This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Puccetti, E. Articles by Ruthardt, M. Search for Related Content PubMed PubMed Citation Articles by Puccetti, E. Articles by Ruthardt, M.

The Integrity of the Charged Pocket in the BTB/POZ Domain Is Essential for the Phenotype Induced by the Leukemia-Associated t(11;17) Fusion Protein PLZF/RAR

¹ Department of Hematology, Johann Wolfgang Goethe-Universität, Frankfurt, Germany; ² Department of Histology and Medical Embryology, University "La Sapienza"; and ³ San Raffaele Biomedical Science Park of Rome, Rome, Italy

Requests for reprints: Elena Puccetti and Martin Ruthardt, Med.Klinik III/Hämatologie, Klinikum der J.W. Goethe Universität, Theodor-Stern Kai 7, 60590 Frankfurt, Germany. Phone: 49-6301-7970 (E. Puccetti), 49-6301-6129 (M. Ruthardt); Fax: 49-6301-6131; E-mail: puccetti{at}em.uni-frankfurt.de and ruthardt{at}em.uni-frankfurt.de.

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Acute myeloid leukemia is characterized by a differentiation block as well as by an increased self-renewal of hematopoietic precursors in the bone marrow. This phenotype is induced by specific acute myeloid leukemia–associated translocations, such as t(15;17) and t(11;17), which involve an identical portion of the retinoic acid receptor (RAR) and either the promyelocytic leukemia (PML) or promyelocytic zinc finger (PLZF) genes, respectively. The resulting fusion proteins form high molecular weight complexes and aberrantly bind several histone deacetylase–recruiting nuclear corepressor complexes. The amino-terminal BTB/POZ domain is indispensable for the capacity of PLZF to form high molecular weight complexes. Here, we studied the role of dimerization and binding to histone deacetylase–recruiting nuclear corepressor complexes for the induction of the leukemic phenotype by PLZF/RAR and we show that (a) the BTB/POZ domain mediates the oligomerization of PLZF/RAR; (b) mutations that inhibit dimerization of PLZF do the same in PLZF/RAR; (c) the PLZF/RAR-related block of differentiation requires an intact BTB/POZ domain; (d) the mutations interfering with either folding of the BTB/POZ domain or with its charged pocket prevent the self-renewal of PLZF/RAR-positive hematopoietic stem cells. Taken together, these data provide evidence that the dimerization capacity and the formation of a functionally charged pocket are indispensable for the PLZF/RAR-induced leukemogenesis.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Acute myeloid leukemia (AML) is characterized by a differentiation block as well as by an aberrant self-renewal of hematopoietic precursors repressing the normal hematopoiesis as a consequence of the increasing accumulation of immature blasts in the bone marrow.

This phenotype can be generated by specific AML-associated translocations, such as the t(15;17) and t(11;17) (1). Both translocations fuse the identical portion of the retinoic acid receptor (RAR) either to the promyelocytic leukemia gene (PML) or to the promyelocytic zinc finger gene (PLZF), respectively (2). In several cell models, the resulting PML/RAR and PLZF/RAR fusion proteins recapitulate the leukemic phenotype by inducing a state of refractoriness to various inducers of myeloid differentiation and an increased potential for self-renewal (1, 2). Accordingly, expression in animal models of both PML/RAR and PLZF/RAR leads to the development of leukemia (3). The treatment with all-trans retinoic acid (t-RA) is able to overcome the block of differentiation of PML/RAR—but not that of PLZF/RAR-positive—blasts (2).

Both fusion proteins form high molecular weight complexes and aberrantly bind several histone deacetylase (HDAC) recruiting nuclear corepressor (HD-NCR) complexes. In vivo, PML/RAR forms high molecular weight complexes through the "coiled-coil" region of PML (4). A PML/RAR lacking the coiled-coil domain is unable to block terminal differentiation and to mediate the response to t-RA (4, 5). Comparably, PLZF/RAR is able to form high molecular weight complexes through its BTB/POZ domain in the PLZF portion (6). Crystallization studies on the PLZF-BTB/POZ domain revealed a structure enabling it to form homodimers. These homodimers themselves assume a quaternary structure forming an oligodimer complex (7). The amino-terminal BTB/POZ domain is indispensable for the capacity of PLZF to form high molecular weight complexes (7). Symmetry-related residues from each of the BTB/POZ monomers form a charged pocket on the dimer by which HD-NCR binds to the BTB/POZ domain. Specific point mutations in the BTB/POZ domain are able to abrogate either the oligomerization or the binding to HD-NCR complexes.

Another common feature of PML/RAR and PLZF/RAR is the formation of stable complexes with several HD-NCR complexes. Both fusion proteins bind the N-CoR–related as well as the SMRT–related HD-NCR through the CoR box region of RAR in a ligand-dependent manner, and binding can be released by pharmacologic doses of t-RA. In contrast to PML/RAR, PLZF/RAR contains a second binding site for HD-NCR complexes in the PLZF portion of the fusion protein. This binding is resistant to t-RA and is mainly located in the BTB/POZ domain (2, 8, 9). The BTB/POZ domain binds to SMRT, mSin3A, and directly to HDAC1 in a ligand-independent manner (2). The aberrant recruitment of the HDAC activity enables PLZF as well as PLZF/RAR to repress transcription (2). Resistance of PLZF/RAR-positive acute promyelocytic leukemia blasts to t-RA–induced differentiation seems to be related to the presence of the BTB/POZ domain.

