| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
The Burnham Institute, La Jolla, California 92037
Address all correspondence and requests for reprints to: Dr. Gregg Duester, The Burnham Institute, 10901 North Torrey Pines Road, La Jolla, California 92037. E-mail: greggd{at}ljcrf.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although much has recently been learned about the role of retinoic acid receptors in vitamin A function, relatively little is known about the spatiotemporal regulation of retinoic acid synthesis from retinol. Biochemical studies have indicated that this is a two-step oxidative process in which retinol is first oxidized to retinal, followed by further oxidation of retinal to retinoic acid (reviewed in Ref. 10). Candidates for a physiologically relevant retinol dehydrogenase have been found among the five known classes of human medium-chain alcohol dehydrogenase (ADH). With the exception of class I ADH, enzymes in the ADH family are inefficient in the oxidation of ethanol and may have evolved to oxidize other alcoholic substrates such as retinol. Indeed, human class I, class II, and class IV ADHs have been shown to have significant retinol dehydrogenase activities in vitro with efficiencies increasing in that order (11, 12), and mouse class IV ADH has been shown to perform this reaction in extracts of epidermis, a retinoid target tissue (13). The mouse ADH family has been demonstrated to be simpler than that of the human with only classes I, III, and IV detected (14). Because class III ADH exhibits no in vitro activity for retinol oxidation in both rodents and humans (11, 15), it appears that only class I and class IV ADHs are conserved as potential alcohol/retinol dehydrogenases in the mouse. The expression patterns of mouse ADHs in gastrulating embryos (16), adult testis and epididymis (17), and adult epidermis (18) suggest that they may participate in local retinoic acid synthesis. Another candidate for a physiologically relevant retinol dehydrogenase could be a rat liver microsomal retinol dehydrogenase, a member of the short-chain dehydrogenase/reductase family (19), although its preference for NADP/NADPH as co-factor suggests that this enzyme could function more in the reductive direction to convert retinal to retinol for retinoid storage in the liver (10).
To further study the possible link between class I and class IV ADHs and physiological retinoic acid production, we have examined the expression of these ADHs in embryonic and adult adrenal glands using in situ hybridization and immunohistochemistry and examined the local production of retinoic acid from these organs by means of a bioassay. Our results are consistent with the possibility that either or both class I and class IV ADHs are involved in retinoic acid synthesis in the adrenal gland, and we postulate that the adrenal gland may serve as an endocrine source of this hormone in both the embryonic and adult mouse.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Tissue sectioning and in situ hybridization
Adult adrenal glands or whole embryos at stage E16.5 were fixed,
embedded in paraffin, sectioned at 6 µm, and processed for in
situ hybridization under high stringency conditions as described
previously (20). The hybridization probes consisted of
35S-labeled antisense RNAs derived from the full-length
complementary DNAs (cDNAs) for mouse class I ADH (21, 22) and class IV
ADH (14) subcloned in pBluescript II KS (Stratagene Cloning Systems,
Inc., La Jolla, CA). Antisense RNA was synthesized using
[
-35S]UTP and T3 or T7 RNA polymerase as described
(20). As a control to monitor background detection, we used a sense RNA
probe transcribed from the plasmid containing the mouse class I ADH
cDNA. We have previously shown that there is no cross-hybridization
observed between the ADH classes under high stringency conditions in
either Northern or Southern blots, probably due to the low ADH
interclass sequence identity that is in the 59–69% range (14).
Slides were exposed to emulsion for 2–3 weeks before development. Staining of hybridized sections was performed with neutral red, and adjacent nonhybridized sections were stained with hematoxylin and eosin as described previously to observe tissue morphology (23). Hybridization results were observed by dark-field microscopy, and stained adjacent sections were observed by bright-field microscopy.
Purification of ADH antibodies
Rabbit polyclonal antisera against mouse class I, III, and IV
ADHs were generated from four monthly injections of 200 µg of each
protein expressed in E. coli as a fusion protein with
Schistosoma glutathione-S-transferase (GST) using the
pGEX-4T-2 plasmid vector system (Pharmacia Biotech, Uppsala, Sweden).
To remove residual cross-reactivity to the other classes of ADH, each
crude antiserum was preabsorbed by incubation with 100 µg of the
other two classes of GST-ADH fusion proteins resolved by SDS-PAGE and
electroblotted onto filter strips.
Affinity purification of either class I or class IV ADH-specific antibodies was performed using 100 µg aliquots of each GST-ADH fusion protein electroblotted to polyvinylidene difluoride (PVDF) membrane. These small filter strips were first blocked for 1 h with 5% BSA in PBS and then added to a 1 ml volume of preabsorbed antiserum diluted 1:10 in the same solution. After incubation for 24–72 h with gentle agitation at 4 C, the filters were washed 3 x 10 min with excess PBS, then placed in 1 ml ice-cold 100 mM glycine (pH 2.2). The filter was incubated 3 min on ice with occasional gentle up and down pipetting to facilitate elution of bound antibodies. The 1 ml volume was removed from the filter and neutralized by addition of a titrated amount of 2 M Tris-base. BSA was added to 5% as a stabilizing agent.
