CA2048673A1 - Metabolic effects of leukaemia inhibitory factor on bone - Google Patents

Abstract

The present invention relates generally to the effect of leukaemia inhibitory factor (LIF) on bone in animals and more particularly, to the use of LIF and/or LIF-like polypeptides to induce, promote and/or enhance metabolic effects on bone in animals.

Description

wo go/l~ 2 0 L~ O E 7 ~ PCT/AU90/~92 ~ ,. ..
.,. 1 METABOLIC ~FF~CIS OF L~UKA~HIA INHI~ITORY YACIOR ON BON~

The present lnventlon relates generally to the effect of leukaemla lnhlbltory faetor (LIF) on bone ln anlmals. More partlcularly, the present lnventlon relates to the use of LIF and/or LIF-llke polypeptides to induce, promote and/or enhance metabolle effeets on bone in anlmals.

The cytokine, leukaemla lnhlbltory factor (LIF), ls a protein that has prevlously been purlfled, cloned and produced ln large quantltles ln purlfled recomblnant form from both Escherlchia coli and yeast eells (Internatlonal Patent Applleatlon No. PCT/AU88/00093).
LIF was orlginally lsolated on the basis of lts capaclty to induce dlfferentlatlon ln and suppress the murlne myelold leukaemic eell line, Ml. LIF has been shown to have a powerful differentiation suppresslng actlon on normal embryonic stem cells. LIF has no apparent proliferatlve effeets on normal hemopoietie cells although LIF receptors were detected on cells of the monocyte-maerophage lineage.

Several eytoklnes and growth faetors have been lmplleated in bone remodelllng and ln partleular have been shown to promote bone resorptlon in vitro. The term 'osteoclast aetivating factor' (OAF) was applied to the bone resorblng aetlvlty deteeted in supernatants from leetin-trangformed lymphoeytes (1), and was used to explaln the excessive bone resorption aeeompanylng eertain hematologieal malignaneies. It is now aecepted that OAF in faet eonsists of several eytokines, lncluding interleukin l (IL-l) and tumour neerosis faetors alpha and beta (~NF~, TNFB)~ all of whieh are potent bone re~orblng agents (2).

W~ 2a4~67; ~ PCT/AU90/M~92 ':,' L~F has no slgnlflcant structural or sequence homology wlth the aforementloned cytoklnes. It was surprlslng, ther~fore, that followlng an analysls of the possible phy~iologlcal and blochemical effects of LIF on bone remodellin~ ln vitro and in vivo, lncluding studies of the cellular targets of LIF actlon, lt was dlscovered, ln accordance wlth the present invention, that LIF has a metabollc effect on bon~ by lnduclng, promoting and/or enhanclng bone resorptlon and/or b~ne formation.
In accordance wlth thls inventlon, lt has been discovered that LlF ls an lmportant paracrine regulator in bone and exerts a number of specific blochemlcal actions on osteoblasts. These lnclude lnhlbltion of plasminogen actlvator (PA) actlvity in osteoblasts, most llkely through lncreased synthesls of PA inhibltor-l formatlon. Furthermore, LIF ltself is expressed ln osteoblastic cells and primary osteoblasts. Thus several lines of evidence establish cells of the osteoblast lineage as targets for LIF action ln a paracrine or autocrine manner.

Accordlngly, one aspect of the present invention relates to a method of lnducing, promoting and/or enhancing metabolic effects on bone in an animal comprising adminlsterlng to sald animal an effective amount of leukaemla lnhibltory factor (~IF) and/or ~IF-llke polypsptides alone or ln comblnatlon wlth other cytoklne~ for a tlme and under condit~ons sufflclent to effect bone metabolism.

Another aspect of the present lnvention is dlrected to a pharmaceutical composition useful in lndu~lng, promoting and/or enhancing metabollc effects on bone ln animal~ comprlsing an effectlve amount of lsuknemla lnhibltory factor (LIF) and/or LIF-llke polypeptldes in comblnation wlth one or more other wo ~/,04~ 2 0 ~ PCT/AU~/~92 s 3 s molecules such a8 cytoklnes, lnorganlc or organlc molecules and whlch affect bone resorptlon, calclflcatlon and/or formatlon and one or more pharmaceutlcally acceptable carrler~ and/or dlluents.

Stlll yet another aspect of the present lnventlon contemplates the use of leukaemla lnhibltory factor (LIF) and/or LIF-llke polypeptldes ln the manufacture of a medlcament useful ln the lnductlon, promotlon and/or enhancement of metabollc effects on bone ln anlmals.

A further aspect of the present lnventlon contemplates a method for the treatment of ln~urles and/or dlseases whlch result ln bone degeneratlon or fracture ln an animal comprising administerlng to said anlmal a bone-forming increaslng-effective amount of leukaemia lnhibitory factor (LIF) andtor LIF-llke polypeptldes alone or in comblnatlon wlth other actlve molecules for a time and under conditions sufficlent to effect bone calclflcatlon.

In a preferred embodlment of the subJect inventlon, the anlmal is mammallan or avian and more preferably ls human.
The followlng abbrevlatlons are used hereln:

LIF Leukaemla inhlbltory factor TNFa Tumour necrosis factor alpha 30 TNFB Tumour necrosls factor beta TGFB Transformlng growth factor beta PTH Parathyrold hormone PA Plasmlnogen actlvator PAI Plasminogen actlvator lnhlbitor 35 CSP Colony stimulatlng factor GM-CSP Granulocyte-macrophage colony-stimulating factor PCT/AU~/~n2 W090/1~ ~ 2~367~ ~

PBS Phosphate buffered sallne SDS Sodium dodecyl sulphate BSA Bovlne serum albumin MEM Minlmal essential medlum 5 DNA Deoxyrlbonucleic acid RNA Ribonucleic acid FCS Fetal calf serum W Ultra vlolet FD cells CSF-dependent continuous hemopoietic cell line FD/LIF cells FD cells expressing LIF gene HEPES 4-(2-Hydroxyethyl)-l-plperazinoethanesulfonic acid The present invention arose, in part, from an analysis of the effects of ~IF on certain responses in osteoblast-like cells and its effects on bone resorption in calvarlae in organ culture. Furthermore, to test the effects of LIF in vivo and to delineate LIF ' 8 physiologlcal functions, mice were generated with chronically-elevated levels of LIF. With respect to this latter test, use was made of cells of the colony stlmulating factor ( CSF ) -dependent continuous hemopoietlc cell line FDC-Pl ("FD" cells) which, although non-leukaemic, have been found to persist and accumulate inthe spleen, lymph n4des ~nd bone marrow of recipient animals. By superlnfecting FDC-Pl cells with a retrovirus containlng the gene which encodes LIF, cloned sublines were developed that constitutively produce high levels of LIF and such cells were inJected into mice to generate animals containing a continuous source of LIF-producing cells. Upon analysis of these animals, together wlth the in vitro data, it was discovered that ~F 18 an lmportant regulator of bone cell functlon causlng the induction, promotlon and/or enhancement of ~uch actlvltles as bone resorption and/or bone formation.

WO ~/1~ ~ ~ PCT/AU~/~92 ~ he ln vlvo exempllfled aspects of the present lnventlon are performed uslng a mouse model. Thls 18 done, however, wlth the understandlng that the present lnventlon extends to the effect of LIF on bone ln all anlmals. Preferably, however, the anlmals are mammallan or avlan but more preferably mammallan. In lts most preferred embodiment, the present lnventlon ls dlrected to humans.

Accordlngly, one aspect of the present lnventlon relates to a method of inducing, promoting and/or enhancing metabolic effects on bone ln an animal comprising adminlstering to said animal an effective amount of LIF and/or LIF-like polypeptides alone or in combination with one or more other cytokines or one or more active substances, for a time and under conditions sufficlent to effect bone anabollsm.

"Metabolic effects", as used ln the specification and claims herein, include, although are not necessarily limited to, such processes as bone formation and/or bone resorption, increased protein synthesis in bone tissue and increased thymidine uptake in bone tissue. ~he term "metabolic effects" also encompasses anabolic and/or catabolic activities of LIF on bone.

By "LIF and/or LIF-like polypeptides~ is meant derivatives and homologues of LIF and extends to the entire natural len~th LIF molecule or derivatives thereof carrying single or multiple amlno acld substltutions, deletions and/or addltions and includes substitution, deletion and/or additlon of any other molecules associatea wlth LIF and lts derivatlves such as, inter ~ , carbohydrate, lipid and polypeptide moieties, provided said derivatlves retain sufficient activity to ~e useful ln the practice of the present lnvention. LIF-llke polypeptldes extend ~o molecules having wo ~/10454 2 0 ~ ~ i; 7 ~ PCI/~U90/01~092 substantlally slmllar actlvity as LIF whlle carrylng amlno acld rearrangements or alteratlons. The preparatlon of various derlvatlves of LIF wlll be apparent from the dlsclosure in PC$/AU88/00093 of the recomblnant LIF molecule and lt~ correspondlng DNA.
Reference hereln to LIF encompasses LIF-llke polypeptldes and vlce verca.