Recently, the critical residues involved in both the interaction with corepressors and the capacity to form high molecular weight complexes has been exactly mapped to the BTB/POZ domain of PLZF (8). The crystal structure of the BTB/POZ domain shows typical features of obligate homodimers with a tightly intertwined dimer characterized by an extensive hydrophobic interface, as well as a surface-exposed groove lined with conserved amino acids formed at the dimer interface, suggestive of a peptide binding site (10); this is also called the charged pocket (7). An extensive structure-function analysis showed that the charged pocket motive plays a major role in transcriptional repression by PLZF. The pocket is formed by symmetry-related residues from each of the monomers, including pairs of aspartates at position 35 and arginines at position 49. Mutations of the BTB/POZ domain that neutralize key charged pocket residues do not disrupt dimerization, but abrogate the ability of PLZF to repress transcription and lead to the loss of interaction with the HD-NCR complexes (8).

In the present work, we tried to disclose the role of the dimerization/high molecular weight complexes formation and the binding to the HD-NCR complexes for the induction of the leukemic phenotype by the PLZF/RAR fusion protein. Therefore, we introduced selected point mutations in the PLZF-BTB/POZ domain. For the complete inhibition of the high molecular weight complex formation, we modified a deep monocore residue that interferes with the overall folding of the BTB/POZ domain, determining a situation comparable with its deletion (7), whereas the surface of the BTB/POZ domain was altered to create a partial inhibition of the high molecular weight complex formation. The recruitment of HDAC activity through the PLZF portion of the fusion protein was abrogated by a modification of the charged pocket, which only slightly interferes with high molecular weight complex formation of PLZF, but abrogates binding to the HD-NCR complexes.

Here, we compared the effect of these PLZF/RAR mutants with that of PLZF/RAR on the biology of hematopoietic stem cells (HSC).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Plasmids. The BTB/POZ domain mutations were created as described elsewhere (7). The following mutations were introduced by PCR-mediated mutagenesis using the indicated specific oligonucleotides (in parentheses): D35N (5'-ACTTTGTGCAATGTGGTCATC-3' and 5'-GATGACCACATTGCACAAAGT-3'), R49Q (5'-CACGCCCACCAGACGGTGCTG-3' and 5'-CAGCACCGTCTGGTGGGCGTG-3'), Y88A (5'-CTGGAGTATGCAGCTACAGCCACG-3' and 5'-CGTGGCTGTAGCTGCATACTCCAG-3'), and L103E (5'-GATGACCTGGAGTATGCGGCC-3' and 5'-GGCCGCATACTCCAGGTCATC-3'). Expression vectors for PLZF/RAR, PLZF, and their BTB/POZ mutants were produced using the Gateway recombination technology (Invitrogen, Karlsruhe, Germany). PINCO (11), pCDNA3, and PCDNA3-HA vectors were previously converted into destination vectors using the Gateway Vector Conversion System (Invitrogen). The cDNA encoding PLZF, PLZF/RAR, and their BTB/POZ mutants were first cloned into pENTR.1A using a BamHI restriction site and then transferred into the destination vectors by LR recombination following the manufacturer's instructions (Invitrogen).

Coimmunoprecipitation, Western blotting, in vitro translation, and in vitro binding. For coimmunoprecipitation, 293 cells were transfected with 5 µg of plasmid DNA by calcium phosphate precipitation according to widely established procedures. After washing, the cells were collected in E1A buffer. An antihemagglutinin (anti-HA) protein affinity matrix constituted of rat monoclonal anti-HA antibody (clone 3F10), covalently coupled to agarose beads (Roche, Basel, Switzerland), was used. The beads were washed five times in E1A buffer (150 mmol/L NaCl) and resuspended in SDS sample buffer. Immunoprecipitated proteins were detected by Western blotting with the indicated antibodies by the enhanced chemiluminescence method. The anti-RAR polyclonal antibody (12) and anti–acetyl-H3 rabbit polyclonal antibody from Upstate Biotechnology, Inc. (Lake Placid, NY), were used.

For in vitro translation, the coupled TnT T7/Sp6 transcription and translation kit was used according to the manufacturer's instruction (Promega, Mannheim, Germany). Five microliters of in vitro–translated ³⁵S-labeled proteins were diluted in SDS-PAGE loading buffer in the presence of ß-mercaptoethanol and 2% SDS, boiled and resolved by SDS-PAGE and visualized by autoradiography.

In vitro-binding assays between GST-N-CoR, GST-SMRT, GST-Sin3A, and GST-HDAC1 fusion proteins and in vitro–translated ³⁵S-labeled PLZF/RAR and its BTB/POZ mutants were done in E1A buffer (NaCl 50 mmol/L) as described elsewhere (9).

Transfection and gel filtration analysis of all-trans retinoic acid binding activity. COS-1 cells were transiently transfected by electroporation with pCDNA3 expression vectors containing cDNAs for wild-type PLZF/RAR or its BTB/POZ mutants as described (6). Apparent molecular weights were calculated on the basis of the elution times of a series of marker proteins used to calibrate the system, such as blue dextran (molecular weight 2,000,000), thyroglobulin (molecular weight 669,000), apoferritin (molecular weight 443,000), ß-amylase (molecular weight 200,000), alcohol dehydrogenase (molecular weight 150,000), bovine albumin (molecular weight 66,000), ovalbumin (molecular weight 45,000), carbonic anhydrase (molecular weight 29,000), and lactalbumin (molecular weight 14,200).

Reporter assays. Reporter constructs used in this study included a vector containing a Luciferase reporter construct under the control of the RARß promoter (6); for normalization, the cmv-Renilla vector for the expression of the Renilla luciferase driven by a cytomegalovirus (CMV) promoter was used. 293 T cells were plated in 12-well tissue culture dishes and transiently transfected (as described above). After 24 hours of exposure to t-RA and valproic acid (VPA) alone or in combination, luciferase activity was determined using the Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer's instructions. All transfection experiments were done at least thrice in triplicates. The fold induction of transcription was calculated in relationship to the transcription of the reporters in the absence of the promoter (pGL3 Basic, Promega).

