Mycotrienins

doi:10.1074/jbc.270.43.25949

HOME

HELP

FEEDBACK

SUBSCRIPTIONS

Volume 270, Number 43, Issue of October 27, 1995 pp. 25949-25955
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.

Mycotrienins
A NEW CLASS OF POTENT INHIBITORS OF OSTEOCLASTIC BONE RESORPTION (*)

(Received for publication, July 20, 1995)

Dominik Feuerbach , Rudolf Waelchli , Theo Fehr (1), Jean H. M. Feyen (§)

From the Departments of Bone & Joints and DAT Lead-finding Unit, Sandoz Pharma Limited, 4002 Basel, Switzerland

ABSTRACT

INTRODUCTION

EXPERIMENTAL PROCEDURES

RESULTS

DISCUSSION

FOOTNOTES

ACKNOWLEDGEMENTS

REFERENCES

ABSTRACT

Pharmacological intervention using selective tyrosine kinase inhibitors has been shown to be an effective approach to inhibit osteoclast function. Here, we report on the structure-activity relations of benzoquinone ansamycins isolated from Streptomyces rishirensis, which form a new class of potent inhibitors of osteoclast-mediated bone resorption. Parathyroid hormone-stimulated bone resorption was inhibited concentration dependently by both mycotrienin I and mycotrienin II, showing half-maximal inhibition in the low nanomolar range in fetal rat long bones in vitro. Structure-activity relation studies indicate that position 19 contained within the quinone/hydroquinone element and the double bonds in position 4, 6, and 8 are crucial for full bioactivity. In contrast, substitutions in position 22 are well tolerated. The lack of a similar effect of 2,6-dimethyl-p-benzoquinone and vitamin K signifies that the mechanism of action is not solely due to the oxygen scavenger capacity of the quinone/hydroquinone moiety. The inhibition of osteoclastic bone resorption is in line with the diminished activity of immunopurified pp60 from bone suggesting that pp60 is a possible target of mycotrienins in the organ culture. Thus, mycotrienins may be useful as pharmacologic inhibitors of osteoclastic bone resorption.

INTRODUCTION

Osteopetrosis, often referred to as ``marble bone disease,'' is a sclerosing bone dysplasia mainly characterized by impaired bone resorption(1) . Over the last few years more insight was gained on the genetic defects involved in the onset of the disease pointing to several different disease mechanisms. The op mouse, a well characterized animal model of the human disease, is unable to develop osteoclasts (2) due to a point mutation in the coding region of macrophage colony stimulation factor(3) . In another animal model, osteopetrosis is manifested after homologous recombination following targeted disruption of the src protooncogene(4) . Detailed examination revealed that osteoclasts express high levels of pp60comparable to the levels expressed in brain and platelets(5) . In the pp60 minus model, multinucleated cells are formed on bone surfaces but neither ruffled border formation nor bone resorption occurs(6) . Thus, in contrast to the op mouse, the inherent defect in the src knock-out mouse is in the mature osteoclast and is autonomous of the bone microenvironment(7) .

pp60 is a non-receptor protein tyrosine kinase(8) , which is expressed ubiquitously but with elevated levels in neurons (9) and platelets(10) . Its function seems to be the transduction of signals arising from the stimulation of cells by growth factors(8, 9, 10, 11) . The pp60-deficient mouse model suggests that src kinase is essential for osteoclast function, but not for osteoclast formation. Pharmacological intervention using relatively selective pp60 tyrosine kinase inhibitors, e.g. herbimycin A, have been shown to be effective inhibitors of osteoclastic bone resorption in vitro and are able to block hypercalcemia in vivo(12, 13) . Mycotrienins, belonging to the class of benzoquinone ansamycins, are structurally related to herbimycin A in that they also contain a quinone/hydroquinone system. Starting from natural-derived mycotrienin I and II, isolated from Streptomyces rishirensis, we prepared a series of benzoquinone ansamycins derivatives and determined their activity to inhibit bone resorption in fetal rat long bones (radii and ulnae) in vitro. We further analyzed their inhibitory effect on an immunopurified pp60 preparation from bone. The observed structure activity relationship allowed for conclusions about a possible mechanism of action for inhibition of bone resorption by this class of compounds.

EXPERIMENTAL PROCEDURES

Materials

The materials used in this study and their sources were as follows: timed pregnant Sprague-Dawley rats, Life Technologies, Inc.; [45Calcium] (specific activity = 0.37-1.5 GBq/mg), [6-

H]thymidine (specific activity = 1.04 TBq/mmol), [2,3-

H]proline (specific activity = 1.51 TBq/mmol); [

P]ATP, >185 TBq/mmol), Hybond(TM) nitrocellulose membrane, ECL(TM), Amersham; BGJ-medium, Life Technologies, Inc.; bovine serum albumin type 5; enolase, Sigma; Optifluor liquid scintillation fluid, Canberra Packard; X-Omat-AR film, Kodak Inc; sheep antibody to c-src oncogenic protein OA-11-863, Cambridge Research Biochemicals Limited, Cheshire United Kingdom.

