Normal hematopoiesis after conditional targeting of RXR ... - CiteSeerX

Normal hematopoiesis after conditional targeting of RXR␣ in murine hematopoietic stem/progenitor cells Mercedes Ricote,*,1,2 Cynthia S. Snyder,*,1 Ho-Yin Leung,* Ju Chen,† Kenneth R. Chien,†,3 and Christopher K. Glass*,‡ Departments of *Cellular and Molecular Medicine and ‡Medicine, School of Medicine, and †Institute of Molecular Medicine, University of California, La Jolla, San Diego

Abstract: Because of the retinoic acid receptor-␣ (RAR␣) gene’s involvement in acute promyelocytic leukemia, the important role of RARs in hematopoiesis is now well established. However, relatively few studies of hematopoiesis have focused on the role of the retinoid X receptors (RXRs), the obligate heterodimeric partners of the RARs. We sought to establish whether conditional targeting of RXR␣ in early hematopoietic progenitors, ideally to the level of the hematopoietic stem cell (HSC), would compromise hematopoiesis. For hematopoietic targeting of RXR␣, we characterized IFN-inducible MxCre mice for use in studying the role of RXR␣ in hematopoiesis. We established that MxCre executes recombination of loxP-flanked RXR␣ in hematopoietic progenitors immunophenotypically enriched for HSC, leading to widespread and sustained targeting of RXR␣ in hematopoietic cells. However, we found no evidence of hematologic compromise in mice lacking RXR␣, suggesting that RXR␣ is dispensable for normal murine hematopoiesis. Nonetheless, RXR␣ null bone marrow cells cultured in methylcellulose form colonies more efficiently than bone marrow cells obtained from control mice. This result suggests that although RXR␣ is not required for murine hematopoiesis, there may be hematopoietic signaling pathways that respond selectively to RXR␣ or settings in which combined expression of RXR (␣, ␤, and ␥) is limiting. J. Leukoc. Biol. 80: 850 – 861; 2006. Key Words: myelopoiesis 䡠 nuclear receptors 䡠 retinoid X receptors 䡠 retinoic acid receptors

INTRODUCTION Differentiation of the various hematopoietic lineages from pluripotent, self-renewing hematopoietic stem cells (HSC) is a tightly controlled process involving coordinate expression of lineage-specific genes. Transcriptional regulation of this process is necessary for normal hematopoiesis, as evidenced by the frequent involvement of transcription factor genes in leukemia-associated chromosomal translocations [1]. Nuclear receptors (NRs) are a family of transcription factors that bind cognate DNA sequences and activate transcription in a ligand850

Journal of Leukocyte Biology Volume 80, October 2006

dependent manner [2]. The important role of NRs in hematopoiesis was highlighted by the discovery that chromosomal translocations involving the gene for the retinoic acid receptor-␣ (RAR␣) lead to acute promyelocytic leukemia (APL) in humans [3]. Furthermore, treatment of APL patients with RA, a ligand for RARs, leads to remission of disease in the majority of patients [4]. These findings paved the way for a significant body of research examining the roles of retinoids and their receptors in hematopoietic, particularly granulocytic, differentiation [5]. Retinoid signaling is transduced by two families of NRs, the aforementioned RARs and the retinoid X receptors (RXRs), each family consisting of three members (␣, ␤, and ␥) [6 –9]. The RARs heterodimerize with RXRs to bind RA response elements, cis-acting DNA sequences found within the promoters of retinoid-responsive genes, and to modulate transcription by interactions with coactivators, corepressors, and other components of the transcriptional machinery [10, 11]. Like RARs, RXRs contain a central DNA binding domain (DBD) and amino- and carboxy-terminal activation domains termed, respectively, AF1 and AF2. The AF2 domain of the retinoid receptors includes the RA binding pocket. All-trans RA (ATRA) binds selectively to RARs, whereas 9-cis RA (9c-RA) binds RARs and RXRs [12, 13]. Many studies addressing the roles of the retinoid receptors in hematopoietic differentiation have used cell lines that differentiate in response to RA treatment or have involved the overexpression of wild-type or mutant retinoid receptors [14 – 20]. More recently, the creation of mice with targeted deletions of the RARs has allowed the study of retinoid signaling in primary hematopoietic cells and in vivo [21–25]. These studies have suggested that RARs are not necessary for granulopoiesis. However, when present, RARs modulate the rate of granulocytic differentiation according to the availability of ligand; liganded RARs accelerate differentiation, whereas unliganded RARs decelerate granulocyte differentiation [26]. Because of the clear associations of the RARs with granulopoiesis, most

1

These authors contributed equally to this work. Correspondence and current address: Centro Nacional de Investigaciones Cardiovasculares (CNIC), Melchor Ferna´ndez Almagro, 3, Madrid 28029, Spain. E-mail: [email protected] 3 Current address: MGH Cardiovascular Research Center and Harvard Stem Cell Institute, Harvard Medical School, Boston, MA. Received February 14, 2006; revised May 17, 2006; accepted June 12, 2006; doi: 10.1189/jlb.0206097. 2

0741-5400/06/0080-850 © Society for Leukocyte Biology

studies examining the role of retinoid signaling in hematopoiesis have focused on the RAR family of retinoid receptors. In this context, RXRs have generally been viewed as the obligate but passive heterodimeric partners to the RARs. However, in contrast to RARs, RXRs can modulate gene transcription as homodimers [27]. In addition, RXRs serve as the heterodimeric partners of the vitamin D receptor, the thyroid hormone receptor, the peroxisome proliferator-activated receptors, and other members of the NR family [28 –31]. Thus, RXRs would be expected to have overlapping functions with RARs but are poised to have an even more widespread influence on hematopoiesis than RARs. Despite this, relatively few studies of NR signaling in hematopoiesis have focused on RXRs [14, 15, 32]. The three members of the RXR family show tissue-specific differences in expression [9]. Previous studies have suggested that the most abundant, or at least the most functionally important, RXR in myeloid cells is RXR␣ [14, 33–35]. To better understand the role of RXRs in various physiological processes, several groups have generated murine knockouts of RXRs [36]. RXR␥ null mice develop normally [37]. Roughly half of RXR␤ knockout mice die before or at birth; survivors appear normal, except that males are sterile [38]. Conventionally targeted RXR␣ knockout mice are embryonic-lethal [39, 40]. To study the roles of RXR␣ in adult mice, we are using mice in which RXR␣ has been conditionally targeted for tissuespecific inactivation [41]. When crossed with Cre recombinase transgenic mice, RXR␣ is inactivated in those tissues in which Cre recombinase is expressed. We hypothesized that inactivating RXR␣ in early hematopoietic progenitors would compromise hematopoiesis and that we would find overlapping (with RARs) and nonoverlapping defects in hematopoiesis, reflecting the role of RXR␣ as a heterodimeric partner to several NR family members and its ability to modulate gene expression as TABLE 1. Name P1 P2 P3 Cre1 Cre2 P4 P5 P6 P7 P8 P9 P10 RXRA-qFor RXRA-qRev RXRB-qFor RXRB-qRev RXRG-qFor RXRG-qRev Cyclo-For Cyclo-Rev Cyclo-qFor Cyclo-qRev

a homodimer. For targeting the expression of Cre recombinase to the hematopoietic system, we obtained MxCre transgenic mice [42]. Early reports using MxCre to accomplish widespread and sustained inactivation of gene targets in hematopoietic cells had yielded inconsistent and sometimes conflicting results (unpublished data and private communications). Therefore, we also sought to systematically characterize MxCre mice for our use in studying the role of RXR␣ in hematopoietic cells.

