Zinc finger protein 521, a new player in bone formation

Ann. N.Y. Acad. Sci. ISSN 0077-8923

ANNALS OF THE NEW YORK ACADEMY OF SCIENCESIssue: Skeletal Biology and Medicine

Zinc finger protein 521, a new player in bone formation

Eric Hesse,1,∗ Riku Kiviranta,1,∗ Meilin Wu,1,∗ Hiroaki Saito,1 Kei Yamana,1 Diego Correa,1

Azeddine Atfi,1 and Roland Baron1

1Department of Medicine, Harvard Medical School and Department of Oral Medicine, Infection and Immunity, Harvard Schoolof Dental Medicine, Boston, Massachusetts, USA

Address for correspondence: Roland Baron, Department of Medicine, Harvard Medical School and Department of OralMedicine, Infection and Immunity, Harvard School of Dental Medicine, 188 Longwood Ave., REB310, Boston, MA [email protected]

Exploration of anabolic pathways in osteoblasts revealed that Zfp521, a 30-zinc finger protein, is highly expressedat the periphery of mesenchymal condensations and in developing bones. In these structures it is expressed inchondroblasts, prehypertrophic chondrocytes, the periosteum, osteoblasts, osteoblast precursors, and osteocytes.Forced expression of Zfp521 in osteoblasts in vivo increases bone formation and bone mass, whereas preliminarydata suggest that germline deletion leads to osteopenia. In contrast, overexpressing Zfp521 in vitro antagonizes,and knockdown favors, osteoblast differentiation and nodule formation. Zfp521 expression is inhibited by bonemorphogenetic protein-2 and stimulated by parathyroid hormone-related protein. Mechanistically, Zfp521 bindsto Runx2, repressing its transcriptional activity. These data support the hypothesis that Zfp521 both opposes theprogression of precursors and promotes the maturation and function of mature osteoblasts. The balance betweenZfp521 and Runx2 may therefore contribute to the regulation of osteoblast differentiation and bone formation.

Keywords: Zfp521; Runx2; osteoblast; bone formation

Introduction

Development and maintenance of the skeleton de-pends on the synchronized actions of bone-formingosteoblasts and bone-resorbing osteoclasts.1,2 Os-teoblasts arise from mesenchymal stem cells thatare able to differentiate into a number of special-ized cell types, including osteoblasts, adipocytes,chondrocytes, and myoblasts.3,4 Downstream of keyosteogenic signaling pathways, such as bone mor-phogenetic protein (BMP) and Wnt, several tran-scription factors have been identified as media-tors and/or targets of these signaling pathways thatregulate specific steps in the progressive transitionfrom early mesenchymal cells to the fully differen-tiated and bone matrix-producing osteoblast.5 Theappropriate timing of the expression of these keytranscription factors is critical to ensure the propercoordination of genes that regulate both the prolif-

∗These authors contributed equally to this work.

eration and progression of osteoblast precursors tomature bone-forming cells, while maintaining thepool of precursors. Among these transcription fac-tors, Runx2, a member of the Runt domain family oftranscription factors, is an essential regulator of os-teoblast commitment and early stages of osteoblastdifferentiation.6–8 However, and in contrast to thefavorable influence of Runx2 on early osteoblast dif-ferentiation, in vivo overexpression of Runx2 at latestages of osteoblast differentiation causes a block ofosteoblast maturation, thereby leading to an accu-mulation of immature osteoblasts, osteopenia, andspontaneous midshaft fractures of long bones inmice.9–11

Similar stage-dependent effects have been re-ported for canonical Wnt activity. For example, ithas been shown that an increased expression ofDkk1 and/or Dkk2 and the concomitant attenua-tion of canonical Wnt signaling is a prerequisite forlate-stage osteoblast differentiation.12 In addition,another group reported independently that over-expression of the Wnt signaling inhibitor Dkk2 in

doi: 10.1111/j.1749-6632.2009.05347.x32 Ann. N.Y. Acad. Sci. 1192 (2010) 32–37 c© 2010 New York Academy of Sciences.