Concentrations of purified antibodies were estimated by comparison of dot blot signals of serially diluted preparations to known concentrations of commercially available rabbit IgG detected by horse radish peroxidase-conjugated goat-antirabbit antiserum (Sigma Chemical Co., St. Louis, MO). Western blot analysis of each of the GST-ADH fusion proteins (24) was used to verify that each preparation of purified antibodies reacted only with their cognate class of GST-ADH fusion protein, and lacked cross-reactivity either to GST or to the other classes of GST-ADH fusion proteins.
Immunohistochemistry
Sections of adult adrenal gland or whole E16.5 embryos embedded
in paraffin as described above were subjected to immunohistochemical
analysis using the Vectastain Elite ABC kit (Vector Laboratories, Inc.,
Burlingame, CA). Incubation was performed for 30 min with purified
antibody serially diluted in normal blocking solution provided in the
Vectastain kit. Incubation with diaminobenzidine (DAB) for color
detection was for 3 min, followed by dilution in tap water to stop the
reaction. Slides were lightly counterstained with hematoxylin before
mounting and observation by bright-field microscopy. Optimal class I
ADH staining was obtained using class I ADH affinity purified antibody
at 0.05 µg/ml. Optimal class IV ADH staining was obtained using class
IV ADH purified antibody at 0.01 µg/ml. Control incubations with
either serially diluted preimmune sera or purified antibody to
Schistosoma GST were used to monitor background detection.
The GST antibody was prepared from the rabbit antisera used in this
study by affinity purification as described above using
Schistosoma GST expressed in E. coli and
electroblotted onto PVDF membrane.
Bioassay detection of retinoic acid
Retinoic acid was detected using a bioassay that employs the
retinoic acid reporter cell line F9-RARE-lacZ (Sil-15). This
cell line was derived from stable transfection of mouse F9
teratocarcinoma cells with a transgene in which expression of
lacZ encoding ß-galactosidase is driven by a promoter
linked to a retinoic acid response element (25). In that study,
F9-RARE-lacZ cells were used to visually detect retinoic
acid in tissue explants grown on top of the reporter cells by
monitoring ß-galactosidase activity produced by the reporter cells in
response to retinoic acid diffusing from the tissue into the
surrounding media. This reporter cell line has also been used to
quantitate retinoic acid in tissue homogenates added to the growth
media (26). As indicated previously, this assay is optimal for
detection of all-trans-retinoic acid but will also detect
the 9-cis and 13-cis isomers of retinoic acid
with lower efficiency (26). Thus, a limitation of the assay is that the
results reflect the combined levels of all biologically active retinoid
ligands in the tissues examined. Also, because the homogenate assay
measures changes in absorbance at 650 nm as an index of
ß-galactosidase activity, there may be other factors present in
tissues that effect 650 nm absorbance.
F9-RARE-lacZ cell stocks were maintained in L15 CO2 medium with G418 selection as described previously (25). The tissue explant assay described above is particularly useful for analysis of tissues with a limited sample size such as the embryonic adrenal gland as examined here. For our tissue explant studies, F9-RARE-lacZ cells were grown in gelatin-coated 24-well plates until 80–90% confluent, at which time dissected adrenal and liver tissues from E16.5 embryos were placed on top of the monolayer and incubation continued for 18 h. After incubation, the reporter cells were fixed in 1% glutaraldehyde, and ß-galactosidase activity was visualized with 5-bromo-4-chloro-3-indolyl-ß-D-galactopyranoside (X-gal) as described (27).
For quantitation of retinoic acid in adult tissues, we used a modification of methods previously described (26). Tissue was dissected, immediately weighed, and frozen. Tissue (10–20 mg) was diluted with 100 µl of L15 CO2 medium, manually homogenized in a 1.5 ml microcentrifuge tube, subjected to three cycles of freeze/thaw (dry ice/ethanol bath for 5 min followed by 37 C bath for 5 min), followed by incubation for 1 h at 4 C and then centrifugation at 12,000 x g for 5 min at 4 C. A 100-µl volume of supernatant was removed and serial dilutions made using L15 CO2 medium. Dilutions of the homogenates (100 µl total volume) were added to F9-RARE-lacZ cells grown to 80–90% confluency in 96-well microtiter plates. After 18 h incubation, cells were fixed in 1% glutaraldehyde and ß-galactosidase activity was visualized as described above. The intensity of the blue reaction product was measured with an ELISA microtiter plate reader at an absorbance of 650 nm. The concentration of retinoic acid in each tissue homogenate was calculated using the dilution factor for each tissue examined plus a standard curve created by treating F9-RARE-lacZ cells with known concentrations of all-trans-retinoic acid and plotting the concentration vs. the A650.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Class IV ADH mRNA at E16.5
was observed in the adrenal gland, but not the kidney, with adrenal
expression appearing to be more diffuse than that observed for class I
ADH mRNA (Fig. 1B). The morphology of these tissues was observed by
hematoxylin-eosin staining (Fig. 1C).
