Regulation of renal 25(OH)D3 la-hydroxylase: signal transduction pathways

JOELLEN WELSH, VALERIE WEAVER, AND MAURA SIMBOLI-CAMPBELL Department of Biochemistry, University of Ottawa, 451 Smyth Road, Ottawa, Ont., Canada K1H 8M5

Received April 22, 1991

Introduction A steroid hormone that functions in calcium (Ca) and

phosphorus (P) homeostasis and modulates cell proliferation and differentiation is 1,25-dihydroxyvitamin D3 (1 ,25(OH)2D3). 1 ,25(OH)2D3 exerts its effects in the classic steroid hormone mode of action: after interaction with the intracellular vitamin D receptor, the hormone-receptor com- plex binds DNA, alters transcription, and ultimately modulates protein synthesis. 1 ,25(OH)2D3 modulated pro- teins are involved in Ca, P, and bone metabolism (calbindins D9K, D28K, osteocalcin), cell signalling (protein kinase C), and cell proliferation and differentiation (c-myc, c-fos). Although extra-renal production of 1 ,25(OH),D3 has been described, the proximal renal tubules are the primary site of 1 ,25(OH)2D3 synthesis in healthy individuals. This com- mentary will focus on the intracellular signalling pathways involved in regulation of the renal 25(OH)D3 la-hydroxy- lase (1-OHASE) which converts 25(OH)D3 to 1 ,25(OH)2D3.

Overview: regulation of 25(OH)D3 la-OHASE The renal 1-OHASE is a mitochondrial mixed function

oxidase whereby reducing equivalents from NADPH are transferred by a flavoprotein (renoredoxin reductase) to a nonheme iron protein (renoredoxin). Renoredoxin subse- quently transfers electrons to a cytochrome P450 that catalyzes hydroxylation of 25(OH)D3 at the la-position. In a similar reaction, 25(OH)D3 can be hydroxylated at the 24-position, generating 24,25(OH)2D3, a metabolite that is not active in Ca or P homeostatis. It is not clear which, if any, components are shared between the 1- and 24-OHASES. Monoclonal antibodies generated against purified bovine, chick, and porcine renal mitochondrial cytochrome P450s recognize distinct proteins that can catalyze either la- or 24-hydroxylation in vitro. Consistent with the known reciprocal relationship between 1,25(OH)2D3 and 24,25(OH)2D3 synthesis, Ghazarian has suggested that post-translational cleavage of the cytochrome P450 determines whether it can function in la- or 24-hydroxylation. The regulatory implications of this model have been reviewed (J. Bone Min. Res., 5, 897-903, 1990) and will not be further discussed here owing to space limitations.

Although many hormones modulate renal 25(OH)D3 hydroxylation in vivo, only parathyroid hormone (PTH) and 1,25(OH)2D3 have been extensively studied in vitro. This review will summarize current evidence on the mechanisms involved in modulation of 1-OHASE activity by PTH and 1,25(OH)2D3. Lack of experimental evidence on other putative endocrine and autocrine modulators of 1 ,25(OH)2D3 production preclude discussion of their poten- tial interactions with PTH and 1 ,25(OH)2D3, although such interactions undoubtedly exist in vivo.

Parathyroid hormone Reconstitution experiments suggest that acute regulation

of 1-OHASE activity by PTH is achieved via reversible Printed in Canada / Imprime au Canada

phosphorylation of the ferrodoxin component, renoredoxin. Pretreatment of renal slices from low Ca fed, vitamin D defi- cient, parathyroidectomized rats with PTH results in dephosphorylation of renoredoxin and stimulation of 1-OHASE activity when renoredoxin is reconstituted with renoredoxin reductase and renal mitochondrial cytochrome P450 (Siege1 et al., J. Biol. Chem., 261, 16 998 - 17 003, 1986). Conversely, la-hydroxylation in the reconstituted sys- tem is inhibited by prior treatment of renoredoxin with mitochondrial type I1 protein kinase (Nemani et al., J. Biol. Chem. 264, 15 361 - 15 366, 1989). Although both renoredoxin and cytochrome P450 are phosphorylated by this kinase, only phosphorylated renoredoxin inhibits la- hydroxylation. These data suggest the existence of an endogenous PTH-responsive phosphatase that activates the 1-OHASE by dephosphorylating renoredoxin.