~ence, the present lnventlon extends to naturally occurrlng but substantlally pure LIF (l.e. greater than or equal to 70~ by welght of LIF relatlve to other protelns or molecules and preferably greater than or equal to 85% and even more preferably greater than or equal to 90%), to recomblnant LIF and to synthetlc LIF
made, for example, by chemlcal means.

Depending on the animal to be treated or the dlsease state to be treated, LIF may be from a range o sources such as, but not llmlted to, human, mouse, dog, cow, plg, sheep or other rumlnant. In some cases lt wlll be preferable to use homologous LIF to the anlmal belng treated, for example, human LIF on humans. 8ut ln other clrcumstances, heterologous LIF may be more convenlent and/or mors effectlve. The cholce of source of LIF may depend on the exlgency of the treatment requlred or the anlmal requlring treatment.

In one embodlment of the sub~ect lnventlon, LIF
and/or LIF-llke polypeptldes are used alone. In another embodiment, they are used ln a comblnatlon wlth one or more other cytoklnes, such as, but not llmlted to, IL-l, TNFa and/or TNFB, and whlch affect varlous aspects of bone resorptlon, and/or formatlon. The present lnventlon also extends to the use of LIF and/or LIF-llke polypeptldes in comblnation wlth other actlve molecules such as agonists, antagonists, calclum or any lnorganlc or or~anlc molecule which d ds ln the bone resorbing and/or forming process WO 90/10454 ~ r~ ~ PCl`tAU90/1~0092 ~ !

elther dlrectly o~ by enhancln~ the effect of LIF or LIF-llke polypeptides.

The inJection of FD cells lnto mlce produces animals ln whlch a progresslvely lncreaslng populatlon of FD cells accumulate. InJectlon of FD cells produclng LIF
("FD/LIF cells") had the same consequence. In agreement wlth prevlous studles, this accumulation was accelerated by pre-lrradlatlng the reclplents. Northern analyses for LIF mRNA and bloassays for LIF lndlcated that actlve productlon of LIF was achleved ln certaln organs known to be the site of accumulatlon of FD cells, l.e. the spleen, lymph nodes and sometimes the marrow and there was a substantlal elevatlon of clrculatlng LIF levels. No elevatlon of LIF productlon or clrculatlng LIF levels were seen ln reclplents of FD cells. The consequence of lnJectlng LIF-produclng F~ cells was lnter alia the development wlthln an interval as short as 10 days of bone overgrowth with consequent extramedullary haemopolesls ln the spleen and llver.

The absence of any of these effects ln reclplents of FD cslls alone or ln cachectlc GM-CSF transgenlc mlce that have been studled extensively in the lnventors' laboratory argue that the pathologlcal changes observed are the speciflc consequence of excess LIF levels. Since such mlce often have few or no FD/LIF cells ln the marrow, the tissue effects seem llkely to be mediated by circulating LIF.
~ he most dramatic primary change was the pattern of exce~s bone formation wlth a hlstological appearance resembllng osteosclerosls. Thls involved three simultaneous changes: an accumulation and actlvation of osteobla~ts; a depletlon of pre-existing hemopoletic cells ~in unirradlated recipients); and, ln many bones, actlvatlo~ of osteoclasts. Osteoblasts express receptors wo 90/104~4 2 0 ~ ~ ~ 7 ~ PC~/AU9o/~3~
;-;,, for LIF (27) and LIF 18 known to relea9e calclum ln vltFo from bone tissue (Reid et al. Personal Communlcatlon).
It ls concelvable, therefore, that the osteoblast changes were due to dlrect actlons of LIF whlch , lf so, would suggest that LIF may be capable of lnfluencing osteoblast precursor mlgratlon , osteoblast prollferatlon and bone formatlon. A well-documented association exists between osteoblast and osteoclast actlvlty but, slnce osteoclasts lack LIF receptors (27), thelr altered behaviour may be secondary to changes ln osteoblast actlvlty. Although the perturbation in bone biology resulted in excess bone formation, many mice can be assumed to have had excess calcium levels because of calcium deposits partlcularly in muscle and llver tissue.
The administration of L~F, theref~re, may, under appropriate conditions and requisite time, lnduce bone formation and may be extremely useful in the repair of fractured bones and to stimulate bone growth and should be partlcularly useful ln the treatment of lnJurles and/or dlseases where bone degeneration or fracture ls a symptom, or a result, such as ln lncreaslng bone calciflcation ln osteoporosis, such as ln post-menopausal osteoporosis, or ln stlmulatlng bone fracture repair.
Admlnistratlon of LIF wlll be by standard procsdures, for example, lntravenous, lnterperitoneal, lntramuscular, or toplcally to the extent that lt i8 applled to the fractured bone. Certaln modlflcations may need to be made to LIF in order to lncrease its serum half life or to protect ls from hydrolytic enzymes. All such modiflcations to LIF or lts pharmaceutical composition comprising LIF or LIF-like polypeptides are encompassed by the present lnventlon. Alternatlvely, the modiflcatlons may comprise adding protective compounds (e.g. enzyme inhlbitors) to pharmaceutlcal compositions.
Suitable carrier vehicles and their formulations, lnclu~lve of other human proteins (e.g. human serum wo ~/~ 7 ~ PCT/AU90/~92 f^', albumln), are de~crlbed, for example, ln Reminqton's Pharmaceutlcal Sclences 16th ed., 1980, Mach Publ~shing Company, edlted by Osol et aI which ls herein incorporated by reference. The effectlve amount of LIF
or LIF-like polypeptides requlred ln pharmaceutlcal composltlons or in vivo to lnduce metabollc effects will vary according to the animal and condition being treated.
For example, an effective amount of from 0.01 to 1 million unlts may be necessary but in some circumstances a range of 0.1 to 10,000 unlts may be required. In other circumstances it may be propitious to administer LIF at 0.01 ~g to 100,000 mg per kg of body weight of anlmal.
Another preferred range may be 0.1 ~g to 10,000 mg of LIF
or LIF-like polypeptides per kg of body weight of animal.
The aforementioned effective amounts may, depending on the situation, be per hour or per day or possibly per week or month. Administration may be single or multiple application. Accordingly, "effective amount" refers to the amount necessary to give the desired effect.
Pharmaceutical compositions useful in the induction of bone formation or increasing bone resorption comprising a bone formation-inducing or bone resorption-effective amount of LIF and/or LIF-like polypeptides may be prepared according to standard practices.
Furthermore, the present lnvention extends to the use of LIF and/or LlF-like polypeptides ln the manufacture of a medicament for the induction, promotion and/or enhancement of metabolic effects on bone in animals and in particular mammals or birds and even more particularly humans. In a preferred embodiment, the present invention contemplates the use of LIF and/or LIF-like polypeptides in the manufacture of a medicament for treating in~uries and/or diseases in animals, and in particular humans, where bone degeneration or fracture is a q ptom or a result such a8 in osteoporosis or bone fracture.

wo ~/1~ 2 ~ 7 ~ PCT/AU~/~N~2 The present inventlon extend~ to parts or derlvatlves of LIF and/or LIF-like polypeptides defined by thelr ablllty to interact with osteoblastc. In partlcular, the interaction may comprlse the blnding of the LIF fragment to a LIF receptor on an osteoblast cell. Such parts or derlvatlves of LIF and/or LIF-llke polypeptides would be isolated by any number of means such as chemlcally treatlng the LIF molecule, lsolatlng cleaved or altered sectlons and testing whether the sectloned or fragmented LIF has activlty ln bioassays as descrlbed ln PCT/AU88/00093. The actlve fragments would then be further analysed as descrlbed hereln uslng ln vitro technlques. Alternatively, selected fragments or portlons of LIF would be synthesized uslng standard recombinant DNA expression systems uslng segments or deriviates of the LIF coding region. The polypeptide so produced would then be tested uslng bioassays and, if actlve, would be further tested ln ln vltro osteoblast cell systems. Therefore, use of the term ~LIF-llke polypeptides" here1n extends to fragment~ or parts of LIF
capable of interacting wlth osteoblasts.

~he in vitro results descrlbed hereln suggest that LIF stlmulates bone resorption by a mechanlsm posslbly ~5 lnvolvlng prostaglandln productlon and lnvolvlng an lncrease in the number of osteoclasts (bone resorblng cell~) ln the calvarla. Its action on thymldlne incorporation probably reflects responses of osteoblasts, and does not appear to involve local prostaglandln production slnce it is unaffected by the presence of indomethacin. ~hese data further suggest that LIF may be an important regulator of bone cell function.
Furthermore, the effect of LIF on the osteoblast-like cells is largely metabolic, in that there i8 clear evldence that LIF effects an increase synthesls of some of the important specific components of bone protein.

