Acta Otolaryngol (Stockh) 1997; 117: 738-743

Localization of 1 l p -Hydroxysteroid Dehydrogenase: Specific Protector of the Mineralocorticoid Receptor in Mammalian Olfactory Mucosa

ROBERT C. KERN and DIMITRI Z. PITOVSKI Laboratory fo r Molecular Reseurch in Otolaryngology, Depurinien t of Otolarygology-- Herid rind Neck Surgery, Nurthivestern University Medical Scliool, Cliicago. IL 6061 1 , USA

Kern RC, Pitovski DZ. Localization of 1 l~-hydroxysieroid deliydrogenase: specrfic protector of’ the minerulocorriroid receptor in niumnialiun oractory mucosa. Acta Otolaryngol (Stockh) 1997; I 17: 738-743.

The enzyme I Ip-hydroxysteroid dehydrogenase ( I IPHSD) plays a major role in the protection o f the mineralocorticoid (type 1) receptor. The cellular mechanism of aldosterone selectivity relies on the coexpression of mineralocorticoid receptors and 1 IBHSD in the same cells. In the current study, 1 IPHSD activity was localized in the mammalian olfactory mucosa by a histochemical technique which links steroid metabolism with the deposition of formazan. The histochemical reaction results from oxidation of the synthetic substrate I la-hydroxyandrostenedione and is dependent on nicotine- adenine dinucleotide (NAD). This demonstrates the presence of a dehydrogenase activity separate from the nicotineamide- adenine dinucleotide phosphate (NADP)-dependent 1 IBHSD. In the olfactory rnucosa, the presence of NAD-dependent l lPHSD activity is localized to the sustentacular cells and acinar cells of Bowman’s glands. N o definite NAD-dependent activity was demonstrated in the olfactory receptor neurons. These data suggest that mineralocorticoid receptors present in acinar cells and sustentacular cells are aldosterone selective. Key words: Boiiwwn’s glund, Na/K-ATPase, ion trunsport, niineralocortfcnids, giucocorticoids.

INTRODUCTION The classic mechanism of corticosteroid hormone action involves diffusion across the cell membrane and binding to intracellular cytoplasmic receptors present in target cells. The resulting steroid-receptor complex enters the nucleus, where it interacts with chromatin and alters the transcription rate of certain genes (1). At least two types of corticosteroid receptors exist: type 1 and type 11. Type I receptors are referred to as mineralocorticoid (aldosterone) receptors and type I1 as glucocorticoid (cortisol, corticosterone), based on their in uivo binding selectivity and physiological re- sponse in epithelial target tissues such as kidney, colon and parotid gland (2, 3). In contrast, in uitro binding studies using cloned receptors have demonstrated near-identical affinity of cortisol and aldosterone for type I receptors (4). Further studies indicated that aldosterone target tissues contain not only type I receptors but also specific steroid-inactivating en- zymes which play a key role in the determination of target cell sensitivity by resolving cross-reactivities between the mineralocorticoid, aldosterone, and the glucocorticoids, cortisol and corticosterone (2, 3). The enzyme 1 Ip -hydroxysteroid dehydrogenase (1 1,BHSD) in mineralocorticoid target cells is respon- sible for converting cortisol and corticosterone to cortisone and dehydrocorticosterone, respectively. The latter two steroids possess a very weak affinity for the mineralocorticoid receptor. Thus, the presence of this intracellular enzyme prevents the binding of glu- cocorticoids to type I receptors even in the presence of a high circulating level of glucocorticoids.

Recent biochemical and immunocytochemical stud- ies have demonstrated the presence of both type I and type I1 corticosteroid receptors in the olfactory mucosa (5-7). The role of specific corticosteroid- inactivating enzymes within the olfactory mucosa re- quires investigation to ascertain the specificity of hormone action on olfactory target cells. The aim of this study was, therefore, to test the hypothesis that specific target cells in the olfactory mucosa pos- sess the specific glucocorticoid-inactivating enzyme, 11pHSD.