Hesse et al. Zinc finger protein 521

Figure 1. Schematic of the Zfp521 molecular structure. Zfp521 is a 180-kDa protein with 30 C2H2 Kruppel-like zincfinger motifs distributed in clusters.

mature osteoblasts stimulates osteoblast differenti-ation and matrix mineralization.13 Consistent withthese findings, Rodda and colleagues reported thatcontinuous activation of the canonical Wnt sig-naling pathway arrests osteoblasts in an immaturestage, suggesting that the progression of osteoblaststo a terminally differentiated stage requires the ces-sation of Wnt/�-catenin signaling.14 These data sug-gest that despite the positive effects of Runx2 andcanonical Wnt signaling on osteoblast lineage com-mitment and early stage osteoblast function, bothneed to be attenuated to allow for terminal os-teoblast differentiation. Thus, while some molecularcues are necessary at specific stages to promote theprogression of cells toward their fully mature stage,the repression of their expression and/or activityby other factors can be equally important at earlierstages of differentiation to prevent the depletion ofthe precursor pool and/or at later stages to allow forfull cellular maturation and function.

Zfp521 (also termed Evi3 in mice and EHZF inhumans) plays such a role in some cell systems.15,16

Zfp521 is a zinc finger protein that consists al-most entirely of 30 C2H2 Kruppel-like zinc fingers(Fig. 1). Indeed, both Zfp521 and Zfp423 (an-other 30-zinc finger protein, also termed olfactory[OLF]/early B-cell factor [Ebf]-associated zinc fin-ger protein, with high domain and sequence homol-ogy to Zfp521) inhibit cell differentiation in somecell lineages.15,17–19 Furthermore, both Zfp521 andZfp423 inhibit the transcriptional activity of Ebfproteins,15 which have recently been linked to os-teogenesis.20,21 Here we describe the role of Zfp521in controlling early osteoblast differentiation andbone formation in vivo.

Zfp521 is a nuclear protein expressed inchondrocytes and osteoblasts

We have found22 that Zfp521 is expressed in manycell types and at high levels in calvarial osteoblastsas well as in C3H10T1/2 and MC3T3-E1 mesenchy-mal cell lines. Zfp521 is most highly enriched in

the nuclei. In the growth plate, Zfp521 is expressedat different intensities in all layers of chondrocytesbut at the highest levels in prehypertrophic cells.Furthermore, Zfp521 is expressed at all stages ofosteoblast differentiation, including preosteoblasts,osteoblasts, and osteocytes. The expression patternand the nuclear localization suggest that Zfp521plays a role in the regulation of gene transcriptionduring osteoblast differentiation.

Zfp521 promotes bone formation

To test the hypothesis that Zfp521 regulates os-teoblast differentiation and function, we gener-ated mouse models of gain of function and havebegun analyzing a model with germline loss ofZfp521 function. Mice in which overexpressionof Zfp521 was targeted to the osteoblasts us-ing the osteocalcin promoter (Zfp521-TG) exhib-ited a high bone-mass phenotype (Fig. 2). His-tomorphometric analysis of tibiae from 4- and12-week-old Zfp521-TG mice revealed thatosteoblast-targeted overexpression of Zfp521 ledto an increase in bone mass due to a marked in-crease in osteoblast number and bone-forming ac-tivity.22 In addition, our preliminary data suggestthat germline deletion of Zfp521 leads to osteope-nia (Fig. 2). These data show that Zfp521 has a cell-autonomous role in osteoblasts where it is essen-tial for osteoblast differentiation and bone matrixformation.

Zfp521 restricts early stages of osteoblastdifferentiation but promotes late-stageosteoblast maturation

To better understand the bone phenotypes ofZfp521 gain- and loss-of-function mice, we exam-ined the ex vivo differentiation of calvarial cellsfrom Zfp521-TG mice. The Zfp521-TG cultureshad less nodule formation and mineralization thanthe control cultures. Moreover, expression of sev-eral early osteoblast marker genes and Runx2 target

Ann. N.Y. Acad. Sci. 1192 (2010) 32–37 c© 2010 New York Academy of Sciences. 33

Zinc finger protein 521 Hesse et al.

Figure 2. Histology of osteoblast-targeted gain- (Zfp521-TG) and germline loss- (Zfp521−/−) of Zfp521 functionmouse models. Four-week-old Zfp521-TG mice demonstrate increased bone volume, whereas 3-week-old Zfp521−/−

mice are osteopenic as shown by Micro-CT.

genes (Runx2, alkaline phosphatase, collagen type 1,osteocalcin, bone sialoprotein, osteopontin) was de-creased in Zfp521-overexpressing calvarial cells,suggesting that Zfp521 antagonizes early stages ofosteoblast differentiation. These findings suggestthat Zfp521 inhibits early osteoblast differentiationbut promotes late stages of osteoblast maturation.