The data from reconstituted systems are not entirely con- sistent with data from whole tissue preparations. For example, although some studies support rapid stimulation of 1,25(OH)2D3 production by PTH, consistent with dephosphorylation of renoredoxin (which peaks within 5 min of PTH exposure), others indicate that 3-4 h of exposure to PTH is required to increase 1-OHASE activity in vivo and in vitro. Armbrecht et al. (Endocrinology (Baltimore), 114, 644-649, 1984) reported that PTH does not enhance 1-OHASE activity in renal slices from normal rats, although slices from low Ca fed, vitamin D deficient, parathyroidectomized rats respond to PTH. If the degree of phosphorylation of renoredoxin determines 1-OHASE activity, tissues from vitamin D and Ca replete animals (with inherently low 1-OHASE activity) should contain highly phosphorylated renoredoxin that would be more responsive to PTH-induced dephosphorylation. The inconsistencies in PTH responsiveness and time course suggest that the phosphorylation state of renoredoxin in vivo is likely modified by a variety of phosphatases and kinases. Since it is well established that PTH signal transduction involves both CAMP and phosphoinositide pathways, the role of each of these signal pathways in regulation of 1-OHASE activ- ity has been examined.

CAMP pathway PTH rapidly stimulates cAMP generation and CAMP-

dependent protein kinase activity in kidney, and 1,25(OH)2D3 production has been correlated to intracellular cAMP levels (modified in vitro by incubation with the phosphodiesterase inhibitor IBMX). Forskolin (an activator of adenylate cyclase and cAMP produc- tion) stimulates dephosphorylation of renoredoxin and 1,25(OH)2D3 production in renal slices from low Ca fed, vitamin D deficient, parathyroidectomized rats. These data suggest that the putative phosphatase that is responsible for dephosphorylation of renoredoxin is a phosphoprotein that is regulated by a CAMP-dependent kinase. However, since cAMP stimulation of 1,25(OH)2D3. production (i.e., in response to forskolin or IBMX) requlres at least 4 h and is

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y SA

VA

NN

AH

RIV

NA

TL

AB

BF

on 1

1/17

/14

For

pers

onal

use

onl

y.

COMMENTARIES / COMMENTAIRES 769

inhibited by cycloheximide, regulatory factors in addition to phosphorylation of renoredoxin are implicated. This long- term regulatory mechanism, possibly involving CAMP- response elements in target genes, analagous to that of adrenal steroid hydroxylases, cannot be fully resolved until the proteins of the 1-OHASE complex have been cloned, sequenced, and characterized. In this context, the recent cloning of the rat renal 24-hydroxylase (Ohyama et al., FEBS Lett., 278, 195, 1991) is a major step towards identi- fying a role for modulation of gene expression in the regula- tion of 25(OH)D3 hydroxylation.

In many systems, PTH stimulation of cAMP generation and 1,25(OH)2D3 production have been dissociated; i.e., high doses of PTH have no effect or inhibit (Fukase et al., Endocrinology (Baltimore), 110, 1073-1075, 1982) 1,25(OH)2D3 production, despite elevated cAMP (Henry, Endocrinology (Baltimore), 108, 733-735, 1981). Also, forskolin, like PTH, stimulates 1 ,25(OH)2D3 production in renal slices from low Ca fed, vitamin D deficient parathyroidectomized rats, but not in renal slices (Armbrecht et al. 1984, op. cit.) or tubules (Ro et al., J. Bone Min. Res. 5, 273-278, 1990) from normal rats, despite high cAMP production. The dissociation of cAMP generation from 1-OHASE activity suggests that la-hydroxylation is also modulated by CAMP-independent factors.