WO gO/10454 2 U l~ ,?j ~, 7 ~, PCr/AU90/~tsi~J~

Additlonally, one of the clear speciflc actlons of LIF
reported herein ls upon the PA-plasmln system ln osteoblasts. Several bone-resorblng hormones stlmulate PA actlvlty ln osteoblast-llke cells ln culture. In order to localize the proteolytlc actlvlty of the PA-plasmln system at sltes where lt i8 required, preclse regulatory control ls needed. Thi~ i8 provlded by the stimulating hormones and paracrine factors and also by PA
inhlbitors (PAIs), which blnd to and llmlt the actlvlty of both tlssue-type PA (tPA) and uroklnase. The speclfic inhibitors, PAI-l and PAI-2, which bind to and inactivate PA, are the products of separate genes.

In the present work, the production of PAI-l by normal ost20blasts ls ldentlfled from rat calvarla and from the osteoblast-like osteogenic sarcoma line, UMR 106-01.
Furthermore, LIF ls shown to enhance productlon of PAI-l protein and mRNA, an effect which is most llkely responslble for the net decrease in PA actlvity released by the cells.

A further posltive action of LIF is the abillty to enhance the stlmulation of alkallne phosphatase activlty by retlnoic acld ln UMR 201 cells, an effect simllar to that obtalned previously with recombinant TNF (11).
Although the effects of LIF resemble ln some respects those of ~GFB the latter has no such effect on alkaline phosphatase activity.

LIF, therefore, exhibits a regulatory effect on the PA-PA
inhibitor system, further evidencing the potential of LIF
as an important regulator in the mlcroen~lronment of bone cells and, therefore, lmportant in bone remodellin~. The effect of LIF on PA does not appear to be acting by the generation of TGFB~

wo go/10454 2 ~ ~ ~ 6 ~ ~ PCr/AU~O/sH~
s ~,.. .

The present lnventlon 18 also descrlbed by the followlng non-limltlng Flgures and Examples. Although the Examples use mlce as the anlmal model, the technlques used and results obtalned wlll be egually appllcable to other animals includlng avlan and mammallan (e.g. human~
specles, and, hence, the pre~ent inventlon encompasses the use of LIF in lnduclng, promotlng and/or enhancing bone formatlon and resorptlon and other metabollc activites in all animals.
In the Figures:-Figure 1 is a graphical representation showingsurvival of mice, with or wlthout exposure to radiatlon, following in~ection of FD cells expresslng LIF.
Figure 2 is a graphical representation showing serum levels of LIF ln irradiated and non-lrradlated mice followlng lnJection of FD cells and FD cells expressing LIF.
Figure 3 is a pictorial representation of the femur from mouse in~ected w~th FD cells (A) and FD/LIF
cells (8). Note the excess new bone formatlon and abnormal trabeculae in FD/LIF recipient.
tHematoxylln/eosln ~H & E; x 54)]. (C) Higher power vlew of marrow from a reciplent of FD/LIF cells show$ng ~5 depletion of hemopoletlc cells, excess numbers of osteoblasts, and nsw bone formatlon. (H & E: x 450). (D) Heart of reclplent of FD/LIF cells showlng calclum deposlts (arrows). (H & E; x54). (E and F) Pancreas from recipient of FD cells (E) and from recipient of FD/LIF
cells (F). Note acinar degeneration and edema but normal islet (arrow). (H & E; x 54). (G and H) Ovary from a reciplent of FD cells (G) (the normal corpora lutea are ln~lcated by arrows) and from a reclplent of FD/LIF cells (H). Note the absence of-corpora lutea in H. ~H & E; x 54)-wo go/1~54 ~ ~ ~ 3 ~ ~ 8 PCT/AU~/~X~2 ~, .

Flgures 4 and 5 are graphlcal repre~entatlons showlng, respectlvely, murlne and human LIF-stlmulated '5Ca release from pre-labelled mouse calvarla ln a dose-dependent manner.
Flgure 6 ls a graphlcal representatlon showlng that the lncrease ln bone re~orptlon 18 assoclated wlth an increase ln osteoclast numbers in the calvaria.
Figure 7 ls a graphlcal representatlon showing the abllity of lO-'M indomethacin to block LIF stlmulated bone resorption.
Figure 8 is a graphical representatlon ~howing that t'H]-thymidine incorporatlon lnto the calvaria is stimulated by LIF.
Figure 9 is a graphical representation showing that stimulation by LIF of [3H]-thymidine lncorporation into the calvarla is not blocked by the presence of indomethacln.
Figure 10 ls a graphical representatlon showlng PA
activity in calvarial osteoblasts (A) and UMR 106-01 20 cells ( B ). UMR 106-01 cells were incubated in the presence (-) and absence (o) of a submaxlmal (20ng/ml) dose of PTH. Results are the means + SEM of triplicate determinations.
Figure 11 18 a photographic respresentation showing (A) ~ime courses of PAI-l mRNA levels followlng treatment of UMR 106-01 cells with L~F (1000 U/ml).
Total RNA (lO~g) was electrophoresed on an agarose/formaldehyde gel, transferred to Hybond N and the RNA hybrldized wlth labelled RNA transcripts; (B) Response of PAI-l mRNA followlng 1 h treatment of UMR
106-01 cells wlth increasing concentrations of LIF.
Condltlons were as descrlbed ln (A).
Figure 12 18 a photographic representatlon of an autoradiograph o a Northern blot 8howing the effect of ~IF on mRNA or al(I) collagen (A) and osteonectln (B) ln UMR201 cells. The 6 lane8 on the left ln each case are the ~I~ treatment samples. ~he cells were lncubated wlth PC~/AU90/~W~2 WO~0/10454 ~o~7 ~

LIF for 24 hours ln 2~ v/v uv lrradlatea ret~l calf serum. An allquot of 20~g total RNA was electrophoresed as described ln Example 5.
Flgure 13 ls a photographlc representatlon showing autoradlographs of cells from newborn rat long bones labelled wlth "sI-LIF. Speclflcally, the photograph shows osteoblasts and an osteoclast 1n the absence of unlabelled competltor (x400).
Flgure 14 ls a dlagrammatlc representation showing (A) saturatlon blndlng lsotherm of total (o), speclflc (~) and non speclflc (o) blndlng of "sI-LIF to UMR 106-06 cells at 4C. (B) scatchard analysls of speclfic blnding data illustrated ln panel A.
Flgure 15 ls a photographlc representatlon showing the detectlon of LIF transcrlpts ln osteoblastic cells.
Equlvalent amounts of RNA from UMR 106-06, UMR 201 or osteoblast-rich primary cultures (1Ca) were probed for LIF or GAP-DH transcripts (as a posltlve control) by PCR
ampllflcatlon as described ln Example 5. ~he negative control track represents a "no cell~ sample carrled ln parallel through RNA extraction, cDNA synthesis and PCR
reaction. In the case of UMR 106-06 and UMR 201 cells, the + and - symbols refer to treatment wlth or without TGF B (2ng/ml) for the number of hours lndlcated.
DMPL~ 1 General effect of ~IF in ~1 oe Commenclng 10 days after the inJection of FD/LIF
cells, mice that had prevlously been glven 6 Gy whole-body irradiatlon began to show evidence of welght loss, ruffled halr and restrlcted movement. Withln days, such mice became morlbund and by 28 days after ln~ection 80~
of these mice had become moribund (Flgure 1). Irradlated mlce ln~ected with control FD cells remalned in good health durin~ thls perlod. Unlrradiated mice ln~ected wlth PD/LIF cells developed an identical clinical wo ~/104~ 2 0 ~ 8 ~ ~ ;. rc~/AU~/~92 .. 15 syndrome but wlth a delayed onset. Unlrradlated mlGe flr~t became morlbund 30 days after lnJectlon but thereafter the rate of dlsease development wa~ parallel to that of ~rradiated mlce, 80% becomlng morlbund by 56 days. Agaln, unlrradlated mlce ln~ected wlth control FD
cells remalned apparently healthy durlng this perlod.
All the observatlons were performed on morlbund mlce ln~ected wlth FD/LIF cells, control mlce lnJected wlth FD
cells belng kllled on each occaslon for comparatlve analysls. ~he serum levels of LIF in lrradiated and non-irradlated mice following in~ection of FD and FD/LIF
cells are shown in Figure 2.