MATERIALS AND METHODS Mice Conditionally targeted RXR␣ mice (RXR␣fl/fl) and MxCre transgenic mice, which have been described previously [41, 42], were crossed. Cre recombinase-positive offspring (MxCre⫹RXR␣fl/⫹) were then crossed with RXR␣fl/fl mice to generate Cre-positive and Cre-negative mice homozygous for the loxP-flanked (floxed) RXR␣ allele, designated MxCre⫹/RXR␣fl/fl and MxCre⫺/RXR␣fl/fl (RXR␣fl/fl), respectively. MxCre⫹/RXR␣fl/fl and RXR␣fl/fl mice were interbred to generate experimental and control animals. Mice were genotyped by PCR using primers P1, P2, and P3, directed against RXR␣ [41], and primers Cre1 and Cre2, directed against Cre. (Primers used in this study are shown in Table 1.) To induce MxCre expression, MxCre⫹/RXR␣fl/fl and RXR␣fl/fl littermate controls received a single induction regimen consisting of three i.p. injections of 300 ␮g poly-inosine, poly-cytosine (pIpC; Sigma Chemical Co., St. Louis, MO), administered every other day. Mice used in these experiments were 3–5 months of age at the time of induction. Blood and tissues were harvested for analysis after an interval of 1 week–2 years after induction. Peripheral blood was drawn into EDTA-coated CapiJect威 T-MQK tubes (Terumo Medical Corp., Elkton, MD). Complete blood profiles were obtained using a Melet Schloesing MS9/-5 (El Cajon, CA). Prior to tissue harvest, animals were killed by carbon dioxide (CO2) asphyxiation according to institutional guidelines. Bone marrow cells were collected from bilateral femurs and tibiae by flushing marrow cavities with PBS through a 25-gauge needle.

Primers

Sequences (5⬘ to 3⬘) ACCAAGCACATCTGTGCTATCT CAACTGTATACCCCATAGTGTT ATGAAACTGCAAGTGGCCTTGA AGGTGTAGAGAAGGCACTTAGC CTAATCGCCATCTTCCAGCAGG TTTTTCCATTGGGTTGGGTA GAATTCTAGAGGATCAGCTTGGG GGACCGCTCCTCAGGTAGGAAT GCTGGACTCTGATCCTCCCTGT CACTATTCTCGGCTCAGGCTT TGCCCATCCCTCAGGAAATATGG TGTTTGCCTCCACGTATGTCTC TAGTCGCAGACATGGACACC GTTGGAGAGTTGAGGGACGA GCACAGAAACTCAGCCCATT CATCCTCATGTCACGCATTT GCCTGGGATTGGAAATATGA ACACCGTAGTGCTTCCCTGA ACGCCACTGTCGCTTTTC GTAAAATGCCCGCAAGTCAA ACAGGTCCTGGCATCTTGTC CATGGCTTCCACAATGTTCA

Purpose genotyping of mice genotyping of mice genotyping of mice/bone marrow genotyping of mice genotyping of mice genotyping of bone marrow genotyping of bone marrow genotyping of bone marrow genotyping of bone marrow genotyping of bone marrow RT-PCR RT-PCR qPCR qPCR qPCR qPCR qPCR qPCR generate qPCR control template generate qPCR control template qPCR qPCR

qPCR, Quantitative PCR.

Ricote et al. Conditional gene targeting of RXR␣

851

Southern blot and DNA PCR analyses DNA was harvested using the Wizard威 genomic DNA purification kit (Promega, Madison, WI). Southern blot analysis was performed according to standard methodologies using BamHI and BglII for restriction digests [43]. The probe for Southern blot analysis and the PCR conditions for genotyping of tissues have been described previously [41]. To genotype flow-sorted cells, we designed a nested multiplex PCR strategy. To harvest genomic DNA from flow sorted cells, cell pellets consisting of 8 –10,000 cells were digested 2 h at 55°C in 30 ␮l PBS containing 0.1 mg/ml Proteinase K (Invitrogen, Carlsbad, CA). Following a 15-min 95°C incubation, 1 ␮l of each digest reaction was used in a 50-␮l first-round PCR reaction containing primers P3, P6, and P7, followed by a second round of PCR using primers P4, P5, and P8. Cycling conditions were 1 min at 95°C; 12 cycles consisting of 30 s at 94°C, 30 s at 66°C (0.5°C decrement per cycle), and 30 s at 72°C; 18 cycles consisting of 30 s at 94°C, 30 s at 62°C, and 30 s at 72°C; and 3 min at 72°C followed by a 4°C hold.

Magnetic cell sorting and FACS Bone marrow cells were enriched for lineage-depleted cells using the StemSep™ murine progenitor enrichment cocktail and StemSep™ device (Stem Cell Technologies, Vancouver, BC, Canada), according to the manufacturer’s instructions. Lineage-depleted cells were incubated with additional enrichment cocktail, washed in PBS/2% FBS, and incubated with PE-labeled strepavidin (PE-SAv) and FITC-labeled anti-stem cell antigen-1 (Sca-1; Ly-6A/E, BD Biosciences, San Jose, CA). Following a second wash, the cells were analyzed and sorted using a FACSVantage SE and CellQuest™ software (BD Biosciences). Freshly harvested bone marrow cells were analyzed by FACS. FITC-labeled (CD3, CD71, and GR-1) and PE-labeled (CD19, Ter119, and F4/80) antibodies were all obtained from PharMingen (San Diego, CA). A FACSCalibur flow cytometer and CellQuest™ software (BD Biosciences) were used for analysis. Propidium iodide was added to allow exclusion of dead cells from the analysis.

Cell culture Bone marrow cells were cultured in preformulated methylcellulose medium (Methocult 3434, Stem Cell Technologies). Cells suspended in 3 mL methylcellulose medium were plated on 35 mm dishes and cultured at 37°C in a humidified, 5% CO2 incubator, and colonies were counted starting at Day 5. When present, 9c-RA (Sigma Chemical Co.) was used at a final concentration of 1 ␮M.

RT-PCR and qPCR RNA was isolated from whole bone marrow using TRIzol (Invitrogen), DNase I-digested, and column-purified (RNeasy MinElute Cleanup; Qiagen, Valencia, CA). First-strand cDNA was synthesized using SuperScript II and random hexamers (Invitrogen). To analyze splicing of the RXR␣ transcript, cDNA was PCR-amplified using Primers P9 and P10. For quantitation of RXR␣, RXR␤, and RXR␥ expression, qPCR was performed using the ABI 7300 real-time PCR system (Applied Biosystems, Foster City, CA). Linearized, gel-purified, plasmid templates (pCMX-RXR␣, -RXR␤, and -RXR␥) were used as standards to quantitate RXR transcript levels. A column-purified (QIAquick PCR purification; Qiagen), 555-bp PCR amplicon served as a standard to quantitate cyclophilin transcript levels. Standards were quantitated by UV spectroscopy. For each standard template, quantitation was performed in triplicate at two different dilutions and on two separate UV spectrometers; the results were averaged. Template copy numbers were calculated based on molecular weights and concentrations of RXR and cyclophilin template controls. All amplification reactions were performed in a 20-␮l reaction volume consisting of 10 ␮l 2⫻ SYBR威 GREEN PCR master mix (Applied Biosystems), 0.24 ␮l each primer (300 nM final), and cDNA or control templates suspended in 9.52 ␮l H2O. For bone marrow samples, the cDNA equivalent of 50 ng reverse-transcribed total RNA served as template for each qPCR reaction. For standards, a dilution series of 10 copies–1 billion copies of template per reaction was amplified to establish the linear amplification range for each primer set. Thermal cycling was initiated for 15 min at 95°C, followed by 50 cycles consisting of 10 s at 94°C, 20 s at 56°C, and 30 s at 72°C. Readings were taken during the 72°C step of each cycle. Amplification cycles were followed by melting curve analysis. Primers designed for each RXR isoform were challenged in mock amplification reactions to confirm that

852

Journal of Leukocyte Biology Volume 80, October 2006

no amplification occurred from the other two isoform templates under the qPCR conditions used. Setups and amplifications of bone marrow samples and template controls were all performed in triplicate and were amplified simultaneously on a single 96-well plate to allow for valid, comparative quantitation. In every experiment, “No RT” controls and “No Template” controls were included.