Preparation, Isolation, and Purification of Mycotrienins from S. rishirensis

Mycotrienin I (SDZ 115-961) and mycotrienin II (SDZ 115-962) were obtained from S. rishirensis T-23 as described(14) . Isolation and purification was performed according to the method described by Sugita et al.(15) . In addition to mycotrienins I and II, a novel compound was isolated. This compound differs from mycotrienin II by having a benzyl group in place of the cyclohexyl ring in the side chain of mycotrienin II and was therefore named hexadehydro-mycotrienin II (SDZ 220-542) (see Fig. 1).

Figure 1: Structure of mycotrienin analogues: A1, representing the basic structural element of mycotrienin I; A2, basic structural elements of mycotrienin II. Mycotrienin I, mycotrienin II, and hexadehydromycotrienin II have been prepared, isolated, and purified from S. rishirensis. Details on the derivatization processes are provided under ``Experimental Procedures.''

Chemical Derivatization of Mycotrienin II

Three methylether products of mycotrienin II (SDZ 221-035, SDZ 221-108, and SDZ 221-150) were formed after treatment of the parent compound with diazomethane. With the same procedure compound SDZ 221-635 was prepared starting from hexadehydro-mycotrienin II. Acylation of mycotrienin II either by benzoylchloride or acetylchloride resulted in the two benzoylmycotrienin II derivatives SDZ 225-420 and SDZ 225-421 or the two corresponding acetylmycotrienin II derivatives (SDZ 225-554 and SDZ 225-555). In a similar way the mycotrienin II-22-trifluoromethanesulfonyl derivative (SDZ 225-419) was prepared using trifluoromethanesulfonylchloride as the acylating reagent. Exhaustive hydrogenation of mycotrienin II in the presence of palladium on activated charcoal in ethanol at room temperature and normal pressure for 72 h resulted in a pair of diastereomers SDZ 225-019 and SDZ 225-020 and a diastrereomeric mixture in the ratio 6:4 of SDZ 225-021. Treatment of mycotrienin II by lithiumaluminum hydride as described (16) followed by 2,2-dimethoxypropane gives the dimethylketalmycotrienin II derivative SDZ 224-957. All newly prepared mycotrienin II analogues were purified by column chromatography on silica gel and were characterized by proton-nuclear-magnetic-resonance (

H NMR) and mass spectrum analysis (MS). Their structures are shown in Fig. 1.

Bone Resorption Assays

Fetal rat long bones were prepared and cultured as described by Raisz(17) . In brief, timed pregnant Sprague-Dawley rats were injected with radiolabeled

Calcium subcutaneous (100 µCi) on the 18th day of gestation. The following day, radii and ulnae were dissected and then precultured in 0.5 ml of BGJ-medium supplemented with 1 mg/ml of bovine serum albumin in 24-well tissue culture plates in a CO (2)

incubator at 37 °C for 24 h. The bone explants were then cultured in the presence or absence of the agents to be tested for 2 days. The medium was removed and replaced with fresh medium supplemented with the test agents, and culture was continued for another 3 days before termination of the experiment. Aliquots of conditioned medium of day 2 and day 5 and the acid extract (trichloroacetic acid, 5 (w/v)) of the bone explants were counted for

Calcium by liquid scintillation. Bone resorption was assessed as the percentage of total

Calcium that was released into the medium.

Immunoprecipitation

Radii and ulnae were dissected from 19-day pregnant Sprague-Dawley rats as described above. After the preculture period, the bone explants were cultured in the presence or absence of the agents to be tested for additional 24 h. Bone explants (eight bone explants/experimental group) were homogenized with a glass to glass homogenizer in 1 ml of lysis buffer (137 mM NaCl, 50 mM Tris-HCl, pH 7.8, 10% glycerol, 0.5 mM sodium-orthovanadate, 2 mM EDTA, 1% Nonidet P-40, 1 mM phenylmethylsulfonyl fluoride, 1 µg/ml aprotinin, 10 µM leupeptin) on ice. The lysate was centrifuged at 10,000 times

g for 30 min and the supernatant clarified by addition of 40 µl of Protein G-Sepharose (Pharmacia). The supernatant was transferred to a new tube, and 2 µl of the pp60

-specific antibody GD11 (approximately 2.5 µg/µl, a kind gift of S. J. Parsons) and 50 µl of Protein G-Sepharose were added. This antibody neither recognizes the kinase domain nor the SH-2 domain of pp60

(18) . Following overnight incubation with gentle agitation at 4 °C, the immunoprecipitate was washed four times with wash buffer (165 mM NaCl, 50 mM Tris-HCl, pH 7.8, 10% glycerol, 0.5 sodium-orthovanadate, 2 mM EDTA, 1% Nonidet P-40) and twice with 10 mM Tris, pH 7.8.