Zfp521 expression is regulated duringosteoblast differentiation

If Zfp521 is a physiologically important brake on os-teoblast differentiation, its expression might be reg-ulated during differentiation of osteoblast precur-sors. We therefore determined if agents that mod-ulate osteoblast differentiation affect endogenousZfp521 expression. Endogenous Zfp521 mRNA lev-els in mesenchymal cell lines and primary calvarialosteoblasts steadily increased with time in culture.Treatment with BMP-2, which induces osteoblast

differentiation and maturation, repressed Zfp521expression. In contrast, continuous treatment of cal-varial osteoblasts with parathyroid hormone-relatedprotein (PTHrP)-1–34, which favors proliferationbut antagonizes osteoblast differentiation, increasedZfp521 expression.22

Zfp521 interacts with Runx2 and inhibitsits transcriptional activity

Runx2 is a key regulatory factor downstream ofBMPs and PTHrP. We therefore examined theeffect of Zfp521 on Runx2-induced Runx2-luciferaseand osteocalcin-luciferase reporter gene activation inseveral cell lines. Consistent with our ex vivo anal-ysis of primary osteoblasts from Zfp521-TG mice,Zfp521 strongly repressed the Runx2-mediated re-porter gene activation, indicating that Zfp521 antag-onizes Runx2 transcriptional activity (Fig. 3). More-over, forced expression of Runx2 dose dependently

34 Ann. N.Y. Acad. Sci. 1192 (2010) 32–37 c© 2010 New York Academy of Sciences.


Figure 3. Zfp521 represses the Runx2-mediated reportergene activation.

reversed the Zfp521-induced restriction of the os-teoblast phenotype, again indicating that Zfp521 an-tagonizes Runx2 activity. This functional relation-ship between Zfp521 and Runx2 suggested that thetwo proteins might interact, and we in fact foundthat both molecules co-immunoprecipitated in thesame complex. Thus, Zfp521 associates with Runx2and represses its transcriptional activity, indicatingthat inhibition by Zfp521 of osteoblast differentia-tion could be a consequence, at least in part, of itsinteraction with Runx2.22

However, this is probably not the only mecha-nism of action of Zfp521. For example, it is knownthat Zfp521 interacts with Ebf1,23 and it has been re-ported that the deletion of Ebf1 results in high bonemass.24 Thus, in addition to antagonizing Runx2,some of the effects of Zfp521 on bone might bemediated through inhibition of Ebf1 function.

Discussion

Bone is constantly formed by osteoblasts, whicharise from mesenchymal precursor cells and un-dergo a carefully controlled differentiation programto enter a matrix-secreting phase and to eventuallyadopt a terminally differentiated stage. Maintain-ing the proper balance between advancing someosteoblasts toward a more committed or maturestate while retaining others in earlier less committedstages requires a tight regulation of the relative activ-

ities of factors that favor the progression of the cellstoward the more committed stages and those that ar-rest that progression. The progression of early mes-enchymal cells toward mature osteoblasts involvesa series of transcription factors (i.e., Runx2, �-catenin, osterix, and several members of the activa-tor protein-1 family) that act at highly specific stepsduring the differentiation program to either retainor advance the cell through specific checkpoints.3

Runx2 regulates both chondrogenesis, where it isrequired for final maturation of hypertrophic chon-drocytes, and osteogenesis, where it plays a dual role,first as a “master regulator” that is required for os-teoblast formation6–8 and later as an antagonist oflate stages of osteoblast maturation.9–11

The concept of stage-specific effects of signal-ing effectors during differentiation is further sup-ported by observations made by several groups oncanonical Wnt signaling. Notwithstanding the factthat Wnt/�-catenin signaling facilitates the com-mitment of mesenchymal precursor cells toward theosteoblast lineage and supports the function of earlystage osteoblasts, it has been reported that canonicalWnt signaling impairs late-stage osteoblast differen-tiation and/or function and must be attenuated toallow terminal osteoblast differentiation.12–14

Our results suggest that Zfp521 is a novel factorthat controls osteoblast differentiation. Both Zfp521and the homologous Zfp423 have been shown toplay similar inhibitory roles in the regulation of dif-ferentiation and maturation of several cell types,suggesting that they might function as “switch pro-teins” controlling the transition from immature tomature cell types.16–19,25,26 Consistent with a possi-ble role of Zfp521 in restricting the progression ofmultipotent cells toward a more committed mes-enchymal precursor stage, we found Zfp521 ex-pressed in early mesenchymal cell precursors aroundearly mesenchymal condensations during skeletaldevelopment, in the perichondrium and periosteumof developing bones, and in specific regions of thegrowth plate. This suggests that Zfp521 expressionis induced at early stages of mesenchymal differ-entiation and possibly repressed as differentiationprogresses, as illustrated in mesenchymal condensa-tions and in the growth plate. Furthermore, expres-sion of Zfp521 was strongly repressed by treatmentof mesenchymal cells with BMP-2, which inducesosteoblast differentiation, and increased by PTHrP,which promotes the proliferation of osteoblast


Zinc finger protein 521 Hesse et al.