|
) as well as a control in
situ hybridization using a sense riboprobe to demonstrate the low
level of background (Fig. 2B). A class I ADH antisense riboprobe
revealed localization of the mRNA to the cortex with a strong signal in
the zona fasiculata and somewhat weaker signal in the zona reticularis
but with little or no detection in the outermost region of the cortex
(zona glomerulosa) and no detection in the medulla (Fig. 2C). Class IV
ADH mRNA was localized in the cortex in a radial variegated pattern in
which only about 50% of the zona fasiculata and zona glomerulosa were
positive; there was minimal detection of class IV ADH mRNA in the zona
reticularis and no detection in the medulla (Fig. 2D). The above
results demonstrated that neither class I ADH nor class IV ADH mRNAs
were detectable in the medulla, indicating that expression of both is
cortex-specific.
|
) and
class IV ADH protein (Fig. 3B) in the developing cortical cells of the
adrenal gland that are located primarily around the circumference of
the organ. Class IV ADH protein appeared to be present in fewer
cortical cells than class I ADH, and both ADHs were absent in most
cells near the center of the gland where the medulla is developing. A
control using Schistosoma GST antibodies indicated a low
level of background detection (Fig. 3C).
|
, D and E). Within the cortex,
class I ADH detection was further limited to the zona fasiculata, with
a lower signal in the zona reticularis and little or no signal evident
in the zona glomerulosa (Fig. 3G). The class I ADH signal in the zona
fasiculata was found to encompass about 90–95% of this tissue;
however, variegated streaks of cells extending centripetally from the
glomerulosa/fasiculata boundary were also evident that exhibited little
or no detection above background (Fig. 3D). Such a variegated pattern
of expression was much more evident for class IV ADH in the cortex,
which exhibited a striking radial pattern of expression that
encompassed about 50% of the cells of the cortex, whereas directly
adjacent cells had little or no signal above background levels (Fig. 3E). In the radial zones where class IV ADH expression was clearly
evident, this enzyme was easily detected in cells of both the zonas
glomerulosa and fasiculata but was found to be nearly absent in cells
of the zona reticularis (Fig. 3H). Controls using purified
Schistosoma GST antibodies revealed the level of background
detection in the adult adrenal gland (Fig. 3, F and I).
Retinoic acid detection in embryonic and adult adrenal gland
A sensitive bioassay using a retinoic acid reporter cell line was
used to determine if the embryonic adrenal gland contains retinoic
acid. A tissue explant bioassay of two E16.5 adrenal glands (< 1 mg
wet weight combined) resulted in a significant level of retinoic acid
detection in the surrounding reporter cells (Fig. 3J). In contrast, a
tissue explant bioassay of a dissected fragment of E16.5 liver (3 mg
wet weight) resulted in a relatively low level of retinoic acid
detection (Fig. 3K).
Retinoic acid concentrations in adult mouse tissues were quantitated by
treating the reporter cells with homogenates of adrenal gland, liver,
brain, heart, and spleen tissue. Retinoic acid concentrations
determined by the homogenate bioassay were found to be 15.5 ± 9.9
pmol/g for adrenal, 24.8 ± 9.2 pmol/g for liver, and < 1.0
pmol/g for brain, heart, and spleen (Fig. 4).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It is well established that retinoic acid administered in the diet can replace the nonvision functions of vitamin A in vitamin A-deficient animals (7, 8), thus showing that under these circumstances retinoic acid can be distributed throughout the body and function essentially in an endocrine fashion. This, along with the detection of low levels of endogenous retinoic acid in the bloodstream (9), raises the question whether an organ exists that naturally functions as an endocrine source of retinoic acid. Embryonic and adult adrenal glands are not affected by vitamin A deficiency (1, 2) or retinoic acid receptor mutations (28), suggesting that the adrenal gland may not be a retinoid target tissue or that its requirement for retinoic acid is quite low. Also, other epithelial tissues such as skin that are clearly dependent on retinoic acid for proper differentiation exhibit very low levels of retinoic acid synthesis (13), whereas high levels have been detected in this study in both embryonic and adult adrenal glands. Therefore, this apparent excess of adrenally produced retinoic acid detected by our studies may serve as a primarily endocrine source to support retinoid target tissues elsewhere in the body. The highly vascularized nature of the adrenal cortex is well designed for export of other lipophilic hormones similar to retinoic acid, particularly glucocorticoids and mineralocorticoids, from adrenocortical cells to the bloodstream (29). Thus, retinoic acid synthesized by this organ could be efficiently delivered to the bloodstream in a manner similar to that for steroid hormones. Our observation that class I and class IV ADHs are not expressed in all cells of each zone of the cortex suggests that the function of these enzymes may be more consistent with production of an exported product (i.e. retinoic acid) rather than a factor needed by the adrenocortical cells themselves, i.e. an export function could be sufficiently performed by a subset of adrenocortical cells.