MATERIAL AND METHODS Preparation of tissues The olfactory mucosae of 18 male Hartley guinea- pigs (Charles River Breeding Laboratories, Kalama- zoo, MI, USA), weighing 200-250 g, were rapidly dissected and fixed by immersion in 2%) paraformaldehyde buffered with 0.1 m cacodylate, pH 7.4, for 15 niin at 4°C. This fixation protocol was found to provide adequate morphological preserva- tion while minimizing inhibition of enzymatic activ- ity. After fixation, tissues were rinsed in several changes of 0.1 m cacodylate buffer, pH 7.4, for 3 h at 4°C and then incubated in reaction or control media for the histochemical detection of 1 IPHSD. Olfac- tory tissues from nine guinea-pigs were incubated to demonstrate 1 l,B HSD activity. Nine guinea-pigs were utilized for the completion of the control experi- ments.

6 1997 Scandinavian University Press. ISSN 0001-6489

Act

a O

tola

ryng

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y N

orth

east

ern

Uni

vers

ity o

n 11

/25/

14Fo

r pe

rson

al u

se o

nly.

Acta Otolaryngol (Stockh) I17 IlPHSD in the olfactory mucosa 739

Rraction media for the demonstration of 1 18 -1zydroxysteroid dehydrogenasc Since the I lPHSD reaction medium described previ- ously (8- 10) was reported to give consistent results for localization of this enzyme, it was used without any major modifications. Microdissected, whole (intact) olfactory mucosae were incubated for 1 h at 37°C in 0.05 mM phosphate buffer, pH 7.4, containing the following reagents: nicotinamide adenine dinucleotide (NAD; final concentration 0.4 mM), nitro blue tetra- zolium (NBT; final concentration 0.5 mg/ml) and 1 1 a -hydroxyandrostenedionel (1 lg-hydroxyandrost- 4-ene-3,17-dione), a steroid substrate, dissolved in dimethyl formamide (final concentration 0.5 mM). Although cortisol and cortisone are the substrates for 1 IBHSD, these glucocorticoids unfortunately cannot be utilized for cytochemistry owing to the presence of two additional hydroxyl groups, whose contribution to the cytochemical reaction cannot be excluded (1 1). Therefore, the substrate used in this investigation was 1 la-hydroxyandrostenedione (7, 8, 10). During the dehydrogenation reaction technique NAD acts as the hydrogen-transferring agent between the hydroxys- teroid and the final hydrogen acceptor, NBT (1 1). Incubation of tissues was carried out in the dark to prevent photo-oxidation of flavoprotein substitutes. After incubation, tissues were thoroughly rinsed in buffer to remove the reaction medium, leaving only the reduced tetrazolium salt, formazan, bound to the tissue. Further reduction of any unreduced NBT remaining in the tissue was halted by post-incubation

fixation in 2% paraformaldehyde (12). To enhance the visibility of the reaction products, samples were post- fixed in 1% osmium tetroxide solution at 20°C for 15 min. Tissues were dehydrated and embedded, and semithin sections were examined by bright-field and phase-contrast microscopy using a Leitz Diaplan mi- croscope.

Control media As a control for the histochemical reactions, the following media were used in connection with the above reaction media: (i) substrate-free reaction me- dia; (ii) media lacking coenzyme (NAD); and (iii) media lacking tetrazolium salt. Three guinea-pigs were used for each control group, making a total of nine guinea-pigs. The care and use of the animals reported on in this study were approved by North- western University Animal Committee.

RESULTS Histochemical detection of l l a H S D with the sub- strate, 1 Ip-hydroxyandrostenedione, and the cofactor NAD elicited an extensive formazan deposition in the apical region of the olfactory epithelium using bright- field microscopy (Fig. 1). At higher magnification, reaction product was clearly apparent in the apical portion of sustentacular cells (Fig. 2). Deeper within the epithelium, at the level of the olfactory receptor nuclei, activity was also detected within cell processes. Definitive identification of the cell type(s) displaying

Fig. 1. Histochemical detection of NAD-dependent 11/3HSD in the apical portion of the olfactory epithelium of the guinea-pig. 1 1/?-hydroxyandrostenedione and the cofactor N A D elicited marked formazan deposition in the sustentacular cell region (*). Deeper regions of the epithelium corresponding to the location of basal progenitor cells were devoid of enzymatic activity. Bright-field microscopy. Bar = 100 mm.