Isolation of murine Sca1⁺/lin^– hematopoietic stem cells. Female C57BL/6N mice from 6 to 12 weeks of age (Charles River, Sulzfeld, Germany) were used as source of Sca1⁺/lin^– murine HSC. The mice were killed by CO₂ asphyxiation; femur and tibia were collected and bone marrow was harvested by flushing the bones with syringe and 26-gauge needle. Sca1⁺/lin^– cells were purified by immunomagnetic beads using the MACS cell separation columns according to the manufacturer's instruction (Miltenyi, Bergisch-Gladbach, Germany). Sca1⁺ cells were "lineage depleted" by labeling the cells with biotin-conjugated lineage panel antibodies B220, CD3, Gr-1, Mac-1, and TER-119 (PharMingen, San Diego, CA). Labeled cells were removed using streptavidin-loaded MACS cell separation columns (Miltenyi). Purified cells were prestimulated before further use for 2 days in medium, containing murine interleukin-3 (mIL-3; 20 ng/mL), mIL-6 (20 ng/mL), and murine stem cell factor (100 ng/mL; Cell Concepts, Umkirch, Germany).

Retroviral infection. Ecotropic Phoenix packaging cells were transiently transfected with the indicated retroviral vectors as described before (1) and transfection efficiency was assessed by the detection of the percentage of green fluorescent protein–positive cells through fluorescence-activated cell sorting (FACS) analysis. Retroviral supernatant was collected at days 2 and 3 after transfection and shock-frozen in liquid nitrogen and stored at –80°C. For the infection, the retroviral supernatant was thawed on ice. Target cells were plated onto retronectin-coated (Takara-Shuzo, Shiga, Japan) non–tissue culture treated 24-well plates and exposed to the retroviral supernatant for 3 hours at 37°C in the presence of 4 µg/mL polybrene (Sigma-Aldrich, Steinheim, Germany). Cells were centrifuged at 2,200 rpm for 45 minutes. Infection was repeated four times and infection efficiency had to be at least 70% for each sample as assessed by the detection of green fluorescent protein–positive cells by FACS. Differences of transduction efficiency between the samples did not exceed 10%.

Colony assays, replating efficiency, and differentiation. Sca1⁺/lin^– cells was cultivated in RPMI supplemented with 10% FCS, mIL-3 (20 ng/mL), mIL-6 (20 ng/mL), and murine stem cell factor (100 ng/mL; StemCell Technologies, Vancouver, BC, Canada) with or without granulocyte colony-stimulating factor (G-CSF; 60 ng/mL) and granulocyte macrophage colony-stimulating factor (GM-CSF; 20 ng/mL). On day 6 after plating, the cells were cytocentrifuged for morphologic analysis and stained for the determination of surface marker expression by FACS. At day 5 after infection, Sca1⁺/lin^– cells were plated at 5,000 cells/mL into methylcellulose supplemented with mIL-3 (20 ng/mL), mIL-6 (20 ng/mL), and murine stem cell factor (100 ng/mL; StemCell Technologies) with or without G-CSF (60 ng/mL) and GM-CSF (20 ng/mL). t-RA was used at the concentration of 10^–6 mol/L. The HDAC inhibitors were used at a concentration of 0.15 mmol/L (VPA), 1 mmol/L (sodium butyrate, NaBu), 10 or 50 nmol/L (LAQ824, kindly provided by Peter Atadja, Novartis, Basel, Switzerland). On day 10 after plating, the colony number was counted and the plates were photographed. After washing out from methylcellulose, cells were cytocentrifuged for morphologic analysis and stained for the determination of surface marker expression by FACS. Five thousand cells per plate were plated again in methylcellulose determining replating efficiency by serial plating. Differentiation was assessed by May-Grünwald-Giemsa staining of the cytospins as well as by the assessment of the expression of c-Kit, Sca1, Gr-1, and Mac-1 (PharMingen) by FACS.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
The modification of the BTB/POZ charged pocket interferes with the ability of PLZF/RAR to dimerize in vitro. To disclose the role of the dimerization/high molecular weight complex formation and the binding to the HD-NCR in the induction of the leukemic phenotype by PLZF/RAR, we introduced point mutations in the BTB/POZ domain of PLZF/RAR known to interfere with the functionality of the BTB/POZ domain of PLZF. Thus, the BTB/POZ mutants of PLZF/RAR were created in analogy to the mutants recently described for PLZF (7, 8).

A correct folding is required for efficient dimerization of the BTB/POZ domain and vice versa. Thus, we created the folding mutant PLZF/RAR^(Y88A) to abolish the dimerization of PLZF/RAR. This mutant was obtained by converting the tyrosine localized in the hydrophobic monomer core of the BTB/POZ domain in position 88 to alanine (Y88A). Y88A leads to a misfolding of the BTB/POZ domain, creating a situation described as equivalent to the deletion of the entire BTB/POZ domain in PLZF (7, 8).

The "surface" mutant PLZF/RAR^(L103E) was created by the substitution of the leucine in 103 by glutamic acid (L103E). This mutation still allows a partial dimerization of PLZF (7, 8).