Kinase Assay

Enolase (46.9 kDa) was denaturated with 50 mM acetic acid at 30 °C for 15 min and neutralized with 1.0 N NaOH. The washed immunoprecipitates were resuspended in 30 µl of substrate kinase buffer (10 mM MnCl (2)

, 20 mM HEPES, pH 7.4, 5 µg of enolase, 10 µCi of [

P]ATP). After 5 min at 30 °C, the kinase assay was stopped by the addition of 2 times

Laemmli buffer and reaction products separated on 10% SDS-PAGE. (

)Autoradiography was performed after the gel was dried and exposed to X-Omat-AR

film using an intensifying screen at -70 °C for 1-3 days. To quantify the extent of enolase phosphorylation and pp60

autophosphorylation, the gel was analyzed using a Molecular Dynamics PhosphorImager (Sunnyvale, Ca).

Western Blot Analysis of pp60

Bone pp60

was immunoprecipitated using the monoclonal antibody GD-11, separated on SDS-PAGE and transblotted to Hybond

nitrocellulose membrane. In preliminary titration experiments, we determined the optimal amount of antibody to quantitatively precipitate pp60

from the bone lysate (results not shown). The membrane was immunostained with a sheep antibody to pp60

(Ab code OA-11-863), washed, and incubated with a horseradish peroxidase-labeled antisheep antibody. Bound antibodies were detected by an enhanced chemiluminescence assay (ECL

, Amersham).

Cell Toxicity Tests

[H]Thymidine Incorporation Test

Neonatal mouse calvariae were prepared and cultured as described previously(19) . In brief, frontal and parietal bones were dissected from 4- to 6-day-old neonatal mouse (strain CD-1), split along the sagittal suture, and placed in culture medium (BGJ medium containing 1 mg/ml bovine serum albumin). After the preculture period, individual bones were transferred to 35-mm culture wells, each well containing 1 ml of fresh culture medium in the presence or absence of the substances to be tested. After 24 h bones were pulse labeled with 5 µCi of [

H]thymidine for the last 2 h of culture. The incorporation of [

H]thymidine into the acid-insoluble bone DNA fraction was determined. Results are expressed as counts/minute/milligram of bone dry weight and are presented as the mean ± S.E.

[H]Proline Incorporation Test

Fetal rat long bones were prepared and cultured as described above (see ``Bone Resorption Assays''). At the end of the preculture, bones were incubated in the presence of various concentrations of mycotrienin II. After 24 h the cultures were supplemented with 5 µCi of [

H]proline and continued for another 48 h. At the end of the culture period, the amount of [

H]proline in the acid-insoluble bone protein fraction was determined. Results are expressed as counts/minute/four bone explants and are presented as the mean ± S.E.

Data Analysis

The results are presented as IC

values and represent the mean value (n = 6) of two independent experiments. Concentration response curves were analyzed using SIGFIT(20) . Statistical analysis was by Student's t test.

RESULTS

Inhibition of Bone Resorption of Fetal Rat Long Bones by Mycotrienins

First, we examined the effect of mycotrienin I and mycotrienin II on the parathyroid hormone-stimulated release of

Calcium into the culture medium from fetal rat long bones (Fig. 2). Both compounds inhibited the release of

Calcium in a concentration-dependent manner with apparent half-maximal inhibition constants (IC

values) of 64 nM (mycotrienin I, code 115-961) and 21 nM (mycotrienin II, code 115-962). Similar inhibitory effects were observed with 1,25-dihydroxycholecalciferol as a bone resorption stimulus (data not shown). Herbimycin A, an antibiotic known to inhibit several tyrosine kinases including pp60

(21) inhibited the PTH-stimulated resorption in fetal rat long bones with apparent IC

values of 400 nM at day 5 (Fig. 3). Bone resorption inhibition was complete (100%) and persisted during the entire culture period.

Figure 2: Concentration dependence of the inhibitory effect of mycotrienin I and mycotrienin II on bone resorption stimulated in the presence of parathyroid hormone in 19-day-old fetal rat long bones in vitro. Fetal rat long bones were cultured with 10M hPTH-(1-34) in the presence or absence of the mycotrienin analogues over the concentration range of 10 to 10M. Bone resorption was assessed as the percentage release of the total Calcium that is released into the culture medium at day 5 of culture. Results are presented as the treated/control (T/C) ratio and represents means ± S.E. (n = 6) for six long bones/group. Results of a representative experiment (n = 3) are shown.