Figure 4. Working model of stage-dependent effects ofZfp521 on osteoblast differentiation.

precursors while preventing their differentia-tion.27–29 Consistent with this role, overexpressingZfp521 in primary early osteoblasts and a multipo-tent mesenchymal cell line delayed or reduced theexpression of early osteoblast differentiation mark-ers and the formation of mineralized nodules.22

Because many of the osteoblast marker genes thatwe found to be downregulated by Zfp521 are Runx2target genes, we hypothesized that this effect couldbe the result of antagonism of Runx2 activity byZfp521. Our observations that Zfp521 physicallyinteracted with Runx2 and antagonized the Runx2-induced activation of reporter genes and that over-expression of Runx2 dose dependently reversed theZfp521-induced repression of the osteoblast phe-notype support that hypothesis.22 Thus, Zfp521 ap-pears to repress the early stages of osteoblast dif-ferentiation, at least in part by antagonizing Runx2activity, but favors late-stage osteoblast function,leading to a high bone mass in mice.22 The effect ofZfp521 on Ebf1 may also contribute to the regula-tory role of Zfp521 in osteoblast differentiation andfunction.

In conclusion, our data suggest that Zfp521 hasa dual effect on osteoblasts, slowing down theirtransition into the committed stage, thereby help-ing to maintain a properly sized pool of precursors,and favoring their progression to the last stages ofdifferentiation and their matrix production, therebyfavoring bone formation (Fig. 4). Both of these ef-fects, that is, the antiosteogenic effect of Zfp521 re-vealed in the in vitro osteoblast cultures and the pro-osteogenic effect revealed in the transgenic mice,could at least in part be the consequence of antag-onizing Runx2 activity in osteoblast precursors inone case and in mature osteoblasts in the other. Wetherefore propose that the balance between Zfp521

and Runx2 may contribute to regulating the rate ofosteoblast differentiation and bone formation dur-ing development and in the mature skeleton.

Acknowledgment

This work was supported in part by a grant from theNational Institutes of Health (National Institute ofArthritis and Musculoskeletal and Skin Diseases) toR.B. (AR48218). Additional support was providedby the Deutsche Forschungsgemeinschaft to E.H.(HE 5208/1-1), the Gideon & Sevgi Rodan Fellow-ship from the International Bone and Mineral Soci-ety (E.H. and R.K.), the Academy of Finland (R.K.),and the Dean’s Scholars Award of Harvard School ofDental Medicine (E.H.). We thank William Hornefor helpful comments and N. Brady for performingthe Micro-CT analysis.

Conflict of interest

The authors declare no conflicts of interest.

References

1. Jilka, R.L. 2003. Biology of the basic multicellular unit

and the pathophysiology of osteoporosis. Med. Pediatr.

Oncol. 41: 182–185.

2. Parfitt, A.M. 2002. Targeted and nontargeted bone re-

modeling: relationship to basic multicellular unit origi-

nation and progression. Bone 30: 5–7.

3. Harada, S. & G.A. Rodan. 2003. Control of osteoblast

function and regulation of bone mass. Nature 423: 349–

355.

4. Karsenty, G. & E.F. Wagner. 2002. Reaching a genetic

and molecular understanding of skeletal development.

Dev. Cell 2: 389–406.

5. Karsenty, G., H.M. Kronenberg & C. Settembre. 2009.

Genetic control of bone formation. Annu. Rev. Cell Dev.

Biol. 25: 629–648.

6. Ducy, P. et al. 1997. Osf2/Cbfa1: a transcriptional acti-

vator of osteoblast differentiation. Cell 89: 747–754.

7. Komori, T. et al. 1997. Targeted disruption of Cbfa1 re-

sults in a complete lack of bone formation owing to

maturational arrest of osteoblasts. Cell 89: 755–764.

8. Otto, F. et al. 1997. Cbfa1, a candidate gene for clei-

docranial dysplasia syndrome, is essential for osteoblast

differentiation and bone development. Cell 89: 765–771.

9. Maruyama, Z. et al. 2007. Runx2 determines bone ma-

turity and turnover rate in postnatal bone development

and is involved in bone loss in estrogen deficiency. Dev.

Dyn. 236: 1876–1890.