Act

a O

tola

ryng

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y N

orth

east

ern

Uni

vers

ity o

n 11

/25/

14Fo

r pe

rson

al u

se o

nly.

140 R. C. Kern & D. Z. Pitovski Acta Otolaryngol (Stockh) I17

F g . 2. At higher magnification, a reaction product indicative of NAD-dependent llPHSD enzyme activity was clearly localized to the apical portion of sustentacular cells (*). Activity was also apparent in cell processes projecting toward the basal surface of the epithelium (u r rmdwd$) . The diameter of these processes suggests that this reflects 1 IDHSD activity u ithin sustentacular cells. Bright-field microscopy. Bar = 50 mm.

F/,q. 3. Examination of the lamina propria revealed NAD-dependent 1 IPHSD enzyme activity in the Bowman’s gland!.. A granular precipitate indicative of enzymatic activity was apparent in the acinar cells (A). The reaction product was absent fi-on1 the remaining portions of the lamina propria. Phase-contrast microscopy. Bar = 20 mm.

Act

a O

tola

ryng

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y N

orth

east

ern

Uni

vers

ity o

n 11

/25/

14Fo

r pe

rson

al u

se o

nly.

Acta Otolaryngol (Stockh) 117 11PHSD in the olfactory mucosa 741

this activity is not possible in this preparation. The fixative used in this experiment was necessarily weak in order to preserve 1 lpHSD activity within the tissue. This resulted in only moderate preservation of the tissue and the subsequent inability to distinguish between processes of olfactory receptor neurons (ax- ons) and basal processes of sustentacular cells. The diameter of the processes (0.2 mm), however, was greater than that normally seen in an olfactory recep- tor axon, although the possibility of fixation artifacts cannot be eliminated (13). The deepest areas of the epithelium, which contain basal progenitor cells, failed to demonstrate enzyme activity. No reaction product was detected in samples incubated in control media (not shown).

The lamina propria of the olfactory mucosa con- tains primarily olfactory receptor nerve axon bundles, Bowman’s glands and blood vessels. When this tissue was incubated to demonstrate 1 lPHSD activity only the Bowman’s glands exhibited significant reaction product. Observation by phase-contrast microscopy revealed the acinar cells of Bowman’s glands to con- tain dense cytoplasmic particles, indicating 1 1 PHSD activity (Fig. 3). Duct cells of Bowman’s glands also exhibit l lPHSD activity (data not shown). No sig- nificant additional activity was demonstrated in the lamina propria. No reaction product was detected in samples incubated in control media (not shown).

DISCUSSION Enzymatic activity for 1 IPHSD was identified in the mammalian olfactory mucosa by microscopic histo- chemistry. In the olfactory epithelium, the activity was specifically associated with the cytoplasm of the sustentacular cells. Activity was also localized within cellular processes projecting toward the basal surface of the epithelium. Definitive identification of these cells is unclear as both sustentacular cells and olfac- tory receptor neurons possess processes which project to the basai surface of the epithelium. The diameter of the basal projections which demonstrated activity, however, exceeded the typical width of mammalian olfactory receptor axons (13). The fixative used in this experiment needed to be weak in order to pre- serve 1 IPHSD activity within the tissue. This resulted in only moderate preservation of the tissue, and post-fixation swelling could account for this increase in diameter. Olfactory nerve axon bundles in the lamina propria, however, failed to demonstrate any enzymatic activity, which further suggests that activ- ity in these basal projections reflects sustentacular cell 1 IBHSD activity. Within the lamina propria, acinar and ductal cells of Bowman’s glands demonstrated 1 IPHSD activity.

