorylation of transcription factors such as GATA-4 (27, 38, 41) or cdx-2 (33).

The pattern of expression we observed with Fos is particularly interesting in relation to the prior literature. Fos was found in the "transcription" category of the Gene Ontology analysis (1) and has been previously postulated to be a mediator of the developmental effects of glucocorticoids on the intestinal epithelium (6, 31). However, our data suggest that it is not a likely candidate in view of the location of its expression in the villus tips. As noted earlier, glucocorticoid-induced precocious maturation reflects initiating events in the crypt that elicit a phenotypic change on the villus as a result of cellular replacement. Thus cells expressing Fos at 4 h after Dex administration would have desquamated ahead of the mature cells emerging from the crypts, making it difficult to conceive a mechanism in which Fos mediates acquisition of the mature phenotype. Our finding of Fos expression at the villus tips agrees well with a report by Yeh et al. (43), which showed that in adult rat jejunum the transient elevation of Fos after ischemia-reperfusion is confined to the upper 60% of the villi. On the other hand, because Fos is induced in many cell types in response to a variety of stimuli, we would predict that this gene is likely to have multiple patterns of expression in the intestinal mucosa. Since Chen et al. (10) have reported that Fos is involved in the developmental onset of expression of the apical sodium-dependent bile acid transporter in the rat ileum, it is possible that Fos is transiently expressed in the upper crypts during normal development. Such a finding would not be inconsistent with our conclusion because there is good evidence that the molecular mechanisms of normal development of the rodent small intestinal epithelium are different from those of glucocorticoid-induced precocious maturation (30, 39).

Although our interests have been in epithelial maturation, the finding that one of the primary response genes (Gene 9) is strongly induced in the muscle layers may be relevant to the clinical use of glucocorticoids in preterm infants. Whereas prenatal administration of glucocorticoid is generally recognized as having a beneficial effect on maturation of the intestinal mucosa (5, 19, 34), in recent years the use of high-dose glucocorticoid therapy in preterm infants following delivery has been associated with reports of significant incidence of intestinal perforation (16). Studies in a mouse model of high-dose Dex immediately after birth have shown thinning of the muscle layers concurrent with growth of the mucosa (17). Gordon et al. (17) have suggested that this adverse effect of Dex on the muscle layers may simply be the consequence of expansion of the mucosa; however, our finding that the transcript for Gene 9 is rapidly and strongly induced in the muscle layers suggests a more specific deleterious effect of glucocorticoid on this component of the intestinal tissue. The fact that this transcript is not expressed in the epithelium raises the possibility of a combination therapy in which the effects of glucocorticoid on the muscle layers might be blocked, thus allowing beneficial effects on the epithelium without the concurrent adverse consequences in the muscle layers.

Several of the technical aspects of the ISH methods used in this study are worthy of comment, particularly for application to intestinal tissue. First, although nonradioactive probes are generally recognized as desirable, the use of alkaline phosphatase for detection has been problematic in the intestine owing to the high levels of endogenous alkaline phosphatase expressed on the brush border membrane of villus enterocytes. The present protocol uses both HCl (0.2 m) and iodoacetamide (20 mM) to destroy the endogenous alkaline phosphatase. Moreover, the robotic aspect allows multiple applications of these procedures as well as the necessary subsequent washes. As can be seen in Supplemental Figs. S1–S5, sections hybridized with sense riboprobes showed no blue deposit on the villi, indicating complete inhibition of the endogenous alkaline phosphatase. The second technical point is that the special freezing chamber allows simultaneous sectioning and hybridization of four different intestinal segments. In our case we used this feature to examine gene expression in duodenum, jejunum, ileum, and colon of each animal. Alternate arrangement of tissue pieces may be more suitable for other studies (e.g., jejunum from 4 animals undergoing different treatments). Although analysis of gene expression along the longitudinal axis of the intestine can also be achieved by the "Swiss roll" technique (13), this method does not offer the diversity of application to individual samples. The third technical aspect is that the present study demonstrates that our robotic-based procedure is highly amenable to the generation of quantitative data. The pseudo-color images themselves (see Figs. 3 and 5) allow easy assessment of whether basal and induced mRNA are in the same or different cellular location. For the five transcripts studied here, the spatial pattern of expression was the same in control and Dex-treated animals, but in other studies (e.g., drug treatments), basal and induced transcripts could well be expected to show different patterns. A further subtlety illustrated by the present study is that this method of quantitation readily detects different levels of induction in different cells within a complex tissue (Fig. 5). The pseudo-color images can be analyzed in various ways, e.g., one could separately quantitate the effect of treatment on signals of strong, moderate, and weak intensity. The data we present show that when all three types of signal are counted in a weighted manner, the resulting composite number shows excellent correlation with quantitative data generated from Northern blots (Fig. 4).