36 Ann. N.Y. Acad. Sci. 1192 (2010) 32–37 c© 2010 New York Academy of Sciences.


10. Geoffroy, V. et al. 2002. High bone resorption in adult ag-

ing transgenic mice overexpressing cbfa1/runx2 in cells

of the osteoblastic lineage. Mol. Cell Biol. 22: 6222–6233.

11. Liu, W. et al. 2001. Overexpression of Cbfa1 in os-

teoblasts inhibits osteoblast maturation and causes os-

teopenia with multiple fractures. J. Cell Biol. 155: 157–

166.

12. Van Der Horst, G. et al. (2005. Downregulation of Wnt

signaling by increased expression of Dickkopf-1 and

-2 is a prerequisite for late-stage osteoblast differen-

tiation of KS483 cells. J. Bone Miner. Res. 20: 1867–

1877.

13. Li, X. et al. 2005. Dkk2 has a role in terminal osteoblast

differentiation and mineralized matrix formation. Nat.

Genet. 37: 945–952.

14. Rodda, S.J. & A.P. McMahon. 2006. Distinct roles for

Hedgehog and canonical Wnt signaling in specification,

differentiation and maintenance of osteoblast progeni-

tors. Development 133: 3231–3244.

15. Bond, H.M. et al. 2004. Early hematopoietic zinc finger

protein (EHZF), the human homolog to mouse Evi3,

is highly expressed in primitive human hematopoietic

cells. Blood 103: 2062–2070.

16. Warming, S. et al. 2003. Evi3, a common retroviral in-

tegration site in murine B-cell lymphoma, encodes an

EBFAZ-related Kruppel-like zinc finger protein. Blood

101: 1934–1940.

17. Hata, A. et al. 2000. OAZ uses distinct DNA- and protein-

binding zinc fingers in separate BMP-Smad and Olf sig-

naling pathways. Cell 100: 229–240.

18. Tsai, R.Y.L. & R.R. Reed. 1997. Cloning and functional

characterization of Roaz, a zinc finger protein that in-

teracts with O/E-1 to regulate gene expression: impli-

cations for olfactory neuronal development. J. Neurosci.

17: 4159–4169.

19. Tsai, R.Y.L. & R.R. Reed. 1998. Identification of DNA

recognition sequences and protein interaction domains

of the multiple-Zn-finger protein Roaz. Mol. Cell Biol.

18: 6447–6456.

20. Horowitz, M.C. et al. 2005. B cells and osteoblast and

osteoclast development. Immunol. Rev. 208: 141–153.

21. Kieslinger, M. et al. 2005. EBF2 regulates osteoblast-

dependent differentiation of osteoclasts. Dev. Cell 9: 757–

767.

22. Wu, M. et al. 2009. Zfp521 antagonizes Runx2, delays

osteoblast differentiation in vitro, and promotes bone

formation in vivo. Bone 44: 528–536.

23. Hentges, K.E. et al. 2005. Evi3, a zinc-finger protein re-

lated to EBFAZ, regulates EBF activity in B-cell leukemia.

Oncogene 24: 1220–1230.

24. Hesslein, D.G. et al. 2009. Ebf1-dependent control of the

osteoblast and adipocyte lineages. Bone 44: 537–546.

25. Warming, S. et al. 2004. Early B-cell factor-associated

zinc-finger gene is a frequent target of retroviral integra-

tion in murine B-cell lymphomas. Oncogene 23: 2727–

2731.

26. Liberg, D., M. Sigvardsson & P. Akerblad. 2002. The

EBF/Olf/Collier family of transcription factors: regula-

tors of differentiation in cells originating from all three

embryonal germ layers. Mol. Cell Biol. 22: 8389–8397.

27. Van Der Horst, G. et al. 2005. Multiple mechanisms are

involved in inhibition of osteoblast differentiation by

PTHrP and PTH in KS483 Cells. J. Bone Miner. Res. 20:

2233–2244.

28. Zerega, B. et al. 1999. Parathyroid hormone [PTH(1-34)]

and parathyroid hormone-related protein [PTHrP(1-

34)] promote reversion of hypertrophic chondrocytes

to a prehypertrophic proliferating phenotype and pre-

vent terminal differentiation of osteoblast-like cells. J.

Bone Miner. Res. 14: 1281–1289.

29. Martinez, M.E. et al. 1997. C-terminal parathyroid

hormone-related protein inhibits proliferation and dif-

ferentiation of human osteoblast-like cells. J. Bone Miner.

Res. 12: 778–785.


Zinc finger protein 521, a new player in bone formation

Documents

Transcript of Zinc finger protein 521, a new player in bone formation