Moribund irradiated and non-lrradlated mlce ln~ected with FD/LIF cells appeared dehydrated and showed a conslstent welght reductlon of 21% and 24%, respectlvely, compared with control mice. The morlbund mice exhlbited circulatory collapse and a reduced circulating blood volume. The hematocrit levels (Table 1) did not exhlbit an expected rlse due to homoconcentration, suggesting a reduction in red cell mass. Collection of serum from such mice was difflcult because of the reduced volume of blood obtalned at killing and a failure of clot retractlon in specimens from FD/~IF mlce.

PCJ /AU90tO0092 wo 90/~04s4 2 ~

TABL~ 1 Paraueter~ of Moribund Mice In~ected with FD~LIF cells 5 Parameter OGY 6Gv FD/LIF FD FD/LIF FD
-Number of Mice Analysed13 9 12 12 Body Welght g 17.8+2.1 22.4+1.8 Haematocrit % 47+8 30+4 49+4 42+2 Total Whlte Cells/~l19,940+ 5470+10,600+ 2370+

Polymorphs 10,780+ 630+3909130+ 630+310 Lymphocytes 7870+ 4600+ 850+/60 1130+590 Monocytes 1270+ 230+40 610+350 3370+200 Eosinophils 30+60 20+30 10~30 10+10 -t~rd d l-tlo u 13~CAI~PI~ 2 He Doietic channes Whlte cell levels were consistently elevated in FD/LIF mice due to a selective 4-5 fold elevation of mature neutrGphils (Table 1). In unirradiated mice, monocyte levels were also signiicantly higher in mice inJected with FD/LIF cells than ln control mice in~ected wlth FD cells.
All mice lnJected with FD/LIF cells exhibited moderate spleen enlargement that was more evident in ~ 7~ ' PCT/AU~/0HWn W090/t04~ ~40~7~ t ~-. ;., I
,..;

Spleen and Marrow Change~ ln Mlce In~ected with FD/LIF Cells Treatment OGy 6Gy -FD FD/LIF FD FD/LIF

SPLE~EN
Number of Mice 12 21 15 16 Spleen Weight mg110+30 230+70 120+40 170+40 pE~r Blasts3.2+ 2.0 6.1+ 3.3 2.5+ 2.17.9+ 4.0 Myeloblasts/ 1.3+ 2.2 7.7+ 6.72.5+ 2.0 9.8+ 5.2 Myelocytes Neutrophils3.2+ 2.6 15.0+ S.95.6+ 2.5 26.0+18.2 Lymphocytes77.6+ 8.0 31.5+13.7 36.9+12.09.71 5.0 Monocytes2.4+ 2.1 7.5+ 3.1 5.8+ 3.111.7+ 4.9 Eosinophlls5.4+ 4.9 3.1+ 3.0 16.6~11.02.1+ 2.0 Nucleated6.8+ 3.0 29.1+17.9 30.6+20.032.7+27.7 Erythroid Cells BON~ MARROW
Number of Mice 10 8/19 14 9/19 ~otal Cells x 106 25.2+ 5.9 5.5+ 0.724.0+ 4.0 6.0~4.0 Blast~3.2~ 1.2 3.0+ 1.93.7~ Z.34.1+ 4.4 Myeloblasts/ 5.8+ 3.411.1+ 3.88.4+ 5.7 10.8+ 7.4 Myelocytes Neutrophils33.6~ 9.166.9+11.0 29.1+12.6 68.6+25.3 Lymphocytes23.7+ 5.37.5+ 5.021.1+ 8.1 2.7+ 3.2 Monocytes5.2+ 2.36.3+ 3.95.9+ 3.77.0+ 5.2 Eosl phils19.2+13.83.1+ 1.914.3+11.7 2.3+ 2.5 Nucleated9.3+ 4.32.1+ 1.117.5+ 7.74.5+ 7.3 Erythrold Cells , Mean Value~ + Standard Deviatlons W0 ~/104~ 2 0 ~ PCT/AU90/~WW~

unlrradlated reclplents (Table 2). The spleen was sllghtly paler than normal but exhibited no surface nodules and in both types of reciplent exhlblted a reduced percentage of lymphoid cells, an elevated percentage of granulocytlc cells and, partlcularly ln unlrradlated reclplents, an elevated percentage of nucleated erythrold cells.

A striking feature of the femoral bone marrow in mice lnJected wlth FD/LIF cells was the dlfflculty experlenced ln insertlng a needle to flush out marrow cells. Marrow cells were not able to be recovered from 10 of 19 lrradlated and 11 of 19 unlrradlated mlce.
Where cells were recovered from mice in~ected with FD/LIF
cells, there was a 4-5-fold reduction ~n total cell number (Table 2) ln marrow from both types of reclpient.
In these marrows there was a signiflcant rise in the proportion of granulocytic cells and a significant reduction in lymphoid and nucleated erythroid cells.
Interestingly, recipients of FD cells exhibited a ma~or increase in eosinophils, usually myelocytes and promyelocytes, a change not seen ln the marrow of recipients of FD/LIF cells.

BXAMPL~ 3 Bone chanaes The femur, tibia and sternum in mice injected with FD/LIF cells showed a remarkable overgrowth of bone, seen as 30 thickening of the bone cortex and the formation of irregular-shaped trabeculae in marrow cavity (Figure 3).
The marrow cavity was convoluted and narrowed in unirradiated and irradiated recipients and was depleted of hemopoietlc cells, those remalning being mainly granulocytic 35 cells althouah in some ~ut not all marrows, megakaryocytes were present, Replscing the hemopoietic tissue, particularly in the ends of the long bones were elongated PCT/AU901~92 ~ . .

cells orlented ln parallel bundles. These cells were enlarged close to bone surface and merged wlth the developed bone. From the calcification evident in the cytoplasm of these cells, they appeared to be osteoblasts but, although 5 the total number seemed well in excess of the number that could have been present ln the normal marrow cavlty, no obvlous mlgration was ~een through bone foramina. In some bones, osteoclasts were prominent and ad~acent bone had irregular scalloped edges. Irregular foramina were also 10 present in many bones particularly at the ends of long bones. These could have been the consequence of increased osteoclastic activity. The open channels between the marrow and the tissues outside the bone could have allowed free entry of osteoblast precursors or exit of marrow cells.

The red pulp of the spleen was enlarged and was packed with hemopoletlc cells, the lymphoid folllcles belng severely depleted of cells. In many llvers, there were also prominent foci of hemopoietlc cells and some llvers showed 20 areas of necrosis and/or calclflcatlon. Less commanly, cirrhotlc changes were present but there was no mitotlc activlty ln parenchymal cells and there was conslstent reductlon ln the volume of individual parenchymal cells.

Areas of calcification were seen ln strlated muscle and the heart ln some but not all anlmals but no foci of cellular lnfiltration were observed in these locations. In several mice, prominent macroscopic areas of calcification were vlsible ln the llver.

Mlce in~ected wlth FD cells showed none of the above pathologlcal changes and the only condltlon evldent in some of these mice was the presence of small focal accumulations of FD cells ln the spleen or mesenterlc node.

7 PCT/AU~/~92 W0 ~tl~ ~ 2 0 ~
~"-"
' In vitro Effect~ of LIF ln neonatal mouse calvarla Bone resorptlon was measured a~ the release of 45Ca 5 from bones whlch had been pre-labelled in vivo. Heml-calvarla were dlssected as prevlously described (3) and prelncu~ated for 24 hours ln minlmal e~sentlal medla plus 1%
v/v fetal calf serum. ~ones were then changed to medla containlng the experimental compounds and incubated for a 10 further 48 hours, at the end of whlch tlme medla and bone were harvested for measurement of 'sCa content. In experiments involvlng the measurement of t'H]-thymldlne and t'H]-phenylalanine, these compounds were added in the last 4 hours of the 48 hour experimental perlod and thelr content lS ln the calvarla quantified by the method of Canalis (4).

Both recombinant murine (Figure 4) and human (Figure 5) LIF stlmulated 45Ca release from pre-labelled calvarla in a dose-dependent manner. The increase in bone 20 resorption was associated wlth an increase in osteoclast numbers in the calvaria (Figure 6). ~he ability of LIF to stimulate bone resorption was blocked by the presence of 10-'M lndomethacin, an inhibitor of prostaglandin synthesis (Figure 7; ~able 3) The effect of LIF on t~]-thYmldine incorporation, a measure of cell proliferatlon, was also 8tudied. ~IF wa~ -found to stimulate t~H] - thymidine incorporation into calvaria but the dose response relationship was diferent 30 from that for bone resorption in that it was a more sensitive response (Figure 8). Furthermore, this effect of LIF was not blocked by the presence o indomethacin (Figure 9) .

LIP also increased the incorporation of t'H]-phenylalanine into bones (control 21420 ~ 1110 cpm; LIF

WO 90/lWS4 ~ 7 ~, PCr/AU90/00092 ~;
.