Quantitation of RXR␣, RXR␤, and RXR␥ transcripts qPCR results were analyzed using sequence detection system software (SDS 1.2, Applied Biosystems). Standard curves were generated by plotting cycle thresholds (CT) against the logarithm of the starting template amount. All primers showed linear amplification over at least an 8-log range of template concentration. CT of bone marrow samples were determined for each amplified target, and the corresponding standard curves were used to convert CT values to transcript copy numbers. As the efficiency of RT is highly variable, quantitative comparisons were only made between qPCR results generated from a single RT reaction. RXR transcript levels were normalized to cyclophilin. For quantitation of RXR␣, RXR␤, and RXR␥ expression in RXR␣fl/fl (wild-type) bone marrow and for comparison of RXR␤ and RXR␥ isoform levels in pIpC-treated MxCre⫹/RXRfl/fl mice and RXR␣fl/fl control mice, RNA was isolated from two mice of each type. Duplicate RT reactions were performed on each isolate, and each resulting cDNA mixture was used in qPCR.

Western blot analysis Western blot analysis was done using standard procedures [43]. Total protein was extracted from bone marrow in SDS sample buffer, resolved by SDS-PAGE and transferred to polyvinylidene difluoride filters. To detect RXR␣, RXR␤, and RXR␥ proteins, rabbit polyclonal antibodies directed against RXR␣ (D-20), RXR␤ (L-20), and RXR␥ (Y-20) were incubated at a dilution of 1:200 for 1 h at room temperature in PBS containing 5% nonfat milk and 0.1% Tween. Primary antibodies were obtained from Santa Cruz Biotechnology Inc. (CA). Secondary antibodies were obtained from Dako (Carpinteria, CA). Immunoreactive proteins were detected using chemiluminescence (Pierce, Rockford, IL).

RESULTS MxCre mediates widespread and sustained loxP recombination in hematopoietic tissues and in bone marrow enriched for HSC To study the role of RXR␣ in hematopoiesis, we first assessed the extent of MxCre-mediated loxP recombination (floxout) in adult hematopoietic tissues. Relative levels of the RXR␣flox and ⌬RXR␣ alleles in hematopoietic tissues were analyzed by PCR and by Southern blotting using the strategy shown in Figure 1A. PCR and Southern blot analyses demonstrated efficient deletion of the targeted RXR␣flox allele and appearance of the ⌬RXR␣ null allele in MxCre⫹/RXR␣fl/fl mice (Fig. 1, B and C). Bone marrow showed the highest level of deletion (88.1%), spleen showed less-efficient gene deletion (67.5%), and thymus showed the least (44.1%). Among other tissues examined, the liver demonstrated the highest level of gene deletion (data not shown). Although the liver serves as a hematopoietic organ during fetal life, it is not a site of ongoing hematopoiesis in the normal adult mouse. Histologic examination confirmed the absence of ongoing hematopoiesis in the livers of MxCre⫹/RXR␣fl/fl and RXR␣fl/fl mice (data not shown). As expected, pIpC treatment alone produced no effect on the floxed alleles in RXR␣fl/fl control mice lacking the MxCre transgene. http://www.jleukbio.org

Fig. 1. Efficiency of MxCre-mediated loxP recombination in hematopoietic tissues. (A) Layout of RXR␣ gene depicting the intron-exon structures of the RXR␣flox (targeted) and ⌬RXR␣ (recombined) alleles and the positions of the loxP sites. In the targeted allele, the fourth exon of RXR␣, containing the DNA-binding domain, is flanked by loxP sites (floxed). After Cre-mediated recombination, Exon 4 is deleted (⌬RXR␣). The primers used in PCR genotyping, the restriction enzyme sites (B, BamHI; Bg, BglII) used for digests, and the intronic probe used for Southern blot analysis are shown. Resulting amplicon and digest band sizes are indicated for each allele. (B) Southern blot and (C) PCR analyses of genomic DNA isolated from hematopoietic tissues of RXR␣fl/fl control mice and conditionally targeted MxCre⫹/RXR␣fl/fl mice 8 –9 days post-pIpC induction. Results shown are from thymus (T), spleen (S), and whole bone marrow (B) DNA isolates. (B) For Southern blot analysis, DNA was digested with BamHI and BglII, and the resulting blot probed using the Intron 2 probe shown. Positions of the 3.3-kb RXR␣flox and the 4.5-kb ⌬RXR␣ bands are indicated. Southern blots were quantified using Image Gauge 4.0 (Fuji Photo Film Co., Ltd., Tokyo, Japan). Percentages of each the RXR␣fl/fl and ⌬RXR␣ alleles, for a combined total of 100%, are shown below each corresponding lane. (C) Primers P1, P2, and P3 generate amplicons of 768 bp (P1/P2) or 441 bp (P1/P3) when used in PCR to amplify RXR␣flox or ⌬RXR␣ DNA, respectively.

Despite high levels of gene deletion in whole bone marrow, a small percentage (11.9%) of the original RXR␣flox allele remained. This level of gene deletion in bone marrow is comparable with other published results [44, 45]. In adult mice, whole bone marrow consists primarily of granulocytic and erythroid precursors in various stages of terminal differentiation. Therefore, it is reasonable to conclude that MxCre-mediated gene deletion is efficient in these maturing cells. However, to study a lasting effect of gene deletion on hematopoietic differentiation or to assess effects on early stages of differentiation, it is important to establish that early progenitors also show efficient gene deletion. Percentage estimates of early bone marrow progenitors depend on the defining criteria used to identify these progenitors. When functional criteria are used to define HSC, estimates drop to 0.05% or fewer of all bone marrow cells (Fig. 2A) [46, 47]. Efforts to pinpoint HSC by cell surface immunophenotyping have shown that in mice, HSC are highly enriched in bone marrow sorted for lineage-negative cells expressing the Sca-1 antigen (Lin–/Sca-1⫹) [48 –50]. Bone marrow cells from MxCre⫹/RXR␣fl/fl mice were immunophenotypically labeled and analyzed by FACS; cells were

sorted into Lin⫹/Sca-1–, Lin⫹/Sca-1⫹, and Lin–/Sca-1⫹ (HSCenriched) populations (Fig. 2B). DNA was genotyped using a nested, multiplex PCR strategy (Fig. 2C). When analyzed on agarose gels, first-round PCR products from whole BM and post-column DNA isolates showed visible amplification of the 601-bp and 409-bp amplicons derived from the RXR␣flox and ⌬RXR␣ alleles, respectively, whereas no amplification could be detected in DNA isolates from the sorted cell populations (data not shown). Aliquots of the first-round reactions were used in a second round of PCR. Products from both secondround reactions are shown in Figure 2D. The upper image depicts the results obtained when all three second-round primers were used, and the lower image demonstrates the findings obtained when primer P5 was excluded from the reaction mixture. When the second-round conditions allowed for simultaneous amplification of the RXR␣flox and ⌬RXR␣ alleles (Fig. 2D, upper image), all marrow samples obtained from MxCre⫹/ RXR␣fl/fl mice demonstrated a strong band at 287 bp, corresponding to the ⌬RXR␣ null allele. A weak 375-bp band corresponding to the residual RXR␣flox allele was present in the whole BM and the post-column sample lanes; when reacRicote et al. Conditional gene targeting of RXR␣