Intracellular enzymes which interconvert circulat- ing hormones between active and inactive forms aid in regulating the biological activity of the ligand in a cell-specific manner. This is particularly important in mineralocorticoid target tissues where glucocorticoids and mineralocorticoids have equivalent affinity for the mineralocorticoid receptor (1 -4). Inactivation of glucocorticoids at the 1 I-hydroxyl position by the action of l lPHSD permits the occupation of the mineralocorticoid receptor by aldosterone despite the presence of much higher levels of circulating cortisol (2). In patients with a deficiency of 1 IpHSD activity, glucocorticoids will activate the mineralocorticoid re- ceptor, leading to excessive mineralocorticoid type effects such as sodium retention and potassium excre- tion (14). The effects of 1 IBHSD deficiency on olfac- tory function are currently unknown.

There is abundant evidence in the literature for the presence of multiple isoenzymes with variable degrees of 1lpHSD activity in different tissues (15). A low- affinity NADP-dependent enzyme termed 1 1P HSDl is found in a wide variety of epithelial tissues such as proximal kidney tubule, brain, liver and inner ear (16). Immunohistochemical studies, however, have shown that this NADP-dependent enzyme does not colocalize with the mineralocorticoid receptor in cells with known aldosterone sensitivity (17). In classic mineralocorticoid target tissues (such as the distal tubule of the kidney) a high-affinity NAD-dependent enzyme is present (1 lPHSD2) and this isoenzyme has been shown to serve as the specific protector of the mineralocorticoid receptor (1 5, 18).

The majority of organs in mammals have measur- able 1 lpHSDl activity, while classic mineralocorti- coid target tissues exhibit 11PHSD2 activity (18). The current study demonstrates NAD-dependent 1 lPHSD activity within the olfactory mucosa, indi- cating the presence of the 11pHSD2 isoenzyme. Ear- lier immunocytochemical work has demonstrated the presence of type I (mineralocorticoid) receptors within the sustentacular cells and the acinar cells of Bowman’s glands (5). Taken together, these data support the hypothesis that the olfactory mucosa is a target site of mineralocorticoid (aldosterone) action.

The classic effect of aldosterone on epithelial target tissues is one of sodium retention and potassium secretion mediated, at least in part, through increases in Na/K-ATPase activity (19). This pattern is seen in ion-transporting tissues such as kidney, colon and salivary gland (19). Recent histochemical and im- munocytochemical work has demonstrated the pres- ence of high levels of the enzyme Na/K-ATPase in the sustentacular cells and Bowman’s glands (20, 21). Although the physiological effects of mineralocor- ticoids within the olfactory system are currently un-

Act

a O

tola

ryng

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y N

orth

east

ern

Uni

vers

ity o

n 11

/25/

14Fo

r pe

rson

al u

se o

nly.

142 R. C. Kern d D. Z. Pirooski Acta Otolaryngol (Stockh) 1 1 7

known, the colocalization of 1 1 ,!I HSD2, Na/K- ATPase and the type I receptor suggest the possibility that aldosterone may act, through Na/K-ATPase, to increase net sodium resorption and potassium secre- tion within the olfactory mucosa.