Despite reports of quantitative methods of analysis more than 10 years ago, ISH continues to be applied primarily to qualitative studies. We suggest that this reflects the fact that most of the older methods have drawbacks of one type or another. For example, although Freeman (13) reported detailed and elegant quantitative data describing longitudinal and vertical gradients of mRNA expression in the rabbit small intestine using 35S-labeled probes followed by micro densitometry, this method has not found widespread use among gastrointestinal system investigators. Likewise, Jonker et al. (21) reported excellent resolution with 35S-riboprobes on rat intestinal sections, as well as good quantitative correlation between signals from 35S and Northern blots, but there are no subsequent citations indicating use of this protocol with intestinal tissue. It is now recognized that there are multiple disadvantages of 35S-labeled probes, specifically limited linearity of signal due to film saturation; inconvenience of procedures using radioisotopes; and minimal ability to reutilize probes as a result of their limited stability (~1 wk compared with nonradioactive probes, which we have successfully kept frozen for up to 3 yr). Several groups have reported quantitative ISH using nonradioactive probes with alkaline phosphatase as the method of detection. Among the earliest of these studies, Kiyama et al. (22) reported that, even without amplification, alkaline phosphatase is as sensitive as 35S, and Larsson et al. (24) documented the importance of several methodological parameters including light source stability, section thickness, probe concentration, and development time. Neither of these groups compared their quantitative data with relative mRNA abundance assessed by other methods. Kiyama et al. stated that their results for reserpine induction of tyrosine hydroxylase mRNA were comparable with those previously observed by Northern analysis. However, no direct comparisons were made and, most importantly, no statistical assessment of the reliability of values from ISH compared with Northern blotting. Leeuw and Pette (26) reported improved ISH quantitation with alkaline phosphatase by kinetic microphotometry. However, this method is very cumbersome and time consuming because it requires individual recording (2–10 min) of accumulation of reaction product in each section studied. Asan and Kugler (4) also reported quantitative ISH with alkaline phosphatase. Although these authors showed a linear increase of signal with section thickness, the variation was extremely wide, with SEs being >50% of the mean. This is in contrast to the very small SEs seen in our study (Fig. 4), which emphasize that this robotic-based protocol generates highly reproducible quantitative data. Moreover, as shown in a recent study of 200 transcripts in mouse brain (9), our automated procedures are readily amenable to high-throughput application.

In summary, in this study the use of robotic-based ISH has demonstrated the complexity of glucocorticoid action on the developing intestine. Although glucocorticoids are often considered "housekeeping" hormones because of their roles in regulation of basal metabolism of all cells, they are now recognized as having critical developmental roles in many tissues (7). In our microarray study (1), Gene Ontology analysis revealed that only 14 of the 66 upregulated transcripts were in the "metabolism" category. Thus the majority of the primary response cascades in the immature intestine may reflect specific rather than generic roles of glucocorticoids. This suggestion is supported by the diverse patterns of cellular gene expression observed in the present study. Since our study examined glucocorticoid effects during the second postnatal week, we would suggest that some of the rapidly induced transcripts reflect hormone action confined to the developmental period, whereas others reflect actions that are maintained into adulthood. Although the former are of greatest interest to our ongoing studies on the regulation of intestinal development, both have pharmacological significance in view of the widespread therapeutic uses of glucocorticoids in both infants and adults. The present study represents the first step in dissecting the molecular mechanisms of glucocorticoid action on intestinal tissue at the cellular level. We propose that further application of this approach to the entire set of glucocorticoid-induced genes in both adult and infant intestine will yield valuable spatial information, which can in turn provide both functional insight and therapeutic benefits.


   GRANTS
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
REFERENCES
 
This work was supported by National Institute of Child Health and Human Development Grant no. R01 HD14094 to S. J. Henning and by pilot/feasibility funds to C. Thaller from National Institute of Diabetes and Digestive and Kidney Diseases Grant P30 DK56338, which supports the Texas Gulf Coast Digestive Disease Center (TGC-DDC). Support by the Morphology Core of the TGC-DCC is also acknowledged.