Figure 3: Concentration dependence of the inhibitory effect of herbimycin A on bone resorption stimulated in the presence of parathyroid hormone in 19-day-old fetal rat long bones in vitro. Fetal rat long bones were culture in the presence of 10M hPTH-(1-34) in the presence or absence of herbimycin A over the concentration range of 10 to 10M. Bone resorption was assessed as the percentage release of the total Calcium that is released into the culture medium at day 2 and day 5 of culture. Results are presented as the treated/control (T/C) ratio and represent means ± S.E. (n = 6) for six long bones/group. Results of a representative experiment (n = 2) are shown.

Structure Activity Relationship of Mycotrienin II Analogues

To determine the essential structure which enables mycotrienins to inhibit bone resorption, we introduced different residues at position 19 and/or 22 (see Fig. 1). The results of these analogues are summarized in Table 1. Methylether substitution of the hydroxyl group at position 19 (SDZ 221-150) in mycotrienin II resulted in a dramatic loss in antiresorption activity with an apparent IC

value of 2.88 times

M. However, full intrinsic activity was retained (intrinsic activity of mycotrienin II = 1 by definition). In contrast, the same substitution at position 22 (SDZ 221-035) did not significantly alter the inhibitory effect on the parathyroid hormone-stimulated bone resorption. In a series of different substitutions at position 22, the following rank order of potency concerning bone resorption was obtained: methylether > acetyl > mycotrienin II > triflat > benzoyl. Mycotrienin-II-22-monomethylether showed even a 4-fold increase in activity. All 22-substituted analogues investigated exhibit full intrinsic activity. Dimethylation or debenzoylation of the phenolic OH groups at position 19 and 22 resulted in total loss of activity. Surprisingly, acetyl residues at both positions and only at position 22 did not affect the activity and inhibited bone resorption with apparent IC

values of 1.5 times

M and 1.1 times

M, respectively.

The acylation at position 22 with benzyol chloride resulted in a 4-fold loss in activity. The natural compound SDZ 220-542, in which the cyclohexane ring of mycotrienin is replaced by a phenyl group and with a methylether at position 22, showed only little change in the activity. Hydrogenation of the double bonds in positions 4, 6, 8, and 14 with or without the removal of the hydroxyl group in position 13 (see Z (4) in Fig. 1) was carried out. The resulting analogues showed no inhibition of bone resorption. Removing the bulky side chain in position 11 as in SDZ 224-957 (structure not shown) also abolished the antibone resorptive activity.

To test for any activity of the quinone/hydroxyquinone structure in other molecules, we assayed 2,6-dimethyl-p-benzochinon and vitamin K1 for inhibition of bone resorption. None of the two compounds had any influence on the release of Calcium from long bones over the concentration range tested (see Table 2), showing that the quinone moiety by itself was not sufficient for antiresorptive activity.

Bioeffect/Toxicity Ratio Determination by [H]Proline and [H]Thymidine Incorporation

To rule out the possibility that inhibition of bone resorption is not due to nonspecific toxic effects, two potent analogues, mycotrienin-II-22-monoethylether and hexahydromycotrienin-II-22-monoethyl ether, were studied in the [

H]thymidine incorporation assay using 4-6-day-old neonatal mouse calvariae in the presence of 10

M hPTH-(1-34) as a measure for cellular toxicity (see Fig. 4). [

H]Thymidine incorporation into the acid-insoluble bone DNA fraction was not affected by these compounds at a concentration range of 10

-10

M. Significant inhibition, however, was observed at µM concentrations, which indicates that the inhibition of bone resorption in the low nanomolar range is not due to unspecific toxic effects.

Figure 4: Effect of two potent mycotrienin analogues on the parathyroid hormone-induced [H]thymidine incorporation into the acid-precipitable bone DNA fraction in 4-6-day-old neonatal mouse calvariae. After a preincubation period (18 h), calvarial halves were stimulated with 10M hPTH-(1-34) in the presence or absence of different concentrations of mycotrienin II for 24 h with [H]thymidine for the last 2 h of culture. Bones were washed in phosphate-buffered saline, extracted with trichloroacetic acid, aceton, and ether. [H]Thymidine incorporation into the acid-precipitable bone DNA fraction is expressed as counts/min/milligram bone dry weight, and is presented as the means ± S.E. for six half-calvariae. *, significantly different from PTH (p < 0.01).

In addition we have tested possible toxic effects of the potent mycotrienin 115-962 in the long bone organ cultures by [H]proline incorporation into the de novo synthesized bone proteins. This compound did not significantly impair [H]proline incorporation into the fetal rat long bones at the concentration range of 10-10M (see Fig. 5).

Figure 5: Effect of mycotrienin II on the parathyroid hormone-induced [H]proline incorporation into fetal rat long bones. After a preincubation period of 24 h, bone explants were labeled with 5 µCi of [H]proline for 48 h. At the end of the culture period, the amount of [H]proline in the acid-insoluble bone protein fraction was determined. [H]Proline incorporation into the acid-precipitable bone protein fraction is expressed as counts/min/four bone explants and is presented as the means ± S.E. (n = 4). *, significantly different from PTH (p < 0.01).