853

Fig. 2. MxCre-mediated loxP recombination in bone marrow enriched for HSC. (A) Schematic of a hematopoietic maturation pyramid, showing the approximate relative numbers of HSC, progenitor cells, and maturing cells in bone marrow. As a result of ongoing cell maturation and the continuous flushing out of maturing cells from the bone marrow to the peripheral circulation, the majority of hematopoietic cells present within the bone marrow at any given moment is soon replaced by the small subpopulation of progenitor (⬍5%) and stem (⬍0.05%) cells which are also present. The genotype of this small subpopulation of stem/progenitor cells will determine the eventual genotype of the bone marrow. To genotype this small subpopultion of cells, an enrichment strategy was necessary. CMP, Common myeloid progenitor; GMP, granulocyte/monocyte progenitor; MEP, megakaryocyte/erythrocyte progenitor; CLP, common lymphoid progenitor. (B) Strategy for enrichment of HSC. Triplicate pools of bone marrow from pIpC-induced MxCre⫹/RXR␣fl/fl mice were immunophenotypically labeled using a cocktail of biotinylated, mAb directed against lineage-specific antigens (CD5, CD45R, CD11b, GR-1, 7– 4, and TER119) and subjected to immunomagnetic depletion of lineage-positive cells (Lin⫹). Whole bone marrow (Whole BM) and lineage-depleted (Post Column) whole bone marrow were serially incubated with fresh aliquots of biotinylated lineage antibody cocktail, with PE-SAv-labeled secondary antibody and with FITC-labeled anti-Sca-1. The cells were then analyzed by FACS. Lin⫹/Sac-1–, Lin⫹/Sca-1⫹, and Lin–/Sca-1⫹ cell populations were sorted for subsequent DNA isolation and PCR genotyping. Sort gates and the percentages of cells within these gates in whole bone marrow and in lineage-depleted bone marrow are indicated. (C) To facilitate detection of small numbers of flow-sorted bone marrow cells, a highly sensitive, nested multiplex PCR strategy using P3 in combination with the newly designed primers P4, P5, P6, P7, and P8 was performed. The first round of PCR used primers P3, P6, and P7. Following column purification, a volume equivalent to 1 ␮l of the original 50-␮l first-round reaction volume was subjected to a second round of PCR using the primers P4, P5, and P8. Amplification products were analyzed by agarose gel electrophoresis. (D) For conditionally targeted MxCre⫹/RXR␣fl/fl mice, DNA isolated from whole bone marrow; lineage-depleted, presorted (Post Column) bone marrow; and three FACS cell populations was subjected to PCR. DNA isolated from the whole bone marrow of RXR␣fl/fl control mice served as a control for the PCR. First-round PCR products (not shown) generated using primers P3, P6, and P7 were subjected to one of two alternative second rounds of PCR, one using the nested primers P4, P5, and P8 (products shown upper gel image) and one using primers P4 and P8 only (products shown lower gel image) to selectively amplify residual RXR␣flox allele. First- and second-round PCR reactions were repeated several times, and the entire procedure of bone marrow harvest followed by cell sorting and PCR analysis was repeated twice.

tion conditions were changed to favor amplification of the RXR␣flox allele, the presence of this 375-bp band was accentuated (Fig. 2D, lanes 2 and 3). In contrast, the sorted cell populations, including the Lin–/Sca-1⫹ subset enriched for HSC, showed complete absence of the RXR␣flox allele. There854

Journal of Leukocyte Biology Volume 80, October 2006

fore, in immunophenotypically defined hematopoietic progenitors highly enriched for HSC, only the ⌬RXR␣ null allele remains after pIpC induction of MxCre⫹/RXR␣fl/fl mice. HSC divide to produce the full repertoire of mature hematopoietic cells, but they are also, by definition, self-renewing. http://www.jleukbio.org

Fig. 3. Longevity of MxCre-mediated loxP recombination in hematopoietic tissues. The diagram indicates the timeline of experiments. Mice were given 3 alternate-day i.p. injections of pIpC to induce expression of Cre recombinase. Tissues were harvested at several time-points following this single MxCre induction regimen, with Day 0 defined as the day of the first pIpC injection. DNA PCR was performed to assess the relative abundance of RXR␣flox and ⌬RXR␣ alleles in spleen (S) and in bone marrow (B) at 1 month, 3 months, 1 year, and 23 months after MxCre induction of loxP recombination. PCR results for control mice lacking the MxCre transgene (RXR␣fl/fl) are shown, for comparison, only from the 1-month time-point. Primers P1, P2, and P3 (Fig. 1A) were used in these PCR reactions. (*, PCR data from bone marrow harvested on Days 8 –9 is shown in Fig. 1C.)

Therefore, if MxCre induction results in RXR␣ gene deletion in HSC; then this deletion should endure over time, unless there is selection pressure against it. To confirm this prediction, we rechecked the mice for levels of gene deletion at several time-points after a single pIpC induction regimen. As shown in Figure 3, bone marrow and spleen harvested from mice killed 1 month, 3 months, 1 year, and nearly 2 years after pIpC induction retain the ⌬RXR␣ null allele. These findings confirm the stability of bone marrow floxout over the normal murine lifespan and support the hypothesis that MxCre induction executes floxout to the level of the HSC.

Expression of RXR␣, RXR␤, and RXR␥ in MxCre⫹/RXR␣fl/fl mice and RXR␣fl/fl controls To look for RNA transcripts arising from the RXR␣flox and ⌬RXR␣ alleles, we performed RT-PCR using primers situated in Exons 3 and 5 of RXR␣ (Fig. 4A). As shown in Figure 4B, RT-PCR of RNA isolated from the whole bone marrows of MxCre⫹/RXR␣fl/fl mice yielded a 221-bp product corresponding to juxtaposition of Exons 3 and 5 following Cre recombination, whereas RNA obtained from the bone marrow of RXR␣fl/fl mice yielded a 401-bp band corresponding to the normal juxtaposition of Exons 3, 4, and 5. To determine if MxCre⫹/RXR␣fl/fl mice continue to express RXR␣ protein following Cre recombination, we performed Western blot analysis on whole cell extracts of bone marrows from RXR␣fl/fl and MxCre⫹/RXR␣fl/fl mice. Western blots show that a 55-kDa protein, corresponding to full-length RXR␣, is expressed in bone marrow from RXR␣fl/fl control mice but not in bone marrow from MxCre⫹/RXR␣fl/fl mice (Fig. 4C). It is of interest that we detected the appearance of a novel protein in bone marrow from MxCre⫹/RXR␣fl/fl mice that corresponds to the expected size for a shortened form (⌬Exon 4) of RXR␣. This short-form RXR␣ was not seen in other tissues or in peritoneal macrophages, wherein expression of RXR␣ is abundant (data not shown), suggesting that this short form RXR␣ may be unstable. Furthermore, deletion of RXR␣ Exon 4, which encodes the DBD, would prevent RXR␣ from binding DNA and activating transcription. Previous studies had suggested that RXR␣ is the predominant RXR in hematopoietic cells [14, 33–35]. To assess the relative expression of the three RXR isoforms in bone marrow and to determine if deletion of RXR␣ results in a compensational up-regulation of RXR␤ or RXR␥, we performed qPCR

on total RNA and Western blot analysis on whole cell extracts of MxCre⫹/RXR␣fl/fl and RXR␣fl/fl control bone marrows. Our qPCR data showed that RXR␣ and RXR␤ transcripts are comparably expressed in whole bone marrow of RXR␣fl/fl control mice; RXR␣ showed only slightly higher expression than RXR␤ (Fig. 5A). Western blot analysis demonstrated detectable levels of RXR␤ protein in bone marrows of MxCre⫹/ RXR␣fl/fl mice and RXR␣fl/fl controls (Fig. 5B, upper blot). There was no compensatory increase in expression of RXR␤ or RXR␥ following conditional targeting of RXR␣ (Fig. 5, A and B). Expression of RXR␥ was extremely low in bone marrow and virtually undetectable in most isolates by Western blot analysis and qPCR (Fig. 5, A and B).