   ACKNOWLEDGMENTS
 
The authors thank Dr. Gregor Eichele for expert advice regarding in situ hybridization.


   FOOTNOTES
 

Address for reprint requests and other correspondence: S. J. Henning, Dept. of Pediatrics, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030 (e-mail: shenning{at}bcm.tmc.edu)

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


   REFERENCES
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 

  1. Agbemafle BM, Oesterreicher TJ, Shaw CA, and Henning SJ. Immediate early genes of glucocorticoid action on the developing intestine. Am J Physiol Gastrointest Liver Physiol 288: G897–G906, 2005.[Abstract/Free Full Text]
  2. Almon RR, Lai W, DuBois DC, and Jusko WJ. Corticosteroid-regulated genes in rat kidney: mining time series array data. Am J Physiol Endocrinol Metab 289: E870–E882, 2005.[Abstract/Free Full Text]
  3. Alonso S, Minty A, Bourlet Y, and Buckingham M. Comparison of three actin-coding sequences in the mouse; evolutionary relationships between the actin genes of warm-blooded vertebrates. J Mol Evol 23: 11–22, 1986.[CrossRef][ISI][Medline]
  4. Asan E and Kugler P. Qualitative and quantitative detection of alkaline phosphatase coupled to an oligonucleotide probe for somatostatin mRNA after in situ hybridization using unfixed rat brain tissue. Histochem Cell Biol 103: 463–471, 1995.[CrossRef][ISI][Medline]
  5. Bauer CR, Morrison JC, Poole WK, Korones SB, Boehm JJ, Rigatto H, and Zachman RD. A decreased incidence of necrotizing enterocolitis after prenatal glucocorticoid therapy. Pediatrics 73: 682–688, 1984.[Abstract/Free Full Text]
  6. Blais S, Boudreau F, Thorneloe K, and Asselin C. Differential expression of fos and jun family members during murine postnatal intestinal development. Biol Neonate 69: 342–349, 1996.[ISI][Medline]
  7. Brown RW and Seckl JR. Glucocorticoid action in development. Curr Options Endocrinol Diabetes 12: 224–232, 2005.[CrossRef]
  8. Carson JP, Eichele G, and Chiu W. A method for automated detection of gene expression required for the establishment of a digital transcriptome-wide gene expression atlas. J Microsc 217: 275–281, 2005.[ISI][Medline]
  9. Carson JP, Ju T, Lu HC, Thaller C, Xu M, Pallas SL, Crair MC, Warren J, Chiu W, and Eichele G. A digital atlas to characterize the mouse brain transcriptome. PLoS Comput Biol 1: e41, 2005.[CrossRef][Medline]
  10. Chen F, Ma L, Al Ansari N, and Shneider B. The role of AP-1 in the transcriptional regulation of the rat apical sodium-dependent bile acid transporter. J Biol Chem 276: 38703–38714, 2001.[Abstract/Free Full Text]
  11. Clerch LB, Baras AS, Massaro GD, Hoffman EP, and Massaro D. DNA microarray analysis of neonatal mouse lung connects regulation of KDR with dexamethasone-induced inhibition of alveolar formation. Am J Physiol Lung Cell Mol Physiol 286: L411–L419, 2004.[Abstract/Free Full Text]
  12. Dean DM and Sanders MM. Ten years after: reclassification of steroid-responsive genes. Mol Endocrinol 10: 1489–1495, 1996.[Abstract]
  13. Freeman TC. Parallel patterns of cell-specific gene-expression during enterocyte differentiation and maturation in the small-intestine of the rabbit. Differentiation 59: 179–192, 1995.[CrossRef][ISI][Medline]
  14. Galand G. Brush border membrane sucrase-isomaltase, maltase-glucoamylase and trehalase in mammals. Comparative development, effects of glucocorticoids, molecular mechanisms, and phylogenetic implications. Comp Biochem Physiol B 94: 1–11, 1989.[CrossRef][Medline]
  15. Gartner H, Shukla P, Markesich DC, Solomon NS, Oesterreicher TJ, and Henning SJ. Developmental expression of trehalase: role of transcriptional activation. Biochim Biophys Acta 1574: 329–336, 2002.[Medline]