T~ble 3 E~r~c~ or indomethacin on lhc LIF r~sponse lo PA Jn calv-rbl osleobl~sls snd UMR 106 01 ctlls.
P~ ~cdvil~ (% Tcpm ) I S x 10-5M
Ostcoblasts Control 13.0~ Q9 10.6 ~0.9 LIF 7.0 + 0.7 SA ~ 0.3 Control 12.~ ~ 2.2 9.4 I Ql LlI: 4.3 ~ O S 4.S ~ 0.4 Calvarial osleoblasts and UMR 106 01 cells were subculturcd onto 12S-I fibrin eoated plates as described in the Melhods seetion. Cells wtre preineubated for 8 h in ~erum ~?ec eondidons as described in the Iegend in Table 4 . The cells were ineubatcd in thc prcsenec nd absence of LII: (1000 Ut~nl ) and indomethacin (I.S x IO SM ).

wo 90/l~ 2 0 ~ ~ 6 7 ~ 22 PCT/AU~/~W~2 26890 + 1253 cpm, p < 0.01), lndlcatlng stlmulatlon o proteln synthesis.

Furthermore, LIF has slgnlflcant effects on t'H]-5 thymidlne $ncorporatlon ln neonatal mouse calvaria even atdoses as low as 100 unlts/ml (Table 4).

Accordlngly, these results lndlcate that LIF
stimulates bone resorptlon by a mechanism posslbly lnvolvlng 10 prostaglandin production and lnvolving an increase ln the number of osteoclasts (bone resorblng cells) ln the calvaria. Its action on thymldine incorporation probably reflects responses of osteoblasts, and does not appear to involve local prostaglandin production since it is 15 unaffected by the presence of indomethacln. These data further indlcate that LIF may be an lmportant regulator of bone cell function.

Experiments were carried out to determine the effect of LIF on the expresslon of mRNA o the bone-speclfic 25 proteins, al(I) collagen and osteonectin, and upon plasminogen actlvator (PA) actlvlty and the productlon of mRNA for plasminogen activator lnhlbitor - l(PAI-l) in osteoblast-like cells.

30 1. MAT~RIALS AND METHODS
(i) Chemical and bioloaical materials Synthetic human parathyroid hormone (PTH) (1-34) was from Beckman Pty. Ltd., Palo Alto, CA, U.S.A., while human recombinant TGFB~ TNFa and TNFB, and full-length rat 35 actln cDNA were gifts from Genentech Inc., San Francisco, CA, U.S.A. 1,25-dihydroxyvitamln D, was generously provided by Dr. M. Uskokovich, Hofman La Roche Inc., New Jersey, and WO 90/10454 ~ (~ L' ~J ~ 7 ~, PCr/AU90/OOOs2 Tablc ~ Effcct of b~ c~nccntratioDs of ~L~ OD pH}tl~ymidiDc iDcolporatioD aDd 4~Ca rclcasc in nconata1 mo~c a~

pHl-Tbymidinc 45Ca Rclcasc (units/ml) Incorporation (trca~ncnt/control) (treatmcnt./control) 100 131 Q12' 1.0110.07 300 1.1720.0P 1 04~0 06 1000 138~0.08~ 1.05~Q07 Data are mcan 2 vam and ue giVCD as b catmcntlcontrol rab~ for case of compar~ono the two fndice~. Significant diffcrcnccs fmm control uc sho~vn: a, pc0.05; b, pcO.01.
4 in cach group. Sim~ar rcsults wcrc obtained in ~vo othcr c~pc~Dents.

2 0 ~ ~ 6 7 ~ PCT/AU~/0NWn prosta~landin E, w~s purchased from Up~ohn Company, Kalamazoo, Michlgan. ~ecomblnant human L~F was produced in both yeast and E.coll and purlfled to homogenelty as descrlbed. ~he speclflc blologlcal actlvlty of pure LIF ls 5 -10~ unlts/mg. Antlserum agalnst TGF~ wa3 purchased from R
and D Systems, Inc. Mlnneapolls, M.D. 55413 U.S.A.

Rabblt antlserum agalnst PAI-2, prepared form U937 cells, was by courtesy of Professor E.K.O. Krulthof, Centre 10 Hospltaller Unlversltalre, Vaudols, Lausanne, Swltzerland.
Rabbit anti-rat hepatoma cell (H~C)PAI-l (S) and rat PAI-l cDNA cloned lnto the EcoRl slte of pBluescrlpt SX(-)(6) were klndly provlded by Drs. R. Zeheb and T.D. Gelehrter, University of Mlchlgan, Ann Arbor, MI. ~he cDNA probe for 15 TGF~ was provlded by Genentech, San Franclsco. The cDNA
probe for murlne LIF was the plasmid pLIF7.2b whlch spans the codlng reglon of the murlne LIF m~NA. ~ flbrlnogen, t~''P]UTP~ [~,2P]CTP and a T7/SP6 palred promoter system were from Amersham Internatlonal, T,RNA polymerase was from 20 International Blotechnologles Inc., New Haven, CT, 06535, U.S.A. and nlck-translatlon klts from Boehrlnger Mannhelm, Mannhelm, West Germany. Sodlum dodecyl sulphate polyacrylamlde gel electrophoresls (SDS-PAGE) M, standards were from 81O-Rad Laboratorles, Rlchmond, CA 94804, U.S.A.
(11) Osetoblast Systems The UMR 106-01 and UMR106-06 cells are a subclone of the VMR106 cells whlch are a clonal llne of rat osteogenic sarcoma cells ln whlch many features of the 30 osteoblast phenotype are preserved (1, 8, 9). The experlments descrlbed were carrled out on cells between the 7th and 15th passages and malntalned ln Eagles Mlnlmal Essential Medlum (MEM) supplemented wlth non-essentlal amlno aclds, 14mM 4-(2-Hydroxyethyl)-l-plperazinoethanesulfonlc 35 acid (Hepes), gentamycln (80~g/ml), mlnocycllnee(l mg/ml) and 5~ (V/V~ fetal cal serum (FCS) at 37C in 5% CO, ln alr.

wO 90/l~ 2 ~ r~ 1 PCT/AU90/~2 The UMR201 clonal cell line, whlch was derlved from nerborn rat calvaria (10), has propertles conslstent wlth those of an osteoblast precursor. Treatment wlth retlno~c acid leads to substantial increase ~n alkallne phosphatase 5 activity and mRNA, and the cells produce predomlnantly type I collagen and have other osteoblastlc eatures (10, 11).

Osteoblast-rich calvarial cell~ were prepared as descrlbed prevlously (7, 8) and were used on the first 10 subculture. Conditions of culture maintenance for UMR201 and osteoblast-rich cells were as described for osteogenic sarcoma cells.

Osteoclasts were isolated from newborn long rat 15 bones as previously described (24), and allowed to settle onto glass coverslips for perlods of greater than 30 mins before washing, to ensure signiflcant contamination with osteoblasts (25).

(iii) RNA PreDaration and HYbridizatlon (Figures 11 and 12) Cells were washed twice with phosphate buffered sallne (PBS) and preincubated in minimal essential medium 25 (MEM) and 0.1% w/v bovine serum albumln (BSA) for 15 h. At the end of thls period the medium was replaced wlth MEM and 0.1~ w/v BSA with and without added LIF. The cells were incubated ~or varying times and concentrations. ~otal RNA
was prepared as previously described (12). An aliquot of lC
30 ~g of total RNA was electrophoresed on a 1.5% w/v agarose/formaldehyde gel as prevlously described (12) and transferred to Hybond N (Amersham International) by Northern blotting. Labelled riboprobes for PAI-1 were prepared from rst PAI-1 cDNA cloned lnto the ~coRI site of Bluescript 35 SK(-) using T3 ~NA polymerase. ~ybridization conditions wexe essentially as aescribed (12) in 50~ v/v formamide, 6 x SSC, 5 x Denhardt's, 0.1~ w/v SDS at 65-C. Filters were wo ~/.~ 2 0 '~ ~ ~ 7 ~ PCT/AU~/~Wn washed to a strlngency of 0.1 x SSC at 65'C and exposed to Kodak XAR-5 X-ray fllm backed by intenslfylng screens.
After removlng DNA bound to the RNA by boillng for 5 min in 0.1 x SSC, 0.1% w/v SDS fllters were probed a second time 5 wlth a nlck-translated full-length rat actln cDNA probe.
The filters were washed to a stringency of 0.1 x SSC, 0.1~
w/v SDS at 42-C and exposed as above. Slmllar hybridlzation conditions were used in detecting mRNA for osteonectin and type I collagen. In each case the probes were labelled by lO nick translatlon.