The cells of the adrenal cortex are divided morphologically into three concentric zones, the outer zona glomerulosa, the intermediate zona fasiculata, and the inner zona reticularis (30). These zones are also functionally distinct, with mineralocorticoids being synthesized in the zona glomerulosa, and glucocorticoids being synthesized in the zona fasiculata and zona reticularis (29). Whereas our studies indicate that cortical expression of both class I and class IV ADHs is variegated and not strictly limited to any one zone, class I ADH is expressed primarily in the zona fasiculata, whereas class IV ADH is expressed primarily in the zona glomerulosa and zona fasiculata. The observation that class IV ADH expression is limited to only a subset of radial columns of cells spanning the cortical zones is of particular interest for studies on cellular differentiation in this tissue. Cellular turnover in the mouse adrenal cortex has been reported to progress over a time period of 200 days, with cell proliferation occurring at the interface of the zona glomerulosa and fasiculata (31). Pulse labeling studies suggest that cells migrate bidirectionally from this interface in a radial manner with some cells migrating centrifugally outward into the narrow zona glomerulosa, whereas most cells migrate centripetally inward through the zona fasiculata to the zona reticularis (31). The striking radially variegated pattern of class IV ADH expression in the cortex provides further support for the above model of radial cellular differentiation and migration in this organ. Also, support for this model has come from the expression pattern of a mouse transgene consisting of a lacZ fusion to a steroid 21-hydroxylase promoter (32). This transgene expresses in a radially variegated pattern in the adult adrenal cortex very similarly to what we have demonstrated for class IV ADH endogenous expression. Collectively, these findings suggest that columns of cells extending from the glomerulosa to the reticularis are in the same cell lineage. Moreover, to our knowledge class IV ADH is the first gene identified that demonstrates this radially variegated pattern in its endogenous expression because native expression of the steroid 21-hydroxylase gene in question was found to be homogenous throughout the cortex (32). Class IV ADH may therefore prove to be a good cell lineage marker for further studies of adrenocortical cell proliferation.
The issue of which enzymes are involved physiologically in the synthesis of retinoic acid in any given organ or at any stage of development is not resolved. Here we have focused upon the potential roles of class I and/or class IV ADHs in retinol oxidation in the adrenal cortex. However, the ability of ADHs to oxidize ethanol or other alcohols should not be overlooked. Ethanol oxidation by class I ADH in the liver and digestive tract may represent a true physiological function designed to eliminate ethanol produced by intestinal flora (33). In contrast to class I ADH, the biochemical properties of class IV ADH (12, 13, 15) indicate that its participation in ethanol oxidation will occur only when this substance reaches toxic pharmacological levels, thus suggesting it catalyzes a physiological function distinct from ethanol oxidation. It remains to be determined whether retinol oxidation is the primary function of class I and class IV ADHs in the adrenal cortex.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received February 10, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  T. Sun, R. L. Surles, and S. A. Tanumihardjo Vitamin A Concentrations in Piglet Extrahepatic Tissues Respond Differently Ten Days after Vitamin A Treatment J. Nutr., June 1, 2008; 138(6): 1101 - 1106. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Koubova, D. B. Menke, Q. Zhou, B. Capel, M. D. Griswold, and D. C. Page Inaugural Articles: Retinoic acid regulates sex-specific timing of meiotic initiation in mice PNAS, February 21, 2006; 103(8): 2474 - 2479. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Molotkov, X. Fan, L. Deltour, M. H. Foglio, S. Martras, J. Farres, X. Pares, and G. Duester Stimulation of retinoic acid production and growth by ubiquitously expressed alcohol dehydrogenase Adh3 PNAS, April 16, 2002; 99(8): 5337 - 5342. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Deltour, M. H. Foglio, and G. Duester Metabolic Deficiencies in Alcohol Dehydrogenase Adh1, Adh3, and Adh4 Null Mutant Mice. OVERLAPPING ROLES OF Adh1 AND Adh4 IN ETHANOL CLEARANCE AND METABOLISM OF RETINOL TO RETINOIC ACID J. Biol. Chem., June 11, 1999; 274(24): 16796 - 16801. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