Phosphoinositide pathway PTH increases phospholipid hydrolysis, diaglycerol

(DAG), and inositol 1,4,5-triphosphate production and elevates intracellular Ca (CaJ in renal cells independently of CAMP. Elevations in DAG and Cq subsequently acti- vate protein kinase C (PKC), the Ca- and phospholipid- dependent protein kinase. PKC activation (by treatment with 12-0-tetradecanoyl phorbol 13-acetate (TPA)) reduces 1 ,25(OH)2D3 production in primary cultures of chick renal tubules (Henry and Luntao, Endocrinology (Baltimore), 124, 2228-2234, 1989), an effect not blocked by cyclo- heximide. We have observed that TPA also inhibits la- hydroxylation in freshly isolated rat proximal tubules and that 4a-phorbol 12,13-didecanoate (a TPA analog that does not cause PKC translocation) does not affect 1 ,25(OH)2D3 production (Weaver and Welsh, Proceedings of the 8th Workshop on Vitamin D, July 1991, Paris. In press). More recently, we measured PKC activity in proximal tubules (as phosphorylation of a PKC-specific peptide) and observed that TPA induces translocation of cytosolic PKC to mem- brane which is temporally consistent with inhibition of la-hydroxylation. Additionally, in this system, the effect of TPA on 1,25(OH)2D3 synthesis is negated after down regulation of PKC (by overnight incubation with TPA) and co-incubation of staurosporine (a PKC inhibitor) with TPA. Consistent with the reciprocity between the 1- and 24-OHASES, TPA stimulates 24,25(OH)2D3. production in mouse and rat renal tubules and in chick primary cultures (Mandla et al., Endocrinology (Baltimore), 127,2639-2647, 1990; Weaver and Welsh, op. cit.; Henry and Luntao, op. cit.).

Translocation of PKC, phosphorylation of endogenous substrates, and inhibition of 1 ,25(OH)2D3 production occur within 10 min of TPA incubation, a time course consistent with a role for PKC in acute regulation of 25(OH)D3 hydroxylation by the phosphorylation-dephosphorylation

model discussed above. PKC phosphorylation occurs on serine and threonine residues, and serine-threonine phosphorylation of ferrodoxin inhibits la-hydroxylation. Most intriguing is the demonstration (Boneh and Tenenhouse, Biochem. Cell Biol., 66, 262-272, 1988) of PKC activity and PKC-dependent phosphorylation in renal mitochondria, the location of the vitamin D OHASES. Although it is tempting to speculate that PKC directly inhibits the 1-OHASE via ferrodoxin phosphorylation, posi- tive identification of PKC's substrates in proximal tubules and their possible involvement in 25(OH)D3 hydroxylation require further study. Most importantly, the physiological relevance of PKC activation remains to be established; the crucial question is whether endogenous regulators of l a - hydroxylation mediate their effects via modulation of PKC.

Interactions between CAMP and PKC signal transduction Although it remains to be resolved how signals generated

at the plasma membrane or in the cytosol are transduced to the mitochondria where 25(OH)D3 hydroxylation occurs, the available data suggest that cAMP and PKC inversely regulate 1 ,25(OH)2D3 synthesis. Since both pathways are activated by PTH, we propose that cross-talk between the two pathways explains the biphasic dose response to PTH as well as the inconsistencies in the litera- ture with respect to PTH responsiveness. Preliminary results from our laboratory suggest that short-term treatment of rat renal tubules with TPA prevents PTH stimulation of 1,25(OH)2D3 production. Conversely, Henry and Luntao (op. cit.) reported that high levels of cAMP (achieved by incubation with forskolin or dibutyl CAMP) attenuate the effects of TPA on 1 ,25(OH)2D3 production. Further stud- ies on CAMP-PKC interactions, as well as investigations into other kinases and (or) phosphatases that may be activated by PI hydrolysis, are necessary to clarify the relative impor- tance of these two signal pathways in PTH regulation of la-hydroxylation. In vivo, other endocrine and autocrine factors likely impinge on both cAMP and PKC signal transduction pathways and modulate basal, as well as PTH- stimulated, 1 -0HASE activity. 1 ,25(OH)2D3 and insulin are discussed below as examples of factors that may modulate la-hydroxylation by interacting with these signal transduction pathways.

Inhibition of la-hydroxylation by 1,25(OH)$3 1 ,25(OH)2D3 represses 1 -0HASE activity and its pres-

ence in vitamin D replete, but not vitamin D deficient, animals appears to modulate PTH stimulation of l a - hydroxylation. Since 1 ,25(OH)2D3 repression of la-hydroxylation in vitro is inhibitable by cycloheximide, 1 ,25(OH)2D3 likely regulates synthesis of one or more pro- teins that affect 1-OHASE activity. In view of the inhibitory effect of TPA on la-hydroxylation, and the recent demon- stration of transcriptional regulation of PKC by 1,25(OH),D3 in HL-60 cells (Obeid et al., J. Biol. Chem. 265, 2370-2374, 1990), it is possible that 1,25(OH)2D3 suppression of la-hydroxylation involves PKC. We have observed that 1 ,25(0H)2D3 (lo-' M) increases PKC activ- ity (measured by phosphorylation assays) and amount (mea- sured by immunoblotting) in an established bovine renal cell line (MDBK cells) in a time course consistent with modula- tion of gene expression. Henry and Luntao (op. cit.) have shown that inhibition of 1-OHASE by 1,25(OH)2D3 and

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y SA

VA

NN

AH

RIV

NA

TL

AB

BF

on 1

1/17

/14

For

pers

onal

use

onl

y.