Mammalian olfactory mucus is primarily secreted by Bowman’s glands and may subsequently be modified by sustentacular cell activity (22). This mu- cus bathes the apical surface of olfactory neurons and serves as the source of ions for the initial phases of olfactory transduction (23). The current study demonstrates the presence of 1 lpHSD2 activity within the cells responsible for olfactory mucus secre- tion and modification, suggesting the possibility that aldosterone may modulate these processes. 1 lPHSD2 will act to protect type I receptors present in these cells from circulating glucocorticoids and thereby per- mit aldosterone binding to take place. Local mediator and neuronal regulation of olfactory mucus secretion is well established, but systemic (hormonal) regula- tion of olfactory mucus secretion is currently un- known. In other ion-transporting epithelial tissues such as kidney and salivary gland aldosterone acts to increase sodium resorption through increases in Na/ K-ATPase activity. This results in the generation of relatively hypotonic saliva and urine (1, 19). By anal- ogy with other systems, the current data suggest that aldosterone may act to regulate the sodium concen- tration of olfactory mucus with the potential for profound effects on olfactory transduction. The sig- nificance of this hypothesis is underscored by the well-established importance of sodium on olfactory transduction (24). Furthermore, 1 lpHSD2 will act to protect type I receptors from glucocorticoids despite wide variations in plasma glucocorticoid (cortisol) levels. Serum cortisol levels are known to exhibit a circadian rhythm and increases also occur in response to stress and infection. The l lpHSD2 enzyme would act to metabolize intracellular cortisol and, therefore, moderate fluctuations in cortisol concentration at the cellular level. Deficiency of 11pHSD2 in the olfactory niucosa might result in cortisol-mediated variations in the electrolyte concentrations of olfactory mucus and perineural fluid. The resulting electrolyte abnor- malities might have secondary (indirect) toxic effects on olfactory neurons. This is particularly interesting given the well-established direct neurotoxicity of ele- vated glucocorticoids in uivo. The molecular mecha- nism for this neurotoxicity is unknown but may involve a glucocorticoid-mediated decrease in cy- toskeletal proteins within the neuron (25). This mech- anism of direct glucocorticoid neurotoxicity would be in addition to the putative indirect neurotoxic effects (such as electrolyte changes) mentioned above. The 1 IPHSD2 enzyme has not been demonstrated within

olfactory neurons in the current study and a,ill not moderate any direct neurotoxic effects of glucocorti- coids in these cells. It is possible, however, that other isoforms of l lPHSD may be present in olfhctory receptor neurons and function in this capacity. The integrated effects of corticosteroids and their degra- dative enzymes on olfactory sodium metabolism, se- cretion and transduction are currently under study.

ACKNOWLEDGEMENTS This investigation was supported by NIH CIDA Award D C 00046 to D.Z.P., NIH CIDA Award D C 001 15 to R.C.K. and the Northwestern University Research Fund.

REFERENCES 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

Wilson, JD. Endocrinology and metabolism. In Issel- bacher KJ, Braunwald E, Wilson JD, Martin JB, Fauci AS, Kaspar DL, eds. Harrison’s principles of internal medicine, 13th ed. New York: McGraw-Hill, 1994:

Funder JW, Pearce PT, Smith R, Smith AI. Mineralo- corticoid action: target tissue specificity is enzyme, not receptor, mediated. Science 1988; 242: 583- 5. Edwards CRW, Stewart PM, Burt D, et al. Localiza- tion of 1 ID-hydroxysteroid dehydrogcnase-tissue spe- cific protector of the mineralocorticoid receptor. Lancet 1988; ii: 986-8. Arriza JL, Weinberger C, Cerelli G, et al. Cloning of human mineralocorticoid receptor complementary DNA: structural and functional kinship with the gluco- corticoid receptor. Science 1987; 237: 268-75. Foster JD, Kern RC, Pitovski DZ. Immunohistochemi- cal localization of mineralocorticoid (type I ) receptors in the mammalian olfactory mucosa. Assoc Res Orolar- yngol Abstr 1994: 33. Foster JD, Kern RC, Pitovski DZ. Immunohistochemi- cal localization of mineralocorticoid (type 11) receptors in the mammalian olfactory mucosa. SOC Neiirosci Abstr 1995: 650. Kern RC, Salem HH, Foster JD, Pitovski, DZ. Miner- alocorticoid (type I) receptors in the olfactory mucosa. Assoc Res Otolaryngol Abstr 1995: (18) 649. Baillie AH, Ferguson, MM, Calman KC, Hart DMcK. Histochemical demonstration of 1 11-hydroxysteroid dehydrogenase. J Endocrinol 1965; 33: 119-25. Baillie AH, Calman KC, Ferguson MM, Hart DMcK. Histochemical utilization of 3-, 6/?-, 1 I - , 12-, 16-, 16p-. 17-, 21- and 24-hydroxysteroids. J Endocrinol 1966; 34: 1-12. Pitovski DZ. Histochemical demonstration of 1 I l -hy- droxysteroid dehydrogenase in ampullae of semicircular canals. Acta Otolaryngol, (Stockh) 1996; 116: 116-9. Wattenberg LW. Microscopic histochemical dernon- stration of steroid-3-01 dehydrogenase in tissue sec- tions. J Histochem Cytochem 1958; 6: 225-32. Balogh K. A histochemical method for the demonstra- tion of 20-hydroxysteroid dehydrogenase activity i n rat ovaries. J Histochem Cytochem 1964: 12: 670-3. Farbman AI. Cell biology of olfaction. New York: Cambridge University Press, 1992: 37.