Hematopoietic consequences of RXR␣ floxout: in vivo analysis Using the MxCre system, we proceeded to investigate the role of RXR␣ in hematopoiesis. We harvested peripheral blood from MxCre⫹/RXR␣fl/fl mice and littermate controls at multiple time-points after pIpC induction and performed complete blood profiles. Peripheral counts were within the normal range for both groups of mice (Table 2). No significant differences were observed between the red cell or white cell counts of MxCre⫹/RXR␣fl/fl mice and RXR␣fl/fl controls. Peripheral blood differentials revealed comparable absolute lymphocyte, neutrophil, eosinophil, and monocyte counts. Bone marrows from MxCre⫹/RXR␣fl/fl mice and RXR␣fl/fl controls demonstrated normal cellularity and a polymorphous mixture of maturing hematopoietic precursors, including the full spectrum of immature and mature neutrophilic precursors (Fig. 6A), consistent with the normal counts seen in the peripheral blood. Bone marrow differential counts showed no significant difference between MxCre⫹/RXR␣fl/fl and RXR␣fl/fl bone marrows (Table 2). The spleens of both groups of mice showed maintenance of normal splenic architecture with small foci of residual, ongoing hematopoiesis concentrated in subcapsular regions (Fig. 6A). FACS analysis performed using a panel of antibodies directed against lymphoid, erythroid, and myeloid antigens demonstrated comparable expression of B- and Tlymphoid (CD19 and CD3, respectively), erythroid (Ter119 vs. CD71), and myeloid (F4/80 vs. GR-1) antigens in MxCre⫹/ RXR␣fl/fl and RXR␣fl/fl control bone marrows (Fig. 6B). No significant hematologic abnormalities were observed in either Ricote et al. Conditional gene targeting of RXR␣

855

A

401 bps P10

P9

RXRα αflox Ex3

Ex4

Ex5

221 bps P9

P10

Ex3

Ex5

RXRα

xC M

B

+

re

fl fl/

α XR

fl fl/

α XR

/R

R

(RXRαflox )

Fig. 5. Relative expression of RXR␣, RXR␤, and RXR␥ in bone marrow. (A) Bone marrow transcript levels of RXR isoforms were assessed by quantitative RT-PCR. The graph shows expression levels of RXR transcripts normalized to expression of cyclophilin A, as explained in Methods. Relative expression levels for RXR␣ (alpha), RXR␤ (beta), and RXR␥ (gamma) were assessed using bone marrow isolated from two RXR␣fl/fl control (solid black and black/white horizontally-striped bars) mice and two conditionally targeted MxCre⫹/RXR␣fl/fl (solid grey and grey/white diagonally-striped bars) mice. The asterisks indicate that the effects of inactivating RXR␣ in conditionally targeted MxCre⫹/RXR␣fl/fl mice are shown in Figure 4. (B) Western blot of total protein isolated from bone marrows of RXR␣fl/fl and MxCre⫹/RXR␣fl/fl mice. The upper and lower gel images show Western blot results for RXR␤ and RXR␥, respectively. The positive control used for RXR␤ is recombinant RXR␤. The positive control shown for RXR␥ is a dilution series of in vitro-transcribed and translated RXR␥. The RXR␥ blot shown represents a long exposure time.

401 bps

( RXRα) 221 bps

xC M

C

+

re

fl fl/

α XR

fl fl/

α XR

/R

R 64 kD RXRα 50 kD

Fig. 4. Expression of RXR␣flox and ⌬RXR␣ alleles in bone marrow. (A) Predicted exon splicing for the RXR␣flox and ⌬RXR␣ alleles. The primers used for RT-PCR and the expected amplicon sizes are indicated. (B) RT-PCR analysis of RNA and (C) Western blot analysis of protein isolated from bone marrow. (B) When used in PCR to amplify reverse-transcribed RNA from RXR␣fl/fl control mice and from conditionally targeted MxCre⫹/RXR␣fl/fl mice, Primers P9 and P10 generate amplicons of 401 bp or 221 bp, respectively. (C) Western blot of total protein isolated from bone marrows of RXR␣fl/fl and MxCre⫹/ RXR␣fl/fl mice. The primary antibody is directed against the N terminus of RXR␣. The antibody detects proteins corresponding to full-length RXR␣ in RXR␣fl/fl control mice and to a novel, short form (presumably ⌬Exon 4) RXR␣ in MxCre⫹/ RXR␣fl/fl mice. Full-length RXR␣ protein is not detected in MxCre⫹/RXR␣fl/fl mice.

856

Journal of Leukocyte Biology Volume 80, October 2006

group of mice. These findings suggest that RXR␣ is dispensable for normal murine hematopoiesis.

Hematopoietic consequences of RXR␣ floxout: effects on in vitro colony formation We performed colony assays on bone marrow cells isolated from MxCre⫹/RXR␣fl/fl mice and from RXR␣fl/fl controls to assess for differences in colony formation. Bone marrow was cultured in methylcellulose culture medium supplemented with hematopoietic cytokines and 9c-RA or vehicle (ethanol). Colony counts were performed after 5 days of culture. Previous studies have shown that addition of RAR ligands (ATRA, 9c-RA) to methylcellulose culture medium suppresses colony formation by normal bone marrow cells [51, 52]. We consistently observed colony suppression by 9c-RA in bone marrow cultures of MxCre⫹/RXR␣fl/fl mice and RXR␣fl/fl controls (Fig. 7). In addition, bone marrow cells obtained from MxCre⫹/ RXR␣fl/fl mice demonstrated a modest but consistent enhancement of colony formation as compared with bone marrow cells obtained from RXR␣fl/fl controls (Fig. 7). The most consistent http://www.jleukbio.org

TABLE 2.

Hematologic Analysis of MxCre⫹/RXR␣fl/fl Mice and RXR␣fl/fl Controls Peripheral blood fl/fl

RXR␣ (n⫽10) White blood cell count, 103/␮l Myeloid counts, 103/␮l or (% myeloid) Neutrophils Metamyelocytes Myelocytes Promyelocytes Myeloblasts Eosinophils Monocytes Erythroid (nRBCs) Lymphocytes, 103/␮l or (per 100 myeloid) Hemoglobin, g/dl Hematocrit, % Red blood cell count, 106/␮l

⫹

Bone marrow fl/fl

MxCre /RXR␣ (n⫽10)

11.3 ⫾ 3.0

12.5 ⫾ 3.9

2.2 ⫾ 1.0 0 0 0 0 0.4 ⫾ 0.2 1.0 ⫾ 0.7 0 7.8 ⫾ 3.1 14.4 ⫾ 0.6 43.0 ⫾ 1.7 9.1 ⫾ 0.5

2.5 ⫾ 0.7 0 0 0 0 0.3 ⫾ 0.5 1.1 ⫾ 0.5 0 8.5 ⫾ 3.5 13.3 ⫾ 1.4 40.6 ⫾ 4.3 8.8 ⫾ 1.0

RXR␣ (n⫽3)

MxCre⫹/RXR␣fl/fl (n⫽3)

(36.4⫾3.1)* (17.0⫾3.7) (9.9⫾2.1) (7.5⫾1.1) (4.0⫾0.9) * (6.4⫾1.1) (18.8⫾3.8) (25.9⫾10.5)