To study the role of the binding to HD-NCR in the biology of PLZF/RAR, a double-mutant PLZF/RAR^(D35N/R49Q) was created. Therefore, we first created a PLZF/RAR^(R49Q) by mutating arginine in position 49 to glutamine. The R49Q mutation does not interfere with the dimerization of PLZF/RAR (data not shown). For the double mutant, the aspartic acid at position 35 was substituted by asparagine (D35N) in the PLZF/RAR^(R49Q) mutant construct. The D35N/R49Q mutation abolishes the binding of HD-NCR complexes to the BTB/POZ domain of PLZF and reduces slightly the capacity to dimerize (8).

To verify the influence of the mutants on the dimerization of PLZF/RAR, we studied the capacity of the PLZF/RAR mutants to bind to the analogous PLZF mutants. The capacity to bind was assessed by coimmunoprecipitation experiments using HA-tagged PLZF mutants as bait for the precipitation of the correspondent PLZF/RAR mutants. As controls, we used wild-type HA-PLZF and wild-type PLZF/RAR.

As described in Fig. 1A, the HA-PLZF is able to coimmunoprecipitate the PLZF/RAR. In contrast to the "folding mutation" Y88A, which abolished the capacity of PLZF to bind PLZF/RAR, the surface mutation L103E, as well as the double D35N/R49Q mutation, did not interfere with the binding (Fig. 1A).

View larger version (37K): [in this window] [in a new window]   Figure 1. Dimerization capability and high molecular weight complex formation of PLZF/RAR BTB/POZ mutants. A, 293 cells were transfected with pCDNA3-PLZF/RAR together with pCDNA3-HA-PLZF constructs, or pCDNA3-PLZF/RAR together with pCDNA3-HA-PLZF harboring the same POZ mutation as indicated. An anti-HA protein affinity matrix was used for coimmunoprecipitation (CoIP). Immunoprecipitated proteins were detected by Western blotting with the anti-PLZF antibody. C, control cells transfected with mock vectors; WT, wild-type PLZF/RAR and/or PLZF; INPUT, lysates from the transfected cells were loaded. B, in vitro–translated and 35S-labeled PLZF/RAR as well as its indicated BTB/POZ mutants were resolved by SDS-PAGE and visualized by autoradiography. The bands running over the 250 kDa ladder band show the formations of high molecular weight complexes. C, COS-1 cells were transfected with pCDNA3 expression vectors for PLZF/RAR and its BTB/POZ mutants. Nuclear extracts labeled with 10 nmol/L [3H]RA were analyzed by a gel filtration size exclusion column. The numbers on the X axis represent the eluted fractions. Apparent molecular weights were calculated on the basis of the elution times of a series of marker proteins used to calibrate the system.

The formation of PLZF/RAR high molecular weight complex is correlated with its capacity to dimerize. PLZF/RAR forms high molecular weight complexes similar to PML/RAR (13). The biology of PML/RAR depends on its capacity to form high molecular weight complexes (4, 14, 15). The PLZF-BTB/POZ domain mediates the formation of high molecular weight complexes (7, 16). To disclose the role of oligomerization of PLZF/RAR for its capacity to form high molecular weight complexes, we analyzed the size of the complexes formed by PLZF/RAR and its BTB/POZ mutants by size exclusion high-performance liquid chromatography of [³H]RA-labeled proteins as described before (6).

The folding mutant PLZF/RAR^(Y88A) exhibited the highest peak at 110 kDa corresponding to the molecular weight of a monomer and were unable to form high molecular weight complexes at 660 kDa. In contrast, PLZF/RAR^(D35N/R49Q) still allowed high molecular weight complex formation peaking at 660 kDa, like PLZF/RAR (Fig. 1C). The high molecular weight complex formed by PLZF/RAR^(L103E) was modified compared with that formed by PLZF/RAR peaking at a slightly lower mass weight. The peak at 50 kDa visible in all samples corresponds to endogenous RAR as previously shown (6).

The influence of the mutations on the stability of the complexes was assessed by running ³⁵S-labeled, in vitro–translated PLZF/RAR and its BTB/POZ mutants in a SDS-PAGE. ³⁵S-labeled in vitro–translated PLZF and PLZF/RAR as well as PML and PML/RAR expressed in cells give origin to a band of over 250 kDa under denaturating conditions in a SDS-PAGE, indicating a high stability of the complexes formed (Fig. 1B and data not shown). As shown in Fig. 1B, none of the mutants, not even the double mutant that is able to form high molecular weight complexes, exhibited the band of over 250 kDa like PLZF/RAR. This finding suggests that the high molecular weight complexes formed by PLZF/RAR^(D35N/R49Q) are less stable than those of PLZF/RAR.

In summary, these data indicate that a folding that still allows to dimerize is a prerequiste condition for the capacity to form high molecular weight complexes in the PLZF/RAR fusion protein, but the stability of this complex is strictly correlated to a correctly folded BTP/POZ.

The dimerization of PLZF/RAR is responsible for its promoter repression activity. To determine the influence of the BTB/POZ mutations on the functionality of PLZF/RAR to repress transcription from a t-RA target promoter, we compared the transactivation activity of the PLZF/RAR mutants with that of PLZF/RAR on the RARß promoter upon exposure to t-RA alone or in combination with VPA (6).

As shown in Fig. 2A, in the absence of t-RA RAR, PLZF/RAR, as well as all mutants, were repressors of the RARß promoter. As known, the exposure to t-RA converted RAR, but not PLZF/RAR, to an activator of the RARß promoter (2). In contrast, the mutations Y88A and D35N/R49Q reduced the capacity of PLZF/RAR to repress the promoter in the presence of t-RA, whereas the surface mutant PLZF/RAR^(L103E) was able to slightly activate the promoter in the presence of t-RA (Fig. 2A). VPA alone did not significantly influence the repressor activity of RAR, PLZF/RAR, and its BTB/POZ mutants. Furthermore, VPA did not modify the effects of RAR, PLZF/RAR, and its BTB/POZ mutants on the RARß promoter upon exposure to t-RA (Fig. 2A).