  16. Gordon PV, Marshall DD, Stiles AD, and Price WA. The clinical, morphologic, and molecular changes in the ileum associated with early postnatal dexamethasone administration: from the baby's bowel to the researcher's bench. Mol Genet Metab 72: 91–103, 2001.[CrossRef][ISI][Medline]
  17. Gordon PV, Price WA, and Stiles AD. Dexamethasone administration to newborn mice alters mucosal and muscular morphology in the ileum and modulates IGF-I localization. Pediatr Res 49: 93–100, 2001.[ISI][Medline]
  18. Henning SJ. Gastrointestinal development: an overview. In: Falk Symposium 94: The Gut as a Model in Cell and Molecular Biology, edited by Halter F, Winton D, and Wright NA. Dordrecht, The Netherlands: Kluwer Academic, 1997, p. 49–60.
  19. Henning SJ, Rubin DC, and Shulman RJ. Ontogeny of the intestinal mucosa. In: Physiology of the Gastrointestinal Tract, edited by Johnson LR. New York: Raven, 1994, p. 571–610.
  20. Jin JY, Almon RR, DuBois DC, and Jusko WJ. Modeling of corticosteroid pharmacogenomics in rat liver using gene microarrays. J Pharmacol Exp Ther 307: 93–109, 2003.[Abstract/Free Full Text]
  21. Jonker A, de Boer PA, van den Hoff MJ, Lamers WH, and Moorman AF. Towards quantitative in situ hybridization. J Histochem Cytochem 45: 413–423, 1997.[Abstract/Free Full Text]
  22. Kiyama H, Emson PC, Ruth J, and Morgan C. Sensitive non-radioisotopic in situ hybridization histochemistry: demonstration of tyrosine hydroxylase gene expression in rat brain and adrenal. Brain Res Mol Brain Res 7: 213–219, 1990.[Medline]
  23. Lang F and Cohen P. Regulation and physiological roles of serum- and glucocorticoid-induced protein kinase isoforms. Sci STKE 2001: RE17, 2001.[Medline]
  24. Larsson LI, Traasdahl B, and Hougaard DM. Quantitative non-radioactive in situ hybridization. Model studies and studies on pituitary proopiomelanocortin cells after adrenalectomy. Histochemistry 95: 209–215, 1991.[CrossRef][ISI][Medline]
  25. Leeper LL and Henning SJ. Development and tissue distribution of sucrase-isomaltase mRNA in rats. Am J Physiol Gastrointest Liver Physiol 258: G52–G58, 1990.[Abstract/Free Full Text]
  26. Leeuw T and Pette D. Kinetic microphotometric evaluation of in situ hybridization for mRNA of slow myosin heavy chain in type I and C fibres of rabbit muscle. Histochemistry 102: 105–112, 1994.[CrossRef][ISI][Medline]
  27. Liang Q, Wiese RJ, Bueno OF, Dai YS, Markham BE, and Molkentin JD. The transcription factor GATA4 is activated by extracellular signal-regulated kinase 1- and 2-mediated phosphorylation of serine 105 in cardiomyocytes. Mol Cell Biol 21: 7460–7469, 2001.[Abstract/Free Full Text]
  28. Montgomery RK, Mulberg AE, and Grand RJ. Development of the human gastrointestinal tract: twenty years of progress. Gastroenterology 116: 702–731, 1999.[CrossRef][ISI][Medline]
  29. Nanthakumar NN and Henning SJ. Ontogeny of sucrase-isomaltase gene expression in rat intestine: responsiveness to glucocorticoids. Am J Physiol Gastrointest Liver Physiol 264: G306–G311, 1993.[Abstract/Free Full Text]
  30. Nanthakumar NN and Henning SJ. Distinguishing normal and glucocorticoid-induced maturation of intestine using bromodeoxyuridine. Am J Physiol Gastrointest Liver Physiol 268: G139–G145, 1995.[Abstract/Free Full Text]
  31. Nsi-Emvo E, Foltzer-Jourdainne C, Raul F, Gosse F, Duluc I, Koch B, and Freund JN. Precocious and reversible expression of sucrase-isomaltase unrelated to intestinal cell turnover. Am J Physiol Gastrointest Liver Physiol 266: G568–G575, 1994.[Abstract/Free Full Text]
  32. Quaroni A and Hochman J. Development of intestinal cell culture models for drug transport and metabolism studies. Adv Drug Delivery Res 22: 3–52, 1996.[CrossRef]