(IV) Plasminogen Activator (PAl Activitv and Alkaline Phos~hatase Cells were subcultured at 20,000/200~1 into '25I-fibr~n-coated wells of 96 well tissue culture plates (13).
After 24 hours ln maintenance medlum the cells were washed twlce wlth Dulbecco's PBS and the medlum was replaced wlth MEM and 0.1~ w/v BSA. After 8 hours the medium was replaced 20 with fresh Eagle' 8 MEM and 0.1~ w/v BSA together with human plasmlnogen tl.56 ug/ml). Pla~mlnogen-independent activity was measured in parallel in the absence of added human plasminogen. After 24 hours an aliquot of medium was removed and counted for solubilized radioactivity.
25 Plasminogen dependent flbrinolytlc activity i8 expressed a~ -percentage o total radioactivlty released from the plate by trypsln.

For studles of regulation of alkallne phosphatase 30 activity, UMR 201 cells were subcultured into 9.6cm' six-plate multiwell dishes in alpha modified MEM (a-MEM) contalning 10~ v/v FCS. When the cells were seml-confluent, the medium was changed to a-MEM with 2~ v/v vitamin A-depleted PCS containing the indicated concentrations of LIF
35 plu8 l~M retinoic acld. Medium was changed after 48 h and fresh treatments added. Speclfic actlvlty of alkaline phosphatAse was measured as described prevlously (lO).

wo go/1~54 ~ 6 7 ~ PCT/AU90/nH~ l .. , (v) L~F ReceDtor Studies UMR 106-01 and 106-06 cells were subcultured, 5 centrlfuged ln Hanks' BSS and kept on lce after whlch cells were resuspended and layered over 180 ~1 fetal calf serum (FCS) ln tapered flexlble plastlc tubes. Cell-assoclated and free '2sI-LIF were separated by centrifugatlon of the cells through FCS. The tlp of the tube contalnlng the cell 10 pellet was cut off and the pellet and supernatant were counted ln a~-counter (Packard Crystal Multl-Detecter, Packard, Downers Grove, IL). Saturatlon lsotherms were analyzed by the method of Scatchard (14) uslng the non-linear curve flttlng programme Llgand (15).
Cells lsolated from newborn rat long bones were allowed to adhere to glass coverslips withln the wells of a 12-well tlssue culture plate (Flow Laboratorles, McLean VA), and were covered wlth 1.0 ml KHF containing 5xlOa cpm '~I-LIF
20 wlth or wlthout a 40-fold excess of unlabelled LIF.
Incubatlon was carrled out at 37C for 30 mlnutes. The medium was removed and the cells were washed wlth 4x4 ml aliquots of warm KHF and 2x4 ml allquots of warm 20mM
sodlum phosphate, 0~15 ml NaCl at pH 7.4. ~he cells were 25 fixed ln 4 ml o 2.5~ u/v glutaraldehyde in PBS and washed wlth 2x4 ml allquots of water. ~he coverslips were attached to gelatln-coated glass slldes uslng DePex mountlng medium, allowed to alr-dry overnight, dlpped in Kodak NTB2 photographic emulslon at 42C ln a dark room and allowed to 30 dry. Slides were then sealed in light proof boxes containing Drierlte and allowed to expose for 4 weeks at 4C. Prior to development, slldes were warmed to room temperature and then developed for 1 min in Kodak Dl9 developer (40g/500 ml water) and fixed in Agfa G333c X-ray 35 fixer for 3 mln.

. .

, wo 90"~ 2 0 ', 3 ~ 7 ~ PCT/AU~/OWK~

(vi) Fibrln and Reverse Flbrin AutoqraDhy Cells were subcultured at 500,000 cells/well (9.6 5 cm') ln MEM and 5% v/v FCS, After 25 h the medium was asplrated, cells washed twlce wlth PBS and the medium replaced wlth 1 ml MEM and 0.1~ w/v ~SA. After 8 h the medlum was asplrated and 1 ml of fresh MEM and 0.1% w/v BSA
added together wlth treatment. At 25 h the medlum was 10 asplrated and centrifuged to remove any cell debris. The cells were washed twice wlth PBS and lysed ln lml. 0.1~ v/v trlton X 100. Conditioned medlum (CM) and cell extracts, prepared as descrlbed abovet were resolved by SDS-PAGE on 10~ w/v slab gels under non-reduclng conditlons as descrlbed (16). PA and PAI activlty were visualized by fibrin (17) and reverse flbrin (18) gel autograph, respectively.

(vli) Immunoblottinq Conditloned medla from confluent cultures of UMR
106-01 cells and calvarial osteoblasts lncubated for 24 h in serum-free MEM were concentrated by ultraflltration with a 30,000 Mr cutoff, while cells were washed twlce wlth P8S, scraped, centrlfuged and lysed ln 0.1~ v/v trlton X 100.
25 The protelns were resolved by SDS-PAGE as above and transferred to nltrocellulose. PAI-l was de~ected by lncubatlon of the nltrocellulose with rabbit polyclonal antl-rat HTC PAI-l (5) followed by antl-rabbit Ig con~ugated wlth alkallne phosphatase and visualization with 5' bromo-4-30 chloro-3-lndolyl phosphate (19).

(vlli)RNA Detection Usin~ the PolYmerase Chain Reaction (Figure 15) ~otal cytoplasmic RNA was isolated from UMR 201 or UMR 106-01 cells or primary osteoblast populatlons ~see above) a~ described 126) and sub~ected to first strand cDNA

wo go/104~4 ~ 3 ~ 7 ~ PCT/AU90/~92 synthesls ln a 20 ~1 reactlon contalnlng 50 mM ~rls-Cl (pH
~.3 at 42C), 20 mM KCl, lOmM MgC12, 5 mM dlthlothretlol, 1 mM of each dNTP, 20 ~g/ml ollgo-DT,s and LIF speclflc oligonucleotlde prlmers, and 20 units AMV reverse 5 transcriptase (80ehringer Mannhelm) for 40 mln at 42C.
After completion of flrst-strand synthesls, the reactlon was diluted to 100 ~1 wlth distllled water und 5 ~1 used to each PCR reactlon. PCR reactions (in a volume of 50 ~1) contained 200 ~M of each dN~P, 1 ~m of each speclflc prlmer, 10 buffer as suppl~ed ln the GeneAmp kit (Cetus Corp., USA) and 1.25 units Taq polymerase. The prlmers used for PCR
reactlons were: LIF, 5' - CCGAATTCGAAAACGGCCTGCATCTAAGG and 3' - CCGAATTCGCCATTGAGCTGTCCAG~TG, deflnlng a 300 bp fragment of mRNA, and GAPDH, 5' - ACCACCATGGAGAAGGCTGC, 15 GAPDH, 5' - ACCACCATGGAGAAGGCTGC and 3'-CTCAGTGTAGCCCAGGATGC, defining a 500 bp fragment. The PCR
reactlon conditlons were: 1.5 mln at 94C; 2 min at 60C; 3 mln at 72C for 25 cycles ln a Perkin-Elmer-Cetus DNA
Thermal Cycler. A portlon of the PCR reactlon was 20 electrophoresed through a 1.2~ w/v agarose gel and transferred to nltrocellulose. Fllters were prehybridized, hybridized and washed ln 2 x SSC-contalning buffers as described (21). The hybridlzatlon probes were the gel-purlfled cDNA inserts of plasmlds pLIF7.2b and hGAP-DH, 25 radlolabeled to a speclfic actlvity of approximately 2xlO~
cpm/ug by nick translatlon and included in the hybridizatlon at approxlmately 2xlO'cpm/ml.

2. R~SULTS
(i) Effect of LIF on bone resorDtlon In the PA response lt was found that LIF lnhibited PA activity in osteoblast-like cells in a dose dependent manner. The inhibltions was apparent ln control cells, or ln 35 cell~ which had been treated with an agent which increases PA activlty, e.g. parathyroid hormone (P~H). An example of ~uch an experiment ls provided in Figure 10, showing wo ~1~54 2 0 ~ ~ 6 7 ~ PCT~AU90/00092 inhlbitlon of PA actlvlty in the UMR106-01 cells. Simllar lnhlbltlon was demonstrated ln the calvarlal osteoblasts.
These experlments are not posslble ln the UMR201 cells, whlch do not survlve sufflclently long enough ln culture ln 5 the absence of serum.

Because the effect of LIF to lnhlblt PA actlvlty mlght be due to an increase ln the amount of a PA lnhibitor, the possible effect of LIF on the formatlon of PAI-l, the PA
10 inhibltor produced by osteoblasts, was examined. Evidence based on reverse flbrin autography, western blotting wlth a specific antiserum to PAI-l, and demonstratlon of PAI-l mRNA
by northern blots of calvarial osteoblasts and UMR106 cells showed thls to be the rase. When UMR 106 cells were treated 15 with LIF, there was an increase in mRNA for PAI-l, consistent with the possibility that the inhibition of PA
activlty of LIF was due to the generation of increased amounts of PAI-l. Regulation of PA actlvity by inhlbltor formation is well recognised as an important factor in 20 determining nett PA activity. Figure 11 shows a time course up to 24 hours of the PAI-l mRNA with response to LIF. AS
early as 1 hour after the beginning of LIF treatment there was a clear lncrease in PAI-l mRNA.