(32.3⫾4.4)* (17.6⫾2.0) (9.7⫾2.5) (7.3⫾0.6) (3.5⫾1.3) * (6.9⫾1.7) (22.8⫾10.0) (28.0⫾6.7)

fl/fl

Numbers shown are the averages ⫾ SD (n equals the number of mice in each group). Peripheral blood differentials are based on 200 cell counts and are shown as absolute counts (⫻103/␮l). Bone marrow differentials are based on 500 myeloid (nonlymphid) cell counts and are shown as percent myeloid cells (%). Bone marrow lymphocytes are shown as numbers of lymphocytes per 100 myeloid (erythroid and granulocytic) cell counts. Megakaryocytes were excluded from the differential cell count. (* For bone marrow differential counts, eosinophilic precursors were not distinguished from neutrophilic precursors. Segmented and band form granulocytes, neutrophilic and eosinophilic, were counted as “neutrophils.”) nRBCs, nucleated RBCs.

findings were obtained from cultures counted on Day 5 of culture. Counts obtained at later time-points, particularly those obtained after 8 days of culture, became more variable between experiments (as described previously by Purton et al. [53]). In colony counts performed on Day 5, the enhancement of colony formation by bone marrow from MxCre⫹/RXR␣fl/fl mice ranged from 25% to 40% as compared with colony formation by RXR␣fl/fl control bone marrow.

DISCUSSION Much of what is known about the roles played by various transcription factors in normal hematopoiesis has come from studies using conventional gene-targeting in mice. However, conventional gene-targeting of key hematopoietic regulators frequently leads to embryonic lethal phenotypes. In this context, embryonic lethality often reflects failure to progress from primitive to definitive hematopoiesis. Although fetal liver can provide a source of hematopoietic cells for in vitro studies of hematopoiesis, this system is less than ideal, as there are significant differences between fetal and adult hematopoiesis. Furthermore, conventional gene-targeting yields ample opportunity for developmental compensations that could obscure a gene’s role under normal conditions. For these reasons, conditional gene-targeting is gaining favor as the targeting method of choice for studying a gene’s role in adult-type hematopoiesis. There are currently several lines of Cre-expressing mice designed to target Cre recombinase expression to specific hematopoietic lineages within the hematopoietic compartment. However, the MxCre system is often used as the method of choice for targeting Cre recombinase expression to the whole hematopoietic system. As early reports using MxCre to accomplish widespread and sustained inactivation of gene targets in

the hematopoietic system had yielded inconsistent results, we wished to establish the use of this system for studying the role of RXR␣ in hematopoiesis. Our findings confirm that MxCre mediates efficient recombination of floxed RXR␣ throughout the hematopoietic system. Furthermore, we demonstrate using a sensitive, nested PCR strategy that recombination is complete in bone marrow cells highly enriched for HSC. Consistent with these findings, we show that the recombined RXR␣ allele endures in bone marrow over the lifespan of the mice after a single induction. These findings validate the use of MxCre transgenic mice for conditional targeting of genes within the hematopoietic compartment and are consistent with the recent, important findings of others [45, 54]. Nonetheless, reports of failures to achieve a sustained floxout in hematopoietic cells when using MxCre are undoubtedly legitimate [55, 56], highlighting at least three interesting points. First, failure to sustain floxout of a conditionally targeted gene within the hematopoietic compartment may be evidence that this gene serves a critical role during hematopoiesis [54 –56]. Repeat testing of bone marrow for maintenance of floxout is important when using MxCre to execute hematopoietic floxout of a conditionally targeted gene. Second, groups who experience failures using MxCre for long-term hematopoietic targeting presumably achieve similarly high levels of initial HSC floxout. Therefore, the regeneration of a bone marrow compartment in which the unrecombined allele again predominates suggests the presence of another, perhaps earlier, MxCre-unresponsive precursor cell outside the immunophenotypically defined (Lin–/Sca-1⫹) HSC subset. It could be suggested that MxCre-unresponsive precursors are present within the Sca-1–/Lin– subset of bone marrow cells, the sole unsorted cell population in Figure 2C. This subset harbors cells that fail to execute complete recombination of the loxP-flanked RXR␣flox allele. (Compare unsorted bone marrow cells shown in lanes 2 and 3 to sorted cells in Ricote et al. Conditional gene targeting of RXR␣

857

Fig. 6. Effects of MxCre-mediated targeting of RXR␣ on hematopoiesis in vivo. (A) Microscopic analysis of adult hematopoietic tissues harvested from RXR␣fl/fl control mice (i–iii) and conditionally targeted MxCre⫹/RXR␣fl/fl mice (iv–vi). Images show histology of sectioned bone marrow (i, iv), bone marrow smears (ii, v), and spleen (iii, vi). Bone marrow smears were stained using Wright-Giemsa. Formalin-fixed, paraffin-embedded bone marrow and spleen sections were stained with H&E. Original magnifications, ⫻400 (i, ii, iv, v) and ⫻100 (iii, vi). Images were acquired using an Olympus BX41 microscope equipped with a Micropublisher 3.3 digital camera and processed using Adobe Photoshop software. (B) FACS analysis of freshly isolated whole bone marrow from RXR␣fl/fl and MxCre⫹/RXR␣fl/fl mice. Cells were immunophenotyped for analysis of lymphoid (CD3 vs. CD19), erythroid (CD71 vs. ter119), and myeloid (GR-1 vs. F4/80) antigen expression. A representative experiment of four is shown. Percentage averages ⫾ SD for RXR␣fl/fl versus MxCre⫹/RXR␣fl/fl mice were as follows: CD3⫹, 1.5 ⫾ 0.5 versus 1.0 ⫾ 0.3; CD19⫹, 13.5 ⫾ 2.9 versus 13.3 ⫾ 4.0; CD71⫹ter119⫹, 25.2 ⫾ 4.8 versus 22.3 ⫾ 4.0; and GR-1⫹, 41.2 ⫾ 4.3 versus 41.1 ⫾ 10.6.

lanes 4 – 6 of Fig. 2D.) However, studies have shown that primitive HSC are not contained within the Sca-1–/Lin– subset of bone marrow cells [57, 58]. Alternatively, it is possible that MxCre-unresponsive precursors are not well-sampled by bone marrow harvest, perhaps as a result of tight adherence to bone marrow stroma. Finally, the fact that we and others see maintenance of MxCre-mediated floxout in bone marrow suggests that MxCre-unresponsive precursors outside the immunophenotypically defined Lin–/Sca1⫹ subset do not normally contribute significantly to the bone marrow cell pool over the murine lifespan; perhaps a significant hematopoietic challenge is required to stimulate these otherwise quiescent precursors. As the MxCre transgene yielded widespread and stable inactivation of RXR␣, we were able to study the long-term hematopoietic consequences of inactivating RXR␣. We predicted that conditionally targeted RXR␣ mice would show defects in hematopoiesis and that we would find overlapping (with RARs) and nonoverlapping defects in hematopoiesis, reflecting the role of RXR␣ as a heterodimeric partner to several NR family members and its ability to modulate gene expression as a homodimer. However, we find that in the absence of RXR␣, murine hematopoiesis proceeds normally. 858