  33. Rings EH, Boudreau F, Taylor JK, Moffett J, Suh ER, and Traber PG. Phosphorylation of the serine 60 residue within the Cdx2 activation domain mediates its transactivation capacity. Gastroenterology 121: 1437–1450, 2001.[CrossRef][ISI][Medline]
  34. Shulman RJ, Schanler RJ, Lau C, Heitkemper MA, Ou C, and Smith EO. Early feeding, antenatal glucocorticoids, and human milk decrease intestinal permeability in preterm infants. Pediatr Res 44: 519–523, 1998.[ISI][Medline]
  35. Solomon NS, Gartner H, Oesterreicher TJ, and Henning SJ. Development of glucocorticoid-responsiveness in mouse intestine. Pediatr Res 49: 782–788, 2001.[ISI][Medline]
  36. Tanaka TS, Jaradat SA, Lim MK, Kargul GJ, Wang X, Grahovac MJ, Pantano S, Sano Y, Piao Y, Nagaraja R, Doi H, Wood WH III, Becker KG, and Ko MS. Genome-wide expression profiling of mid-gestation placenta and embryo using a 15,000 mouse developmental cDNA microarray. Proc Natl Acad Sci USA 97: 9127–9132, 2000.[Abstract/Free Full Text]
  37. Traber PG and Wu GD. Intestinal development and differentiation. In: Gastrointestinal Cancers: Biology, Diagnosis, and Therapy, edited by Rustgi AK. Philadelphia, PA: Lippincott-Raven, 1995, p. 21–43.
  38. Tremblay JJ and Viger RS. Transcription factor GATA-4 is activated by phosphorylation of serine 261 via the cAMP/protein kinase a signaling pathway in gonadal cells. J Biol Chem 278: 22128–22135, 2003.[Abstract/Free Full Text]
  39. Tung J, Markowitz AJ, Silberg DG, and Traber PG. Developmental expression of SI is regulated in transgenic mice by an evolutionarily conserved promoter. Am J Physiol Gastrointest Liver Physiol 273: G83–G92, 1997.[Abstract/Free Full Text]
  40. Van Belzen N, Dinjens WN, Diesveld MP, Groen NA, van der Made AC, Nozawa Y, Vlietstra R, Trapman J, and Bosman FT. A novel gene which is up-regulated during colon epithelial cell differentiation and down-regulated in colorectal neoplasms. Lab Invest 77: 85–92, 1997.[ISI][Medline]
  41. Wang J, Paradis P, Aries A, Komati H, Lefebvre C, Wang H, and Nemer M. Convergence of protein kinase C and JAK-STAT signaling on transcription factor GATA-4. Mol Cell Biol 25: 9829–9844, 2005.[Abstract/Free Full Text]
  42. Yaylaoglu MB, Titmus A, Visel A, Alvarez-Bolado G, Thaller C, and Eichele G. Comprehensive expression atlas of fibroblast growth factors and their receptors generated by a novel robotic in situ hybridization platform. Dev Dyn 234: 371–386, 2005.[CrossRef][ISI][Medline]
  43. Yeh KY, Yeh M, Glass J, and Granger DN. Rapid activation of NF-kappaB and AP-1 and target gene expression in postischemic rat intestine. Gastroenterology 118: 525–534, 2000.[CrossRef][ISI][Medline]
  44. Yeh KY, Yeh M, and Glass J. Glucocorticoids and dietary iron regulate postnatal intestinal heavy and light ferritin expression in rats. Am J Physiol Gastrointest Liver Physiol 278: G217–G226, 2000.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Am. J. Physiol. Gastrointest. Liver Physiol.Home page
T. Bosse, J. J. Fialkovich, C. M. Piaseckyj, E. Beuling, H. Broekman, R. J. Grand, R. K. Montgomery, and S. D. Krasinski
Gata4 and Hnf1{alpha} are partially required for the expression of specific intestinal genes during development
Am J Physiol Gastrointest Liver Physiol, May 1, 2007; 292(5): G1302 - G1314.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Figures
Right arrow All Versions of this Article:
291/6/G1041    most recent
00139.2006v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Yaylaoglu, M. B.
Right arrow Articles by Henning, S. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Yaylaoglu, M. B.
Right arrow Articles by Henning, S. J.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2006 by the American Physiological Society.