View larger version (24K): [in this window] [in a new window]   Figure 2. Transcriptional activity of PLZF/RAR and its BTB/POZ mutants. A, PLZF/RAR and its BTB/POZ mutants were tested for their activity on the RARß promoter. These plasmids were transfected into 293 cells along with a reporter plasmid containing RARß promoter. Renilla under the control of a CMV promoter was used as internal control. The transcriptional effects are expressed as fold induction of transcriptional activity compared with the activity of the basic vector containing the luciferase reporter without promoter (control). B, in vitro binding assays between GST fusion proteins of N-CoR, SMRT, Sin3A, and HDAC1 and the 35S-labeled PLZF/RAR or its BTP/POZ mutants in presence/absence of t-RA. INPUT represents 10% of the labeled protein added to the binding.

The charged pocket has no influence on the corepressor and histone deacetylase 1 binding of PLZF/RAR. The mutation of amino acids that determine the charge of the charged pocket abrogate in PLZF the interaction between the corepressors and the BTB/POZ domain without disturbing the normal folding of the dimer. To assess the effects of the D35N-R49Q double mutation on the capacity of PLZF/RAR to bind N-CoR, SMRT, and HDAC1, we did in vitro binding experiments. To exclude the binding of the corepressors to the RAR portion of the fusion protein, we studied the binding in absence, as well as in presence, of t-RA. The mutants were able to bind to the corepressors and to HDAC1 to the same extent than the PLZF/RAR even in the presence of t-RA (Fig. 2B).

Interestingly, t-RA completely abolished the binding of PLZF/RAR and its mutants to SMRT, indicating that SMRT does not efficiently bind to the PLZF portion of PLZF/RAR, whereas it is well known to bind PLZF.

These data indicate that, in contrast to PLZF, the integrity of the charged pocket formed by the BTB/POZ domain of PLZF/RAR has no influence on the binding to N-CoR, SMRT, and HDAC1.

The integrity of the BTB/POZ domain is essential for the PLZF/RAR-related differentiation block in murine hematopoietic stem cells. To investigate the influence of the BTB/POZ mutations on the PLZF/RAR-induced leukemic differentiation block, we expressed PLZF/RAR and its BTB/POZ mutants in murine Sca1⁺/lin^– HSC by retroviral transduction. Protein expression was controlled by indirect immunofluorescence with an anti-RAR antibody (data not shown). Mock-infected cells were used as controls. The cells were cultivated in liquid culture in presence or absence of G/GM-CSF, and differentiation was assessed by morphology and expression of differentiation-specific surface markers, such as Gr-1 and Mac-1.

In contrast to PLZF/RAR, the CP mutant PLZF/RAR^(D35N/R49Q), as well as the BTP/POZ folding mutant PLZF/RAR^(Y88A), and the surface mutant PLZF/RAR^(L103E) were unable to block G/GM-CSF–induced differentiation of Sca1⁺/lin^– cells (Fig. 3A and B) as revealed by the increased number of mature granulocytes and monocytes present in the transduced cells (Fig. 3A). These findings were confirmed by the induction of the differentiation-specific surface markers Gr-1 and Mac-1 in the absence and presence of G/GM-CSF. In fact, Gr-1 and Mac-1 induction reached to the same level in cells expressing PLZF/RAR^(D35N/R49Q), PLZF/RAR^(Y88A), or PLZF/RAR^(L103E) than the mock-infected cells (Fig. 3B).

View larger version (62K): [in this window] [in a new window]   Figure 3. Effect of the BTB/POZ mutations on the block of myeloid differentiation induced by PLZF/RAR. Sca1+/lin– HSCs were transduced with PLZF/RAR or its BTB/POZ mutants. Mock-transduced cells were used as control. The cells were seeded in semisolid medium ± G-CSF/GM-CSF for the induction of granulocytic/monocytic differentiation. At day 10, the cells were harvested and analyzed. A, morphologic analysis of cells stained with May-Grünwald-Giemsa. Original magnification, x400. The mutations are indicated. B, expression of differentiation-specific surface marker: Gr-1 and Mac-1 are markers for myeloid differentiation. One representative experiment of at least three is given. C, expression of the stem cell markers Sca1 and c-Kit. One representative experiment of at least three is given.

The increased replating efficiency of PLZF/RAR–expressing hematopoietic stem cells directly depends on the integrity of the charged pocket of PLZF/RAR. Recently, we have shown that the expression of PLZF/RAR considerably increases the replating efficiency of HSC as a sign of an augmented capacity for self-renewal contributing to the leukemic phenotype (1).

To define the impact of the high molecular weight complex formation and the charged pocket of PLZF/RAR on its capacity to increase self-renewal of HSCs, we compared the replating efficiency of Sca1⁺/lin^– stem cells expressing PLZF/RAR with that of Sca1⁺/lin^– stem cells expressing PLZF/RAR BTB/POZ mutants. The PLZF/RAR-expressing cells were able to form colonies beyond the ninth plating. The CP mutant PLZF/RAR^(D35N/R49Q) as well as the BTP/POZ null mutant PLZF/RAR^(Y88A) were unable to increase the replating efficiency of HSC compared with mock-transduced control cells, whereas the surface mutant PLZF/RAR^(L103E) allowed replatings until the fifth passage (Table 1).

View this table: [in this window] [in a new window]   Table 1. Serial replating of Sca1+/lin– HSCs expressing PLZF/RAR and PLZF/RAR BTB/POZ mutants

Taken together, these data show that the mutations interfering with the folding of the BTB/POZ domain as well as with the charged pocket abolish the capacity of PLZF/RAR to aberrantly increase the replating efficiency of HSC.