1883-8.

Act

a O

tola

ryng

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y N

orth

east

ern

Uni

vers

ity o

n 11

/25/

14Fo

r pe

rson

al u

se o

nly.

Acta Otolaryngol (Stockh) 117

14. Stewart PM, Corrie JET, Sheckleton CL, Edwards CRW. Syndrome of apparent mineralocorticoid excess. J Clin Invest 1987; 82: 340-9.

15. Krozowski Z.S, Provencher PH, Smith RE, Obeyesekere VR, Mercer WR, Albiston AL. Isozymes of 1 1 beta-hy- droxysteroid dehydrogenase: which enzyme endows mineralocorticoid specificity? Steroids 1994; 59: 1 16-20.

16. ten Cate W J F , Monder C, Marandici A, Rarey KE. I I-Hydroxysteroid dehydrogenase in the rat inner ear. Am J Physiol 1994; 266: E269-73.

17. Rundle SE, Funder JW, Lakshmi V, Monder C. The intrarenal localization of mineralocorticoid receptors and 1 1 beta-dehydrogenase: immunocytochemical stud- ies. Endocrinology 1989; 125: 1700-4.

18. Albiston AL, Obeyesekere VR, Smith RE, Krozowski ZS. Cloning and distribution of the human Ilp-hy- droxysteroid dehydrogenase type 2 enzyme. Mol Cell Endo 1994; 105: R l l - 7 .

19. Funder JW. Aldosterone action. Ann Rev Physiol

20. Kern RC, Kerr TP, Getchell TV. Ultrastructural local- ization of Na, K-ATPase in rodent olfactory epithe- lium. Brain Res 1991; 546.

2 I . Foster JD, Kern RC, Pitovski DZ. Localization of Na, K-ATPase in the mammalian olfactory mucosa. Assoc Res Otolaryngol Abstr 1995; 650: 163.

22. Getchell M , Mellert, T. Olfactory mucus secretion. In: Getchell TV, ed. Taste and smell in health and disease. New York: Raven Press, 1991.

1993; 55: 115-30.

1 lPHSD in the olfactory nzucosa 743

23. Getchell T, Su Z, Getchell M. Mucus domains: micro- chemical hetetogeneity in the mucociliary complex of the olfactory epithelium. In: the molecular basis of smell and taste transduction. Ciba Symposium, 179. New York: Wiley, 1993.

24. Firestein S, Werblin, F. Odor-induced membrane cur- rents in vertebrate receptor neurons. Science 1989; 244: 79-82.

25. McEwen B, Cameron H, Chao H, et al. Adrenal steroids and plasticity of hippocampal neurons: toward an understanding of underlying cellular and molecu- lar mechanisms. Cell Molec Neurobiol 1993; 13: 457- 82.

Submitted Septeinber 3, 1996; accepted November 7, 1996

Address for correspondence: Dimitri Z. Pitovski, M.D. Laboratory for Molecular Research in Otolaryngology Department of Otolaryngology-Head and Neck Surgery Northwestern University Medical School Searle Building 12-56 1 303 East Chicago Avenue Chicago, IL 60611 USA Tel: + 1 312 503-8920 Fax: + I 312 503-1616 E-mail: DZ-Pitovski@nwu.edu

Act

a O

tola

ryng

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y N

orth

east

ern

Uni

vers

ity o

n 11

/25/

14Fo

r pe

rson

al u

se o

nly.