Influence of Mycotrienin II on pp60 Kinase Activity

To clarify the mechanism of action of the mycotrienins, we have examined whether this class of compounds affects the pp60

kinase activity in the fetal rat long bone explants, in analogy to the inhibitory effect of the structurally related herbimycin A on pp60

in bone marrow cultures. A 24-h incubation with hPTH-(1-34) resulted in a 2.5-fold increase of pp60

kinase activity using enolase as substrate (compare lanes 1 and 2 in Fig. 6A). Simultaneous treatment of the bone explants with PTH-(1-34) and mycotrienin II or mycotrienin I led to a concentration-dependent decrease of pp60

activity. The IC

for pp60

activity in this assay is 100 nM for mycotrienin I and 50 nM for mycotrienin II (see Fig. 6B) which is clearly higher than the IC

which is found for bone resorption inhibition in the fetal rat long bone assay. The IC

observed for herbimycin A in the kinase assay (400 nM) is comparable with the concentration which is needed in the organ culture to get half-maximal inhibition. Note that at the concentration where inhibition of substrate phosphorylation occurred, c-src autophosphorylation was affected as well (Fig. 6A). Both mycotrienin I and herbimycin A did not affect total pp60

protein levels as measured by Western blot analysis (see Fig. 7). The structure activity relations for the different mycotrienin analogues to inhibit bone resorption closely resembles the ability of these compounds to inhibit pp60

after treatment of the bone explants in the absence or presence of PTH-(1-34) (see Fig. 8).

Figure 6: Influence of mycotrienin II and herbimycin A on pp60 kinase activity in fetal rat long bones in vitro. Panel A, fetal rat long bones were cultured for 24 h with hPTH-(1-34), mycotrienin II, and herbimycin A at concentrations indicated in the figure. Total long bone lysates were immunoprecipitated with pp60 antibody GD-11 and assayed for kinase activity using enolase as a substrate in the presence of [-P]ATP. Panel B, quantitative analysis of the phosphorylation of enolase by the pp60oncoprotein. After separation on 10% SDS-PAGE, radioactivity of the band corresponding to enolase was determined using a PhosphorImager (Molecular Dynamics, model PI400S). *, significantly different from PTH (p < 0.01),**, significant different from control (p < 0.01).

Figure 7: Influence of mycotrienin II and herbimycin A on the total protein amount of pp60 in fetal rat long bones in vitro. Fetal rat long bones were cultured for 24 h with hPTH-(1-34), mycotrienin II and herbimycin A as indicated in Fig. 6. Total bone lysates were immunoprecipitated with pp60 antibody GD-11. Immunoprecipitates were subjected to SDS-PAGE and immunoblotted by a sheep antibody to the pp60oncoprotein (OA-11-863).

Figure 8: Structure activity relations for a selective series of mycotrienin analogues on pp60 kinase activity in fetal rat long bones in vitro. Fetal rat long bones were cultured for 24 h with hPTH-(1-34) (10M), in the presence or absence of the mycotrienin analogues at concentrations indicated in the figure. Total long bone lysates were immunoprecipitated with pp60 antibody GD-11 and assayed for kinase activity using enolase as a substrate in the presence of [-P]ATP. Radioactivity of the band corresponding to enolase was determined using a PhosphorImager (for details see legend to Fig. 6). *, significantly different from PTH (p < 0.01).

Next, we were interested to see whether mycotrienin I inhibits pp60 kinase directly or whether it interferes with an earlier step in the signaling cascade. For this purpose pp60 was immunoprecipitated from fetal rat long bone homogenates. To obtain optimal pp60 activity, bone explants were stimulated with 10M hPTH-(1-34) during 24 h in culture prior to homogenization. The immunocomplexed enzyme was preincubated with the different compounds for 30 min before the addition of the substrate and [-P]ATP. As shown in Fig. 9, both herbimycin A and mycotrienin I inhibited the kinase activity by 50% at a concentration of 0.5 µM.

Figure 9: Direct inhibitory effect of mycotrienin II and herbimycin A on the pp60 kinase. Fetal rat long bones were cultured with hPTH-(1-34) (10M) for 24 h and lysed. Total bone lysates were immunoprecipitated with pp60 antibody GD-11. Immunoprecipitates were treated for 30 min at room temperature with the compounds as indicated, washed, and assayed for kinase activity using enolase as a substrate in the presence of [-P]ATP. Quantification of the radioactivity within the band corresponding to enolase was carried out as described in the legend of Fig. 6B. *, significantly different from control (p < 0.01).