This effect of LIF on inhibition of PA activity and promotion of PAI-l mRNA i~ very similar to the effect of transforming growth factor B (TGFB) in the same cells. To exclude the possibility that LIF might be acting through the generation of TGFB in the cells, two experiments were 30 carried out. In the first, an antibody against TGFB was used to determine whether it blocked the action of LIF. No blocking action was found (Table 5). The same antibody was able to block the action of TGFB.

The second experiment was to determine whether LIF
lnfluenced mRNA for TGFB in the osteoblast-like cells. In the UMR106 cells and in calvarial osteoblasts, no effect of wo 90~1WS4 ~ O ~ v ~v 7 ~ PCT/AU~/~
! . . ', 31 LIF on TGFB mRNA was found. It 18 concluded that the LIF
actlon of PAI~l was not medlated ln the osteoblast-llke cells by any generatlon of TGFB as an lntermedlate ln lts actlon.

TABL~ 8ff~ct of ~ntl hu~an TGF ant~body (ab) on PA Actlvlty ln LIF UMR106-01 cells.

1 0 ~
ExDt. 1 PA Activity (~ TcDm) Control39.7 + 1.4 LIF (100 u/ml) 20.9 + 0.8 TGF~ ab (30 ug/ml) 43.5 + 0.4 LIF + TGF~ ab 23.1 + 0.7 ExDt. 2 Control20.8 + O.3 PTH (10 ng/ml) 46.0 + 1.1 TGFB (0.5 ng/ml) 3.2 + 0.1 PTH ~ TGFB 12.0 + 0,7 PTH + TGFB + ab (30 ug/ml) 51.7 + 4.7 PTH + TGFB ~ TGFB ab 38.6 + 2.2 LIF treatment of UMR201 cells resulted ln a dose-dependent increase ln mRNA for both type I collagen and osteonectin.
Northern blots to demonstrate thls are provlded in Flgure 30 12. Both these effects are also demonstrated by TGFB in UMR
201 cell3. However, not all TGFB-llke effects are demonstrat~d by LIF, even ln these cells. For example, whereas ~GFB attenuates the rlse ln alkaline phosphatase actlvity and m~NA ln response to retlnolc acid ln these 35 cells, the reverse takes place wlth LIF. LIF actually enhances the alkaline phosphatase response, whlch ls an 2 ~ ~ ~ 6 7 ~ PCr/Augo/00092 ,. . .

effect very slmllar to that of another cyto~lne, tumour necrosls facto~ a ( TNFa ) .

The effect of LIF on the osteoblast-llke cells ls 5 largely anabolic, in that there is clear evldence of a LIF
effect to lncrease cynthesis of some of the lmportant speclflc components of bone proteln, and a regulatory effect on the PA-PA lnhlbitor system, whlch glve LIF the potentlal of belng an important regulator ln the mlcroenvlronment of 10 bone cells, and therefore lmportant in bone remodelling.

The effect of LIF to lncrease bone resorption in mouse calvariae is one which is dependent upon the synthesls withln those bones ln organ culture of prostaglandins. The 15 evldence for thls is that blocklng prostaglandin synthesis inhlblts the LI~ stlmulatlon of resorption. On the other hand, the LIF stlmulatlon of 3H-thymldlne into acid-precipltable macromolecules was not lnhlbited hy indomethacin (Example 4). T~FB al~o promotes re~orption ~n 20 the mouse calvarlal organ culture system, but lt is inhlbited when prostaglandln synthesls ls lnhibited, however, TGFB is not a promoter of resorption in a system uslng fetal rat long bones. The mouse calvarisl system 18 one in which the generatlon of prostaglandlns is vlrtually 25 certain to result ln lncreased resorption. How~ver, TGFB
has lts predominant effect on promoting formatlon of bone, as does LIF. In other types of experimental sy~tems there is good evidence that prostaglandins in fact stimulate bone formation.
(ii) ReceDtor Studies Receptor autoradlography with '25I-LIF was carried out to determine whlch cells of newborn rodent bone bound 35 LIF speclfically. Osteoclast preparations from newborn rat lon~ ~One8 were prepared 80 as to ensure that the cultures were heavily cont mlnated with osteobla8t8, macrophages and WO ~/l~ ~ ~ . PCT/AU90/~W~2 other cells. Thls was achleved by allowlng a long settllng tlme onto coversllps after lsolatlon of cells (Z5). In these experlmentæ, speclflc blndlng of '2sI-LIF was demonstrated on cells wlth the characterlstlcs of 5 osteoblasts (Figure 14) and to macrophages as the lnventors have shown prevlously (22), but no evldence of LIF blnding to osteoclasts or to mononuclear tartrate-resistant acld phosphatase-posltive cells (osteoclast precursors) could be found.
In view of the identiflcation of LIF receptors on prlmary osteoblasts, biochemica~ studies of LIF binding were carrled out on the osteoblast-like, clonal osteogenic sarcoma cell llne, UMR106-06. Speciflc blnding was readily demonstrated, 15 and Scatchard analysis indicated a receptor number of 300 per cell, and X~ of 60 pM (Flg. 14). The receptor number compares favourably with the 400-~00 per cell which had been previously demonstrated on the murine myeloid leukemia cell line, Ml (20). Specific binding was also demonstrated on 20 UMR 201 cells, although at lower receptor numbers.

In vlew of the growing reallzatlon that locally produced factors are lmportant ln regulatlng bone metabolism (7), and the demonstration of GM-CSF production by osteoblasts (23) 25 evidence was sought for production of LIF by osteoblast-like cellæ.

(iii) Production of LIF by osteoblasts 30 Biological assays revealed the presence of very low levels of Ml cell differentiating activity in medium conditioned by primary osteoblast populations and by the ostesblastic cell llnes UMR 201 and 106-01. In order to confirm that this actlvlty was due to LIF, these cells were assayed for LIF
35 tran8crlpt8. LIF transcrlpts were not detected ln RNA from these sources using conventional Northern blot~ (not shown), pos8ibly because the amounts of LIF formed were very low.

wo ~"~ 2 ~ ~ Q ~ 7 ~ PCT/AU90/0~92 f LIF transcrlpts were, however, demonstrable ln these cells by uslng a sens~tlve RNA detection approach based on the polymerase chain re~ctlon. It appears from such analyses (eg Figure 15) that LIF ls transcribed constltutlvely ln 5 both the UMR 201 and UMR 106-01 cells, as well as ln prlmary osteoblast populatlonR, although ln some experlments -modulation of the level ln UMR 106-01 was apparent on treatment with T~FB (Figure 15). However, glven the posslble dlfflcultles ln quantlfylng PCR reactlons lt cannot 10 be dermined with certainty the level of inductlon ln response to TGFB. It was consistently found that UMR 201 cells contained higher levels of LIF transcripts than UMR
106-01 cells.

(lv) Effect of LIF on alkallne ~hosDhatase actlvlty in UMR201 cells Treatment of UMR 201 cells with LIF for 3 days resulted in dose-dependent potentlation of the retlnolc acld-lnduced 20 alkallne phosphatase activity (Table 6). Thls effect 18 very similar to that we have prevlously obtalned with TNFa in the same cells (11). In the latter work evidence was obtained that the effect was due to stabilization of mRNA
for alkallne phosphatase. Thus, although LIF and TN~a 25 resemble each other ln thls actlon, they have effects in opposlng dlrections on PA actlvity in osteoblasts (Table 7).

35 2~'~g~7~
~ 90/10454 pcr/Augo/ooo92 ~b~ 6 Eft~c~ Or LIF on P~ ~ct;~ in Control nd Sltmubted UMR.Ig6.0 c~lls. .
P~ ~c~ivi~ (% T epm) Tr~tmcnl -LIF +LlF(lOOOU/ml) t~ lO.S~0.3 2.0+0.1 PIH (3.3 x l~9M) 42.1 + 1.1 14.7 +1.0 PGE2 (1~7M) 22.0 ~ 1.6 3.5 + 0.2 1,2S (OH)2 D3 (10-8M) 22.3 ~ 0.2 5.S + 01 TNFa aoo U/ml) 2S.3 + Q3 ~.6 ~ 0.1 (SOOOU/!~nl) 32.4 ~ I S S3+Q3 Cclls wcrc subcullurcd on to 12S-I-fibnn-coatcd pbtcs as dcsc ibcd in thc Mclhods scction. Cclls wcrc prcincubated for 8 h in MEM plus 0.1~ BSA bcforc thc addition of frcsh mcdium containing thc bonc rcsorbing factors. Thc incubation w~s c nicd ouî for 24 h.

wo ~/1~ 2 0 '~ ~ 6 ~ ~ PCT/AU90/~X~2 ~S:
1) Horton, J. E., Ralsa, L.G., Slmmons, H.A., Oppenhelm, J.J., & Meyerhager, S.E. Sclence, 177:793-795, 1972.