770 BIOCHEM. CELL BIOL. VOL. 69, 1991

TPA are additive. Armbrecht et al. (op. cit.) reported that preincubation of renal slices with 1 ,25(OH)2D3 prevents stimulation of la-hydroxylation by forskolin, suggesting that 1 ,25(OH)2D3 blocks CAMP-mediated enhancement of la-hydroxylation. Taken together, these results suggest that 1 ,25(OH)2D3 enhances PKC gene expression and (or) mem- brane localization, thereby increasing the amount of PKC available for activation. The elevated PKC could theoretically affect the phosphorylation state of renoredoxin and reduce basal and (or) PTH-stimulated 1-OHASE activ- ity. Since phosphorylation of the vitamin D receptor upon hormone binding (Pike and Sleator, Biochem. Biophys. Res. Commun., 131, 378, 1985; Brown and DeLuca, J. Biol. Chem., 265, 10 025, 1990) may modulate its biological activ- ity, the possibility that PKC-dependent phosphorylation modulates target cell sensitivity to 1,25(OH),D3 should also be explored.

Impaired la-hydroxylation in diabetes Insulin exerts a permissive effect on PTH stimulation of

1-OHASE activity in vitro (Henry, op. cit.) and 1 ,25(OH)2D3 production is reduced in both streptozotocin- induced and spontaneously diabetic rats (Weaver and Welsh, Fed. Proc., 49, A1047, 1990). Since circulating PTH and PTH stimulation of renal adenylate cyclase are not reduced in diabetic rats, we are investigating the possibility that impaired la-hydroxylation in diabetes is related to altered PKC activity. As discussed above, PKC activation inhibits 1,25(OH)2D3 production and thus mimics the diabetic situation. Incubation of diabetic renal slices with PTH does not result in dephosphorylation of renoredoxin or stimula-

1-OHASE turned off in diabetic rats even in the presence of PTH. Further studies are necessary to determine exactly how insulin, which operates via a tyrosine kinase receptor, modulates renal PKC and 1-OHASE activity. Altered renal PKC has also been reported in the genetically hypophosphatemic mouse and in uninephrectomy; both con- ditions are associated with altered 1 ,25(OH)2D3 produc- tion. These animal models should prove useful in probing the physiological relevance of PKC in regulation of 1 ,25(OH)2D3 production.

Summary In vitro, activation of the CAMP signalling pathway

stimulates, whereas activation of PKC inhibits, 1,25(OH)2D3 synthesis. Since PTH activates both pathways, the ultimate effect of PTH on la-hydroxylation in vivo likely depends on other endocrine-autocrine factors that impinge onto these signal transduction cascades. For example, 1 ,25(OH)2D3, a known repressor of la-hydrox- ylation, may increase renal PKC amount-activity, thereby enhancing the inhibitory arm and preventing PTH stimula- tion of the 1-OHASE. In contrast, studies with diabetic rats suggest that insulin may allow CAMP-mediated stimulation to override PKC-mediated inhibition of 1-OHASE activity. Analogous to models proposed for regulation of adrenal steroid hydroxylases, it is likely that regulation of renal vitamin D hydroxylation involves both acute (reversible phosphorylation) and chronic (modulation of gene expres- sion) mechanisms. However, the molecular details of these regulatory mechanisms remain to be resolved.

- - tion of 1-OHASE, although both responses occur in con- Acknowledements - - - - - -- - - trol tissue (Wongsurawat a/. , J. Bone Min. Res. 4 , 9 4 5 , This research was supported by operating grants to 1989). We have recently found that membrane PKC activ- J. Welsh from The Medical Research Council and The ity in renal tubules increases within 2 days of diabetes onset Kidney Foundation of Canada, and a Medical Research and is temporally correlated with the reduction in circulating Council post graduate studentship to V. Weaver. 1 ,25(OH)2D3. We hypothesize that increased PKC prevents dephosphorylation of renoredoxin and, thus, keeps the

Bio

chem

. Cel

l Bio

l. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y SA

VA

NN

AH

RIV

NA

TL

AB

BF

on 1

1/17

/14

For

pers

onal

use

onl

y.