DISCUSSION

The results which we obtained in the fetal rat long bone resorption assay indicate that mycotrienins form a potent class of inhibitors of osteoclastic bone resorption. Both mycotrienin I and mycotrienin II inhibited the release of Calcium into the culture medium by fetal rat long bones with apparent half-maximal inhibition (IC) values of 64 and 21 nM, respectively. The inhibition was complete and persisted during the entire culture period of 5 days. In contrast, no significant change in the basal bone resorption rate was observed at the concentrations tested (data not shown). Our results are in agreement with a previous study by Yoneda et al.(12, 22) who demonstrated that the structurally related herbimycin A inhibited osteoclast formation in bone marrow cultures, diminished the function of matured osteoclast in the pit assay, and prevented hypercalcemia by interleukin-1 or tumor-induced hypercalcemia.

From an analysis of the structure relationship of the mycotrienins, it is concluded that substitution in position 19 is essential for potent inhibition of bone resorption. Methylation of the hydroxyl group in position 19 resulted in a significant loss in activity to an apparent IC value of 2.88 times 10M. Despite a 10-fold drop in activity the compound retained full intrinsic activity (intrinsic activity of mycotrienin II is 1 by definition). Further structural elements for the inhibition of bone resorption are the double bonds in position 4, 6, 8, and 14, the hydroxyl group at position 13, and the amide function at position 25. In distinction to the elements required in position 19, substitutions in position 22 are well tolerated indicating that this position is not crucial for bioactivity. Methylation even led to a small increase of the inhibitory effect of the analogue on bone resorption.

It has been hypothesized that oxygen-derived free radicals produced by the osteoclast serve as intermediates in the recruitment and activation of osteoclasts(23, 24, 25) . Depletion of superoxide anions in tissue by superoxide dismutase led to an inhibition of stimulated bone resorption, e.g. by parathyroid hormone or interleukin 1 (25) . Vitamin K, a quinone/hydroquinone derivative, abolished the induction of interleukin 1 beta by the phorbol ester phorbol myristate acetate, an effect which has been related to the radical scavenging properties of vitamin K(26) . As the quinone/hydroquinone moiety is a common element of the different mycotrienin analogues tested, we have investigated whether the inhibition of bone resorption was related to the radical scavenging capacity of these analogues. Results obtained with 2,6-dimethyl-p-benzoquinone and vitamin K do not support this notion. A further argument against the scavenger theory may be seen in the observation that structural modifications within the quinone moiety, at position 22, are well tolerated. Second, esterification of both positions 19 and 22 in the quinone ring preserved the bone antiresorptive activity. We therefore conclude that the ability of the mycotrienins to inhibit bone resorption is not explained by the oxygen radical scavenging capacity of these analogues. The bioeffect/toxicity ratios of these compounds indicates that the inhibition of bone resorption in the low nanomolar range probably is not just a consequence of their nonspecific toxic effects. Inhibition of cell proliferation and protein synthesis, markers for general toxicity, was only seen at high micromolar concentrations for two of our potent analogues.

Mycotrienins which belong to the benzoquinone ansamycin class of compounds were first isolated from different strains of Streptomyces and described to have antifungal and antitumor activity. Related compounds in this group include geldanamycin (27) and herbimycin A (28) . Both compounds have been reported to revert the morphology of fibroblasts transformed by many oncogenic tyrosine kinases e.g. src, fyn, bcr-abl, and erbB2(29) . Our findings indicate that pp60 kinase may serve as a possible target for mycotrienins in the fetal rat long bones. The structure activity relations for the different mycotrienin analogues to inhibit bone resorption closely resembles the ability of these compounds to inhibit pp60. Fetal rat long bones treated with parathyroid hormone showed a 2-fold increase in pp60 kinase activity. Our observation that parathyroid hormone increased pp60 kinase activity rather than to increase the total content of pp60 at the protein level seems to be at variance with results previously reported by Yoneda et al.(12, 22) . However, the observation period of 24 h in our studies was probably to short to detect any significant increase in protein expression.

It has been speculated that herbimycin A irreversibly binds to active SH groups of target proteins resulting in the sterical hindrance of the active site of the kinase(21) . Recent evidence suggests that benzoquinone ansamycins may inhibit pp60 activity in a more indirect fashion via inhibition of the heat shock protein HSP90-pp60 heterocomplex formation(30, 31) . From our results it is evident that herbimycin A as well as the mycotrienins can exert a direct inhibitory effect on pp60 itself as demonstrated in the in vitro kinase assay.

The half-maximal inhibition (IC) of mycotrienin I and mycotrienin II detected in the pp60 kinase assay differ from the IC found in the fetal rat long bone assay in that a 2-5-fold higher concentration is needed for the inhibitory effect in the kinase assay. Several possible explanations could explain this apparent discrepancy. If the mycotrienin binds reversible to the kinase, events during several washing steps and the immunopurification procedure could result in the loss of inhibitor. This loss could explain the difference in half-maximal inhibition values obtained in the bone resorption assay and kinase assay. In addition our results do not exclude that mycotrienins might have further targets in the fetal rat long bones which potentiate their antibone resorptive capacity in vitro.