5 2) Martln, T.J., Ng, K.W., & Nlcholson, G.C. Clln. Endocr.
Metab. 2:1-29, 1988.

3) Canalls, E. Endocrlnol. 118:78-81, 1986.

10 4) Reld, I.R., Katz, J., Ibbertson, H.K., & Gray, D.H.
Clacid. Tlssue Int. 38:38-43, 1986.

5) Zeheb, R., Rafferty, U.M., Rodriguez, M.A., Andreasen, P., & Gelehrter, T.D. Thromb. h Haemost. 58:1017-1022, 15 1987.

6) Zeheb, R., Gelehrter, T.D. Gene 73:459~468, 1988.

7) Martin, T.M., Ng, K.W., Partridge, N.D. & Livesey, S.A.
20 Methods ln Enzvmol. 185: 324-336, 1987.

8) Partridge, N.C., Alcorn, D., Michelangeli, V.P., Ryan, G., & Martln, T.J. Cancer Res., 43:4388-4394, 1983.

25 9) Forrest, S.M., Ng, K.W., Findlay, D.M., Mlchelangeli, V.P., Llvesey, S.A., Partridge, N.C., ZaJac, J.D., h Martin, T.J. Calc. Tiss. Int. 37:51-56, 1985.

10) Ng, K.W., Gummer, R.R., Michelangeli, V.P., Bateman, 30 J.F., Mascara, T., Cole, W.G., & Martin, T.J., J. Bone. Min.
Res. 3:53-61, l9B8.

11) Ng, K.W., Hudson, P.J., Power, B.E., ManJi, S.S., Gummer, P.R., & Martin, T.J., J. Mol. Endocrlnol., 3:57-64, 35 1989.

wo go/l04~ ~ ~ ~ v {; 7 ~ Pcr/Augo/0oo92 12) Manlatls, ~., Frltsch, E.F. & Sambrook, J. Molecular clonlna: a Laboratorv Manual, Cold sprlng Harbour Laboratory, Cold Sprlng Harbour, N.Y., USA, 1982.

5 13) Allen, E.H., Hamilton, J.A., Metcalf, D.C., Kabota, M.
& Martln, T.J. Blochlm. Blo~hys. Acta 888: 199-207, 1986.

14) Scatchard, G. Ann. NY. Acad. Scl. 51: 600-672, 1949.

10 15) Munson, P.J., & Rodbard, D. Anal. Biochem. 107:220-237, 1980.

16) Laemmll ( 1970) .

lS 17) Granelll-Plperlno, A., & Relch, E. J. Ex~. Med.
148: 223-234, 1978.

18) Erlckson, L.A., Lawrence, P.A., & Loskutoff, D/J. Anal.
Blochem. 137:454-463, 1984.

19) Knecht, D.A., & Dimond, R.L. Anal. Biochem. 136: 180-184, 1984.

20) Hllton, D.J., Nlcola, N.A., & Metcalf, D. Proc. Natl.
25 Acad. Scl. (USA) 85:5971-5975, 1988.

21) Gearlng, D.P., King, J.A., Gough, N.M. & Nicola, N.A.
EMB0 J. 8: 3667-3676, 1989.

30 22) Hilton, D.J., Nlcola, N.A. & Metcalf, D. J. Cell.
Ph~slol. (In Press), 1990.

23) Felix, R., Elford, P.R., Stoerile, D., Cecchin, M., Wetterward, A., ~rechsel, U., Flelsch, H. & Shadler, B.M. J.
35 Bone_Mln. R~8. 3:27-32, 1988.

wo 90/l~ 2 0 ~ 8 6 7 ~ PCT/AU90/00092 24) Nlcholson, G.C., Moseley, J.M., Sexton, P.M., Mendelsohn, F.A.O. & Martln, T.J. J. Clln. Invest. 78: 355-360, 1986.

5 25) McSheehy, P.M.J. & Chambers, T.J. Endocrlnology 118:
824-828, 1986.

26) Gough, N.M. Anal. Blo~hem. 173 : 93-95, 1988.

10 27) Allen, E.H., D.J. ~llton, M.A. Brown et al. J. Cellular PhYsioloov (submitted for publlcatlon), 1990.

Claims (36)

THE CLAIMS DEFINING THE INVENTION ARE AS FOLLOWS:

1. A method for inducing, promoting and/or enhancing metabolic effects on bone in an animal comprising i administering to said animal an effective amount of leukaemia inhibitory factor (LIF) and/or LIF-like polypeptides for a time and under conditions sufficient to effect bone metabolism.

2. The method according to claim 1 wherein the animal is mammalian or avian.

3. The method according to claim 2 wherein the animal is mammalian.

4. The method according to claim 3 wherein the mammal is a human.

5. The method according to any one of claims 1 to 4 wherein the metabolic effect comprises bone formation and/or bone resorption, increased protein synthesis in bone tissue and/or thymidine uptake in bone tissue.

6. The method according to any one of claims 1 to 5 optionally further comprising the administration of one or more active molecules in combination with LIF and/or LIF-like polypeptides.

7. The method according to claim 6 wherein the active molecule is another cytokine.

8. A pharmaceutical composition useful in inducing, promoting and/or enhancing metabolic effects on bone in an animal comprising an effective amount of leukaemia inhibitory factor (LIF) and/or LIF-like polypeptides and one or more active molecules and one or more pharmaceutically acceptable carriers and/or diluents.

9. The pharmaceutical composition according to claim 8 wherein the animal is mammalian or avian.

10. The pharmaceutical composition according to claim 9 wherein the animal is mammalian.

11. The composition according to claim 10 wherein the mammal is a human.

12. The composition according to claim 8 wherein the active molecule is a cytokine.

13. The composition according to claim 8 wherein the active molecule is an inorganic or organic molecule.

14. The use of leukaemia inhibitory factor (LIF) and/or LIF-like polypeptides in the manufacture of a medicament for the induction, promotion and/or enhancement of metabolic effects on bone in animals.

15. The use according to claim 14 wherein the animal is avian or mammalian.

16. The use according to claim 15 wherein the animal is mammalian.

17. The use according to claim 15 wherein the mammal is human.

18. The use according to any one of claims 14 to 17 wherein the metabolic effect comprises bone formation and/or bone resorption, increased protein synthesis in bone tissue and/or increased thymidine uptake in bone tissue.

19. A method for the treatment of injuries and/or diseases which result in bone degeneration or fracture in an animal said method comprising the administration to said animal of a bone forming - increasing - effective amount of leukaemia inhibitory factor (LIF) and/or LIF-like polypeptides for a time and under conditions sufficient to effect bone formation, and/or bone fracture repair.

20. The method according to claim 19 wherein said animal is a human.

21. The method according to claim 19 or 20 wherein said injury is a bone fracture.

22. The method according to claim 19 or 20 wherein said disease is post-menopausal osteoporosis.

23. The method according to any one of claims 19 to 22 optionally further comprising administering one or more active molecule in combinationn with LIF and/or LIF-like molecules.

24. The method according to claim 23 wherein the active molecule is a cytokine.

25. The method according to claim 23 wherein the active molecule is an inorganic or organic molecule.

26. A method for stimulating bone formation in a mammal comprising administering to said human a bone forming-increasing-effective amount of leukaemia inhibitory factor (LIF) and/or LIF-like polypeptides for a time and under conditions sufficient to effect bone formation.

27. The method according to claim 26 wherein said mammal is a human.

28. A method for stimulating bone fracture repair in a mammal comprising administering to said mammal a bone-forming effective amount of leukaemia inhibitory factor (LIF) and/or LIF-like polypeptides for a time and under i conditions sufficient to effect bone formation.

29. The method according to claim 28 wherein said mammal is human.

30. The use of leukaemia inhibitory factor (LIF) and/or LIF-like polypeptides in the manufacutre of a medicament for treating injuries and/or diseases in animals where bone degeneration or fracture is a symptom or a result.

31. The use according to claim 30 wherein the animal is avian or mammalian.

32. The use according to claim 31 wherein the animal is mammalian.

33. The use according to claim 32 wherein the animal is human.

34. The use according to any one of claim 30 to 33 wherein the injury is bone fracture.

35. The use according to any one of claims 30 to 33 wherein the disease is osteoporosis.

36. The use according to claim 35 wherein the disease is post-menopausal osteoporosis.