Journal of Leukocyte Biology Volume 80, October 2006

We see no evidence of hematologic compromise in mice lacking RXR␣. It is surprising that we are able to detect a ⌬RXR␣ protein product in bone marrow from MxCre⫹/RXR␣fl/fl mice (Fig. 4C). But given the presence of detectable short form (presumed ⌬Exon 4) RXR␣ protein in bone marrow, the absence of a hematopoietic phenotype in pIpC-treated MxCre⫹/ RXR␣fl/fl mice raises the question of whether targeting of RXR␣ by deleting its DBD results in a true null phenotype. As RXR␣ cannot function to activate gene transcription in the absence of its DBD, RXR␣ is truly null for this activity following deletion of Exon 4. Furthermore, when crossed with protamine-Cre transgenic mice, this line of conditionally targeted RXR␣fl/fl mice recapitulates the embryonic lethal phenotype observed in conventionally targeted RXR␣⫺/⫺ mice [41]. This argues that conditional targeting of RXR␣ through removal of its DBD is equivalent to conventional targeting for the production of a null phenotype. A concern raised by the presence of a short-form (DBD-deleted) RXR␣ would be that this aberrant protein could interact with normal NRs and interfere with their normal functions. Thus, dominant-negative activity by a short-form RXR␣ could be expected to yield false-positive phenotypes not attributable to absence of RXR␣. http://www.jleukbio.org

Fig. 7. Effects of MxCre-mediated targeting of RXR␣ on bone marrow colony formation. Bone marrow cells from pIpC-treated MxCre⫹/RXR␣fl/fl and RXR␣fl/fl mice were each prepared in triplicate and the bone marrows separately pooled for parallel analysis. Clonogenic culture was performed using preformulated methylcellulose medium containing cytokines (erythropoietin, IL-3, IL-6, and stem cell factor) and 9c-RA (diagonally hatched black/grey, RXR␣flfl, and grey/white, MxCre⫹/RXR␣fl/fl, bars) or ethanol (solid black, RXR␣fl/fl, and gray, MxCre⫹/RXR␣fl/fl, bars). The results shown are from a single representative experiment. Colonies were counted on the days indicated. Each data point depicts the average count ⫾ SD of five replicate plates, each of which was plated with 10,000 cells pooled from three RXR␣fl/fl control or conditionally targeted MxCre⫹/RXR␣fl/fl mice. Colony assays were performed five times, each time with from two to five replicates plated per bone marrow/ ligand combination. (P values calculated using the Student’s t-test. *, P⬍0.01, as compared with ethanol treatment alone; **, P⬍0.01, as compared with ethanol-treated control group; ***, P⬍0.05, as compared with ethanol-treated control group.)

The absence of a phenotype in these mice argues that no dominant-negative activity is occurring, perhaps due to instability of this aberrant protein as our data suggest. Although it remains possible that hematopoietic challenge might reveal hematopoietic defects in MxCre⫹/RXR␣fl/fl mice, our findings suggest that RXR␣ is dispensable for normal murine hematopoiesis. Although previous studies looking at the role of RXRs in hematopoietic growth and differentiation have focused primarily on RXR␣, we find that in bone marrow, RXR␤ is expressed at a level comparable with RXR␣. Furthermore, when RXR␣ is conditionally inactivated by expression of Cre recombinase, there is no compensatory up-regulation of RXR␤. Previous studies have found that RXR␥ is scarce in hematopoietic cells. Similarly, we find that RXR␥ is virtually undetectable in bone marrow. Taken together, these findings suggest that RXR␤, at baseline levels of expression, compensates for the loss of RXR␣ in MxCre⫹/RXR␣fl/fl bone marrow cells and fulfills the normal (steady-state), hematopoietic requirement for RXR. This result is somewhat surprising, as RXR␤ has not been implicated to serve an important role in hematopoiesis. Nonetheless, a recent study found that RXR␤ expression exceeds that of RXR␣ in human CD34⫹ hematopoietic progenitor cells [59]. This finding is consistent with our results in the mouse system and could further explain the absence of a significant hematopoietic deficit in MxCre⫹/RXR␣fl/fl mice. It is possible that RXR␤ is the functionally dominant RXR in early myeloid precursors and that the emphasis on RXR␣ reflects a bias

toward studying macrophages and other mature myeloid cell types, wherein RXR␣ expression significantly exceeds that of RXR␤. In colony assays, we see that colony formation by RXR␣ null bone marrow cells is suppressed by 9c-RA. This result is well-described and has been attributed to RAR/RXR heterodimer function [51, 60]. The fact that we continue to see this effect in RXR␣ null bone marrow further supports the idea that RXR␤ satisfies the requirement of RAR for a heterodimeric partner in hematopoietic cells. It is interesting that we also observed that in the absence of ligand, RXR␣ null bone marrow cells formed colonies more efficiently than bone marrow cells obtained from control mice. Although the biological significance of this finding is unclear, it could suggest that, although RXR␣ is not required for murine hematopoiesis, there may be hematopoietic signaling pathways that respond selectively to RXR␣ or settings in which baseline expression of RXR␤ is limiting when there is competition for RXR. These questions regarding the functional redundancy of RXR␣ and RXR␤ in hematopoietic cells could be better-addressed using mice in which both RXR␣ and RXR␤ are simultaneously targeted within the hematopoietic system. This remains an area for future study.

ACKNOWLEDGMENTS This work was supported by an American Heart Association Grant-in-Aid (M. R.) and National Institutes of Health Grant (NIH) CA52599 (C. K. G.). C. S. S. is supported by NIH Grant HL67034. M. R. is supported by the Ramo´n y Cajal Program (MCYT) and a Marie Curie Reintegration grant. DNA sequencing was performed by the DNA Sequencing Shared Resource, University of California San Diego (UCSD) Cancer Center, which is funded in part by National Cancer Institute Cancer Center Support Grant #2 P30 A23100-18. Conditionally targeted RXR␣ mice were provided by J. C. and K. R. C. MxCre transgenic mice were obtained with permission from Ralf Ku¨hn and Klause Rajewsky (University of Cologne, Germany). Ron Evans (Salk Institute, San Diego, CA) provided plasmid templates pCMX-RXR␣, -RXR␤, and -RXR␥. We thank Dennis J. Young in the UCSD Cancer Center’s Flow Cytometry Shared Resource for flow cytometer expertise. M. R. and C. S. S. contributed to the design, performance, and analysis of this research. M. R. generated, maintained, and supervised the breeding of the mice used in this study. C. S. S. wrote the manuscript. M. R. and C. S. S. shared in the editing of the manuscript. H-Y. L. assisted with the performance of the experiments. C. K. G. provided support for all aspects of this project.

REFERENCES 1. Tenen, D. G., Hromas, R., Licht, J. D., Zhang, D. E. (1997) Transcription factors, normal myeloid development, and leukemia. Blood 90, 489 –519. 2. Mangelsdorf, D. J., Thummel, C., Beato, M., Herrlich, P., Schutz, G., Umesono, K., Blumberg, B., Kastner, P., Mark, M., Chambon, P., Evans, R. M. (1995) The nuclear receptor superfamily: the second decade. Cell 83, 835– 839.