The polypeptide hormone calcitonin plays an important role in the physiological regulation of bone resorption. Calcitonin exerts its inhibitory effect on bone resorption, at least in part, via an inhibition of pp60(22, 32) . Short term exposure of murine bone marrow cells to the hormone led to an inhibition of pp60 kinase activity while long term exposure also diminished the total amount of the pp60 protein. In the fetal rat long bone resorption assay salmon calcitonin is a powerful inhibitor of osteoclast-mediated bone resorption (IC value around 7.4 times 10M,(33) ) at the initial phase of culture. After an extended culture time (up to day 5), the system becomes less sensitive to calcitonin and bone resorption returns to the parathyroid hormone-stimulated resorption level (``escape phenomenon''). In this respect it is interesting to note that the mycotrienin-induced inhibition was complete and persisted for the entire culture period of 5 days. Thus, inhibiting the osteoclast downstream of the calcitonin receptor would be an approach to overcome the escape phenomenon that is associated with a calcitonin.

Parathyroid hormone has been shown to stimulate both osteoclast activity and osteoclast recruitment(34) . Bone resorption, as assessed by the fetal rat long bone system, is almost entirely dependent on the activation of mature osteoclasts. Thus, this particular system does not allow discriminating whether the mycotrienins inhibit either one or both, parathyroid hormone-stimulated osteoclast recruitment and osteoclast function. Results obtained with the pp60 knock-out mouse indicate that osteoclast formation is normal but that the osteoclasts are not functional in the animals as indicated by the absence of the ruffled borders(6) . However, previous findings using bone marrow cultures indicated that the total amount of tartrate-resistant acid phosphatase-positive multinucleated cells in 1,25-dihydroxyvitamine D (3) -treated cultures is significantly inhibited in the presence of herbimycin A(22) . Taking these findings into consideration it cannot be excluded that the mycotrienins interfere with cellular processes distinct from the pp60 signaling cascade which lead to an inhibition of osteoclast formation as well.

In conclusion, our results are consistent with the observation that pp60 is essential for normal osteoclastic bone resorption and point to a potential for pharmacologic intervention in the bone resorption process at the level of pp60. Therefore, the mycotrienins may have a therapeutic potential as bone resorption inhibitors in diseases where bone resorption is increased.

FOOTNOTES

*: The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked ``advertisement'' in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§: To whom correspondence should be addressed: Sandoz Pharma Ltd., Preclinical Research 360/406, CH-4002 Basle, Switzerland. Tel. 41-61-324.77.36; Fax: 41-61-324.47.74.
(): The abbreviations used are: PAGE, polyacrylamide gel electrophoresis; hPTH-(1-34), human parathyroid hormone peptide fragment(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34) .

ACKNOWLEDGEMENTS

We are grateful to Dr. J. T. Parsons for kindly providing monoclonal antibody GD11 and to Heinz Anklin, Funda Kalkuloglu, and Genéviève M. M. Kuntzelmann for their skilled technical assistance. We thank Karl-Heinz Wiederhold for the production of the photographs. Finally, we thank Dr. Francis Cardinaux for critical reading of the manuscript.