The treatment with histone deacetylase inhibitors does not revert the differentiation block and the aberrant replating efficiency of PLZF/RAR-expressing hematopoietic stem cells. The PLZF/RAR^(D35N/R49Q) abolished the capacity to induce the PLZF/RAR-related phenotype in HSC, but was still able to repress transcription even in the presence of t-RA. The analogous D35N/R49Q mutation led to a partial loss of stability and folding in PLZF. Also, the surface mutant PLZF/RAR^(L103E) exhibited a reduced capacity to increase the replating efficiency. Therefore, we wanted to exclude the possibility that the abrogation of the PLZF/RAR-related phenotype by the D35N/R49Q is due to the abrogation of the binding to the HD-NCR and not to this slight defect in dimerization. Hence, we treated the PLZF/RAR-expressing HSC with different HDAC inhibitors: VPA, NaBu, and LAQ824. The first two are known to induce differentiation in AML blasts by inhibiting HDAC activity (17), the third was shown to have an hyperacetylating activity exceeding that of trichostatin A (18). We first selected HSC-expressing PLZF/RAR by serial replating of Sca1⁺/lin^– HSC retrovirally transduced with PLZF/RAR in semisolid medium as described above. The following platings were done in presence/absence of VPA (150 µg/mL) and NaBu (1 mmol/L). The effect of VPA and NaBu on the differentiation was assessed by the expression of differentiation-specific surface markers. Neither VPA nor NaBu was able to overcome the PLZF/RAR-induced differentiation block as revealed by their incapacity to increase the expression of Mac-1 and Gr-1 and their inability to decrease the expression of the stem cell markers Sca1 and c-Kit (Fig. 4A). Furthermore, HDAC inhibition did not reduce the replating efficiency of PLZF/RAR-positive HSC, as shown by the fact that the exposure to VPA or NaBu still allowed at least three further platings (Table 2A), and did not inhibit the selection of immature HSC by the expression of PLZF/RAR, as shown by the levels of the stem cell markers Sca1 and c-Kit (Fig. 4A). Both HDAC inhibitors were able to increase the amount of acetylated histone 3 without any influence on the expression level of PLZF/RAR (Fig. 4B).

View larger version (33K): [in this window] [in a new window]   Figure 4. Effect of HDAC inhibition and exposure to t-RA on the phenotype of PLZF/RAR-expressing HSCs. A, differentiation of PLZF/RAR-expressing HSCs in absence/presence of VPA or NaBu. Gr-1 and Mac-1 were detected as markers of myeloid differentiation and Sca1 and c-Kit were detected as stem cells markers. B, Western blot of Sca1+/lin– HSCs at the sixth passage. Expression of PLZF/RAR was detected with a polyclonal anti-RAR antibody. Sca1+/lin– cells were used as control (C). Western blot of Sca1+/lin– HSCs expressing PLZF/RAR at the sixth passage. Histone 3 acetylation was detected in absence/presence of VPA or NaBu with the anti-acetyl-H3 rabbit polyclonal antibody. Coomassie staining was used as loading control. C, differentiation of PLZF/RAR-expressing HSC in absence/presence of VPA or LAQ824, 10 or 50 nmol/L. Gr-1 and Mac-1 were detected as markers of myeloid differentiation and Sca1 and c-Kit were detected as stem cell markers. The percentage of positive cells is reported for each of three serial platings.

View this table: [in this window] [in a new window]   Table 2. Effect of HDAC inhibition and exposure to t-RA on the replating efficiency of HSCs expressing PLZF/RAR

Upon exposure to Laq824, the percentage of Sca1-positive as well as of c-Kit–positive cells reached the same levels during the three replatings as obtained in VPA-treated cells. In comparison to VPA-treated cells or untreated controls, the percentage of Gr-1–positive as well as of Mac-1–positive cells was increased only in the first two replatings in the presence of Laq824 (Fig. 4C).

These data show that the exposure of PLZF/RAR-positive HSC to HDAC inhibitor alone or in combination with t-RA is unable to completely abrogate the selection of immature HSC by PLZF/RAR.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
The aim of this study was to clarify the relevance of the aberrant recruitment of HD-NCR separately from that of the formation of high molecular weight complexes for the capacity of PLZF/RAR to determine the leukemic phenotype in HSC.

First, we showed that the BTB/POZ domain is not only crucial for the function of PLZF but, moreover, plays a fundamental role in the biological function of PLZF/RAR. This is proven by the fact that the mutations that interfere with the dimerization of PLZF also hinder PLZF/RAR, which confirms data on a BTB/POZ deletion mutant of PLZF/RAR (20). Comparably to PLZF, the BTB/POZ domain of PLZF/RAR seems to be the only dimerization interface of PLZF/RAR and the only structure by which PLZF/RAR forms high molecular weight complexes in vivo. The destruction of the folding caused by the hydrophobic monomer core mutant PLZF/RAR^(Y88A) results in the formation of a complete null mutant for dimerization as well as for transcriptional repression and in consequence for the PLZF/RAR-related leukemic phenotype. This is in accordance with the effects of this mutation in PLZF, where the all folding mutants abolish the capacity to dimerize as well as to form high molecular weight complexes and reduce the transcription repression activity of PLZF (7, 8).

One could speculate that a PLZF/RAR fusion protein unable to form high molecular weight complexes could be again accessible for cofactors by its RAR portion, such as retinoid X receptor, the main cofactor of RAR for t-RA–induced activation of transcription (21).