REFERENCES

Whyte, M. P. (1993) in Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism , 2nd Ed., pp. 327-344,, Raven Press, New York
Kodoma, H., Yamasaki, A., Nose, M., Niida, S., Ohgame, Y., Abe, M., Kumegawa, M., and Suda, T. (1991) J. Exp. Med. 173, 269-272 [Abstract/Free Full Text]
Yoshida, H., Hayashi, S., Kunisada, T., Ogawa, M., Nishikawa, S., Okamura, H., Suda, T., Shulz, L. D., and Nishikawa, S. (1990) Nature 345, 442-444 [CrossRef][Medline] [Order article via Infotrieve]
Soriano, P., Montgomery, C., Geske, R., and Bradley, A. (1991) Cell 64, 693-702 [CrossRef][Medline] [Order article via Infotrieve]
Horne, W. C., Neff, L., Chatterjee, D., Lomri, A., Levy, J. B., and Baron, R. (1992) J. Cell. Biol. 119, 1003-1013 [Abstract/Free Full Text]
Boyce, B. F., Yoneda, T., Lowe, C., Soriano, P., and Mundy, G. R. (1992) J. Clin. Invest. 90, 1622-1627
Lowe, C., Yoneda, T., Boyce, B. F., Chen, H., Mundy, G. R., and Soriano, P. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 4485-4489 [Abstract/Free Full Text]
Cooper, J. A. (1989) in Peptides and Protein Phosphorylation , pp. 85-113, (Kemp, B., ed) CRC Press, Boca Raton, FL
Levy, J. B., and Brugge, J. S. (1989) Mol. Cell. Biol. 9, 3332-3341 [Abstract/Free Full Text]
Golden, A., Nemeth, S. P., and Brugge, J. S. (1986) Proc. Natl. Acad Sci. U. S. A. 83, 852-856 [Abstract/Free Full Text]
Koch, C. A., Anderson, D., Moran, M. F., Ellis, C., and Pawson, T. (1991) Science 252, 668-674 [Abstract/Free Full Text]
Yoneda, T., Lowe, C., Lee, C. H., Gutierrez, G., Niewolna, M., Williams, P. J., Izbicka, E., Uehara, Y., and Mundy, G. R. (1993) J. Clin. Invest. 91, 2791-2795
Hall, T. E., Schaeublin, M., and Missbach, M. (1994) Biochem. Biophys. Res. Commun. 199, 1237-1244 [CrossRef][Medline] [Order article via Infotrieve]
Sugita, M., Natori, Y., Sasaki, T., Furihata, K., Shimazu, A., Seto, H., and Otake, N. (1982) J. Antibiot. 35, 1460-1466 [Medline] [Order article via Infotrieve]
Sugita, M., Saski, T., N., Furihata, K., Seto, H., and Otake, N. (1982b) J. Antibiot. 35, 1467-1473 [Medline] [Order article via Infotrieve]
Sugita, M., Natori, Y., Sueda, N., Furihata, K., Seto, H., and Otake, N. (1982) J. Antibiot. 35, 1474-1479 [Medline] [Order article via Infotrieve]
Raisz, L. G. (1965) J. Clin. Invest. 44, 103-116
Moyers, J. S., Bouton, A. M., and Parsons, S. J. (1993) Mol. Cell. Biol. 13, 2391-2400 [Abstract/Free Full Text]
Feyen, J. H. M., Evans, D. B., Binkert, C., Heinrich, G. F., Geisse, S., and Kocher, H. P. (1991) J. Biol. Chem. 29, 19469-19474
Gubler, H. P. (1989) SIGFIT V1.0: an IBM PC Program for Dose-response Curve Analysis , Sandoz Ltd., Basel
Uehara, Y., and Fukazawa, H. (1991) Methods Enzymol. 201, 370-379 [Medline] [Order article via Infotrieve]
Yoneda, T., Niewolna, Lowe, C., Izbicka, E., and Mundey, G. R. (1993) Mol. Endocrinol. 7, 1313-1318 [Abstract/Free Full Text]
Alam, A. S., Huang, C. L., Blake, D. R., and Zaidi, M. (1992) Biosci. Rep. 12, 369-380 [CrossRef][Medline] [Order article via Infotrieve]
Ries, W. L., Key, L. L., Jr., and Rodriguiz, R. M. (1992) J. Bine Miner. Res. 7, 931-939
Garrett, I. R., Boyce, B. F., Oreffo, R. O., Bonewald, L., Poser, J., and Mundy, G. R. (1990) J. Clin. Invest. 85, 632-639
Akeson, A. L., Woods, C. W., Mosher, C. B., Thomas, C. E., and Jackson, R. L. (1991) Atherosclerosis 86, 261-270 [CrossRef][Medline] [Order article via Infotrieve]
DeBoer, C., Meuleman, P. A., Wnuk, R. J., and Peterson, D. H. (1970) J. Antibiot. 23, 442-447 [Medline] [Order article via Infotrieve]
Omura, S., Iwai, Y., Takahashi, Y., Sadakane, N., and Nakagawa, A. (1979) J. Antibiot. 32, 255-261 [Medline] [Order article via Infotrieve]
Uehara Y., Musakami, Y., Mituno, S., and Kawai, S. (1988) Virology 164, 294-299 [CrossRef][Medline] [Order article via Infotrieve]
Whitesell, L., Shifin, S. D., Schwab, G., and Beckers, L. M. (1992) Cancer Res. 51, 1721-1728
Whitesell, L., Mimnaugh, E. G., De Costa, B., Myers, C. E., and Neckers, L. M. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 8324-8328 [Abstract/Free Full Text]
Izbicka, E., Yoneda, T., Niewolna, M., Lowe, C., and Mundy, G. R. (1992) J. Bone Miner. Res. 7, Suppl. 1, S145
Feyen, J. H. M., Cardinaux, F., Gamse, R., Bruns, C., Azria, A., and Trechsel, U. (1992) Biochem. Biophys. Res. Commun. 187, 8-13 [CrossRef][Medline] [Order article via Infotrieve]
Lorenzo, J. A., Raisz, L. G., and Hock, J. M. (1983) J. Clin. Invest. 72, 1924-1929