The stability of the complexes is important for the functionality of the fusion protein as well as for the induction of the leukemic phenotype by PLZF/RAR. The L103E surface mutation, which allows the formation of unstable high molecular weight complexes, render PLZF/RAR able to activate ligand-induced transcription to an intermediate extent between PLZF/RAR and the null mutant PLZF/RAR^(Y88A). This might be related to the fact that PLZF/RAR^(L103E) is still able to dimerize but is deficient in binding to partner proteins indispensable for full transcription repression. Furthermore, it completely abolishes the ability of PLZF/RAR to block differentiation and partially abrogates the aberrant replating efficiency of HSC. The double-mutant PLZF/RAR^(D35N/R49Q) forms high molecular weight complexes with reduced stability. The reduced stability of the high molecular weight complex formed by PLZF/RAR^(D35N/R49Q) might explain why it is able to repress t-RA–induced transcription but not to block differentiation or to aberrantly increase the replating efficiency of HSC.

One could hypothesize that the major feature of PLZF/RAR in determining the leukemic phenotype is the capacity to form high molecular weight complexes in vivo as a consequence of the ability to self-associate and not the aberrant recruitment of HDAC activity. Even if PLZF/RAR^(D35N/R49Q) is completely impaired in its binding to the HD-NCR complexes and even if it cannot be completely excluded that the PLZF/RAR^(L103E) might be impaired in its capacity to bind HD-NCR complexes, the fact that both mutants exhibit nearly the same phenotype in HSC strongly indicates that the aberrant recruitment of HDAC activity plays a minor role in determining the leukemic phenotype. This concept is supported by our findings that the targeting of the HDAC activity recruited by the PLZF as well as by the RAR portion of the fusion protein through the exposure to HDAC inhibitors, such as VPA, NaBu, and Laq824, does not revert the PLZF/RAR-related phenotype in HSC.

These results seem to be in contrast to recent findings that show that HDAC inhibitors, such as VPA, trichostatin A, and suberoylanilide hydroxamic acid, are able to revert the phenotype induced by PLZF/RAR alone or in combination with t-RA (17, 22–25). All these studies are based on leukemia models working with either primary or cell line–derived blast populations. In these blast populations, the HDAC inhibitors induce an arrest of cell growth associated with the induction of differentiation or apoptosis. All these results are fully compatible with the targeting of the blast population by HDAC inhibitors, without hitting the leukemic stem cell population. In fact, in vivo, there is response to the treatment with the HDAC inhibitors leading only to a slight prolongation of the overall survival (23). This is most likely due to the fact that the HSC population is not targeted by the HDAC inhibitors. In contrast, our model here is based on an early HSC population transduced with PLZF/RAR in which there is no effect of HDAC inhibitors on the leukemic phenotype induced by PLZF/RAR. In combination with the previous findings on the blast populations, our findings are of importance because they clearly show that the aberrant recruitment of HDAC activity is critical for overcoming the differentiation block of the leukemic blast population but not for the altered biology of leukemic HSC, the source of the relapse in patients suffering from AML.

Taken together, our data support the notion that the charged pocket is one of the major targets for hitting effectively the PLZF/RAR fusion protein. In fact, the neutralization of the charged pocket inhibits the leukemic phenotype induced by PLZF/RAR to the same extent than the abolition of the high molecular weight complex formation.

	Acknowledgments

 
Grant support: Deutsche Forschungsgemeinschaft grant DFG-Ru. 728/2-2 (M. Ruthardt), Alfred und Angelika Gutermuth Stiftung (E. Puccetti), Italian Association for Cancer Research, and Istituto Pasteur Fondazione Cenci Bolognetti and Universita di Roma "La Sapienza" (C. Nervi).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received 10/ 8/04. Revised 4/20/05. Accepted 4/22/05.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Zheng X, Beissert T, Kukoc-Zivojnov N, et al. -Catenin contributes to leukemogenesis induced by AML-associated translocation products by increasing the self-renewal of very primitive progenitor cells. Blood 2004;103:3535–43.[Abstract/Free Full Text]
2. Melnick A, Licht JD. Deconstructing a disease: RAR, its fusion partners, and their roles in the pathogenesis of acute promyelocytic leukemia. Blood 1999;93:3167–215.[Free Full Text]
3. Puccetti E, Ruthardt M. Acute promyelocytic leukemia: PML/RAR and the leukemic stem cell. Leukemia 2004;18:1169–75.[CrossRef][Medline]
4. Grignani F, Gelmetti V, Fanelli M, et al. Formation of PML/RAR high molecular weight nuclear complexes through the PML coiled-coil region is essential for the PML/RAR-mediated retinoic acid response. Oncogene 1999;18:6313–21.[CrossRef][Medline]
5. Grignani F, Pelicci PG. Pathogenetic role of the PML1RAR fusion protein in acute promyelocytic leukemia. Curr Top Microbiol Immunol 1996;211:269–78.[Medline]
6. Benedetti L, Levin AA, Scicchitano BM, et al. Characterization of the retinoid binding properties of the major fusion products present in acute promyelocytic leukemia cells. Blood 1997;90:1175–85.[Abstract/Free Full Text]
7. Melnick A, Ahmad KF, Arai S, et al. In-depth mutational analysis of the promyelocytic leukemia zinc finger BTB/POZ domain reveals motifs and residues required for biological and transcriptional functions. Mol Cell Biol 2000;20:6550–67.[Abstract/Free Full Text]
8. Melnick A, Carlile G, Ahmad KF, et al. Critical residues within the BTB domain of PLZF and Bcl-6 modulate interaction with corepressors. Mol Cell Biol 2002;22:1804–18.[Abstract/Free Full Text]
9. Puccetti E, Sennewald B, Fosca-Ferrara F, et al. Down-stream regions of the POZ-domain influence the interaction of the t(11;17)-associated PLZF/RAR fusion protein with the histone-deacetylase recruiting co-repressor complex. Hematol J 2001;2:385–92.[CrossRef][Medline]
10. Ahmad KF, Engel CK, Prive GG. Crystal structure of the BTB domain from PLZF. Proc Natl Acad Sci U S A 1998;95:12123–8.[Abstract/Free Full Text]
11. Grignani F, Kinsella T, Mencarelli A, et al. High-efficiency gene transfer and selection of human hematopoietic progenitor cells with a hybrid EBV/retroviral vector expressing the green fluorescence protein. Cancer Res 1998;58:14–9.[Abstract/Free Full Text]
12. Ruthardt M, Testa U, Nervi C, et al. Opposite effects of the acute promyelocytic leukemia PML-retinoic acid receptor (RAR) and PLZF-RAR fusion proteins on retinoic acid signalling. Mol Cell Biol 1997;17:4859–69.[Abstract]
13. Nervi C, Poindexter EC, Grignani F, et al. Characterization of the PML-RAR chimeric product of the acute promyelocytic leukemia-specific t(15;17) translocation. Cancer Res 1992;52:3687–92.[Abstract/Free Full Text]
14. Lin RJ, Evans RM. Acquisition of oncogenic potential by RAR chimeras in acute promyelocytic leukemia through formation of homodimers. Mol Cell 2000;5:821–30.[CrossRef][Medline]
15. Minucci S, Maccarana M, Cioce M, et al. Oligomerization of RAR and AML1 transcription factors as a novel mechanism of oncogenic activation. Mol Cell 2000;5:811–20.[CrossRef][Medline]
16. Beissert T, Puccetti E, Bianchini A, et al. Targeting of the N-terminal coiled coil oligomerization interface of BCR interferes with the transformation potential of BCR-ABL and increases sensitivity to STI571. Blood 2003;102:2985–93.[Abstract/Free Full Text]
17. Gottlicher M, Minucci S, Zhu P, et al. Valproic acid defines a novel class of HDAC inhibitors inducing differentiation of transformed cells. EMBO J 2001;20:6969–78.[CrossRef][Medline]
18. Romanski A, Bacic B, Bug G, et al. Use of a novel histone deacetylase inhibitor to induce apoptosis in cell lines of acute lymphoblastic leukemia. Haematologica 2004;89:419–26.[Abstract/Free Full Text]
19. Remiszewski SW, Sambucetti LC, Bair KW, et al. N-hydroxy-3-phenyl-2-propenamides as novel inhibitors of human histone deacetylase with in vivo antitumor activity: discovery of (2E)-N-hydroxy-3-[4-[(2-hydroxyethyl)[2-(1H-indol-3-yl)ethyl]amino]methyl]phenyl]-2-propenamide (NVP-LAQ824). J Med Chem 2003;46:4609–24.[CrossRef][Medline]
20. Dong S, Zhu J, Reid A, et al. Amino-terminal protein-protein interaction motif (POZ-domain) is responsible for activities of the promyelocytic leukemia zinc finger-retinoic acid receptor- fusion protein. Proc Natl Acad Sci U S A 1996;93:3624–9.[Abstract/Free Full Text]
21. Nagpal S, Friant S, Nakshatri H, Chambon P. RARs and RXRs: evidence for two autonomous transactivation functions (AF-1 and AF-2) and heterodimerization in vivo. EMBO J 1993;12:2349–60.[Medline]
22. Ferrara FF, Fazi F, Bianchini A, et al. Histone deacetylase-targeted treatment restores retinoic acid signaling and differentiation in acute myeloid leukemia. Cancer Res 2001;61:2–7.[Abstract/Free Full Text]
23. He LZ, Guidez F, Tribioli C, et al. Distinct interactions of PML-RAR and PLZF-RAR with co-repressors determine differential responses to RA in APL. Nat Genet 1998;18:126–35.[CrossRef][Medline]
24. Hong SH, David G, Wong CW, Dejean A, Privalsky ML. SMRT corepressor interacts with PLZF and with the PML-retinoic acid receptor (RAR) and PLZF-RAR oncoproteins associated with acute promyelocytic leukemia. Proc Natl Acad Sci U S A 1997;94:9028–33.[Abstract/Free Full Text]
25. Guidez F, Ivins S, Zhu J, Soderstrom M, Waxman S, Zelent A. Reduced retinoic acid-sensitivities of nuclear receptor corepressor binding to PML- and PLZF-RAR underlie molecular pathogenesis and treatment of acute promyelocytic leukemia. Blood 1998;91:2634–42.[Abstract/Free Full Text]

This article has been cited by other articles:

				M. Ishibashi, K. Nakayama, S. Yeasmin, A. Katagiri, K. Iida, N. Nakayama, M. Fukumoto, and K. Miyazaki A BTB/POZ Gene, NAC-1, a Tumor Recurrence-Associated Gene, as a Potential Target for Taxol Resistance in Ovarian Cancer Clin. Cancer Res., May 15, 2008; 14(10): 3149 - 3155. [Abstract] [Full Text] [PDF]

				P. J. Stogios, L. Chen, and G. G. Prive Crystal structure of the BTB domain from the LRF/ZBTB7 transcriptional regulator Protein Sci., February 1, 2007; 16(2): 336 - 342. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Puccetti, E. Articles by Ruthardt, M. Search for Related Content PubMed PubMed Citation Articles by Puccetti, E. Articles by Ruthardt, M.