Ricote et al. Conditional gene targeting of RXR␣

859

3. Melnick, A., Licht, J. D. (1999) Deconstructing a disease: RARalpha, its fusion partners, and their roles in the pathogenesis of acute promyelocytic leukemia. Blood 93, 3167–3215. 4. Fenaux, P., Chomienne, C., Degos, L. (2001) All-trans retinoic acid and chemotherapy in the treatment of acute promyelocytic leukemia. Semin. Hematol. 38, 13–25. 5. Lawson, N. D., Berliner, N. (1999) Neutrophil maturation and the role of retinoic acid. Exp. Hematol. 27, 1355–1367. 6. Chambon, P. (1996) A decade of molecular biology of retinoic acid receptors. FASEB J. 10, 940 –954. 7. Kastner, P., Mark, M., Chambon, P. (1995) Nonsteroid nuclear receptors: what are genetic studies telling us about their role in real life? Cell 83, 859 – 869. 8. Giguere, V. (1994) Retinoic acid receptors and cellular retinoid binding proteins: complex interplay in retinoid signaling. Endocr. Rev. 15, 61–79. 9. Mangelsdorf, D. J., Borgmeyer, U., Heyman, R. A., Zhou, J. Y., Ong, E. S., Oro, A. E., Kakizuka, A., Evans, R. M. (1992) Characterization of three RXR genes that mediate the action of 9-cis retinoic acid. Genes Dev. 6, 329 –344. 10. Rosenfeld, M. G., Glass, C. K. (2001) Coregulator codes of transcriptional regulation by nuclear receptors. J. Biol. Chem. 276, 36865–36868. 11. Shibata, H., Spencer, T. E., Onate, S. A., Jenster, G., Tsai, S. Y., Tsai, M. J., O’Malley, B. W. (1997) Role of co-activators and co-repressors in the mechanism of steroid/thyroid receptor action. Recent Prog. Horm. Res. 52, 141–164 (discussion 164-5). 12. Levin, A. A., Sturzenbecker, L. J., Kazmer, S., Bosakowski, T., Huselton, C., Allenby, G., Speck, J., Kratzeisen, C., Rosenberger, M., Lovey, A., et al. (1992) 9-cis retinoic acid stereoisomer binds and activates the nuclear receptor RXR alpha. Nature 355, 359 –361. 13. Heyman, R. A., Mangelsdorf, D. J., Dyck, J. A., Stein, R. B., Eichele, G., Evans, R. M., Thaller, C. (1992) 9-cis retinoic acid is a high affinity ligand for the retinoid X receptor. Cell 68, 397– 406. 14. Brown, T. R., Stonehouse, T. J., Branch, J. S., Brickell, P. M., Katz, D. R. (1997) Stable transfection of U937 cells with sense or antisense RXRalpha cDNA suggests a role for RXR-alpha in the control of monoblastic differentiation induced by retinoic acid and vitamin D. Exp. Cell Res. 236, 94 –102. 15. Johnson, B. S., Chandraratna, R. A., Heyman, R. A., Allegretto, E. A., Mueller, L., Collins, S. J. (1999) Retinoid X receptor (RXR) agonistinduced activation of dominant-negative RXR-retinoic acid receptor alpha403 heterodimers is developmentally regulated during myeloid differentiation. Mol. Cell. Biol. 19, 3372–3382. 16. Sakashita, A., Kizaki, M., Pakkala, S., Schiller, G., Tsuruoka, N., Tomosaki, R., Cameron, J. F., Dawson, M. I., Koeffler, H. P. (1993) 9-cisretinoic acid: effects on normal and leukemic hematopoiesis in vitro. Blood 81, 1009 –1016. 17. Shiohara, M., Dawson, M. I., Hobbs, P. D., Sawai, N., Higuchi, T., Koike, K., Komiyama, A., Koeffler, H. P. (1999) Effects of novel RAR- and RXR-selective retinoids on myeloid leukemic proliferation and differentiation in vitro. Blood 93, 2057–2066. 18. Tocci, A., Parolini, I., Gabbianelli, M., Testa, U., Luchetti, L., Samoggia, P., Masella, B., Russo, G., Valtieri, M., Peschle, C. (1996) Dual action of retinoic acid on human embryonic/fetal hematopoiesis: blockade of primitive progenitor proliferation and shift from multipotent/erythroid/monocytic to granulocytic differentiation program. Blood 88, 2878 –2888. 19. Walkley, C. R., Yuan, Y. D., Chandraratna, R. A., McArthur, G. A. (2002) Retinoic acid receptor antagonism in vivo expands the numbers of precursor cells during granulopoiesis. Leukemia 16, 1763–1772. 20. Minucci, S., Zand, D. J., Dey, A., Marks, M. S., Nagata, T., Grippo, J. F., Ozato, K. (1994) Dominant negative retinoid X receptor beta inhibits retinoic acid-responsive gene regulation in embryonal carcinoma cells. Mol. Cell. Biol. 14, 360 –372. 21. Mendelsohn, C., Mark, M., Dolle, P., Dierich, A., Gaub, M. P., Krust, A., Lampron, C., Chambon, P. (1994) Retinoic acid receptor beta 2 (RAR beta 2) null mutant mice appear normal. Dev. Biol. 166, 246 –258. 22. Lufkin, T., Lohnes, D., Mark, M., Dierich, A., Gorry, P., Gaub, M. P., LeMeur, M., Chambon, P. (1993) High postnatal lethality and testis degeneration in retinoic acid receptor alpha mutant mice. Proc. Natl. Acad. Sci. USA 90, 7225–7229. 23. Lohnes, D., Kastner, P., Dierich, A., Mark, M., LeMeur, M., Chambon, P. (1993) Function of retinoic acid receptor gamma in the mouse. Cell 73, 643– 658. 24. Li, E., Sucov, H. M., Lee, K. F., Evans, R. M., Jaenisch, R. (1993) Normal development and growth of mice carrying a targeted disruption of the alpha 1 retinoic acid receptor gene. Proc. Natl. Acad. Sci. USA 90, 1590 –1594. 25. Labrecque, J., Allan, D., Chambon, P., Iscove, N. N., Lohnes, D., Hoang, T. (1998) Impaired granulocytic differentiation in vitro in hematopoietic

860

Journal of Leukocyte Biology Volume 80, October 2006

26. 27.

28. 29.

30. 31. 32.

33.

34. 35. 36. 37.

38. 39.

40. 41.

42. 43. 44.

45.

46. 47. 48. 49.

cells lacking retinoic acid receptors alpha1 and gamma. Blood 92, 607– 615. Kastner, P., Lawrence, H. J., Waltzinger, C., Ghyselinck, N. B., Chambon, P., Chan, S. (2001) Positive and negative regulation of granulopoiesis by endogenous RARalpha. Blood 97, 1314 –1320. Zhang, X. K., Lehmann, J., Hoffmann, B., Dawson, M. I., Cameron, J., Graupner, G., Hermann, T., Tran, P., Pfahl, M. (1992) Homodimer formation of retinoid X receptor induced by 9-cis retinoic acid. Nature 358, 587–591. Zhang, X. K., Hoffmann, B., Tran, P. B., Graupner, G., Pfahl, M. (1992) Retinoid X receptor is an auxiliary protein for thyroid hormone and retinoic acid receptors. Nature 355, 441– 446. Kliewer, S. A., Umesono, K., Noonan, D. J., Heyman, R. A., Evans, R. M. (1992) Convergence of 9-cis retinoic acid and peroxisome proliferator signalling pathways through heterodimer formation of their receptors. Nature 358, 771–774. Kliewer, S. A., Umesono, K., Mangelsdorf, D. J., Evans, R. M. (1992) Retinoid X receptor interacts with nuclear receptors in retinoic acid, thyroid hormone and vitamin D3 signalling. Nature 355, 446 – 449. Mangelsdorf, D. J., Evans, R. M. (1995) The RXR heterodimers and orphan receptors. Cell 83, 841– 850. Sunaga, S., Maki, K., Lagasse, E., Blanco, J. C., Ozato, K., Miyazaki, J., Ikuta, K. (1997) Myeloid differentiation is impaired in transgenic mice with targeted expression of a dominant negative form of retinoid X receptor beta. Br. J. Haematol. 96, 19 –30. Defacque, H., Commes, T., Legouffe, E., Sevilla, C., Rossi, J. F., RochetteEgly, C., Marti, J. (1996) Expression of Retinoid X Receptor alpha is increased upon monocytic cell differentiation. Biochem. Biophys. Res. Commun. 220, 315–322. Fritsche, J., Stonehouse, T. J., Katz, D. R., Andreesen, R., Kreutz, M. (2000) Expression of retinoid receptors during human monocyte differentiation in vitro. Biochem. Biophys. Res. Commun. 270, 17–22. Ohata, M., Yamauchi, M., Takeda, K., Toda, G., Kamimura, S., Motomura, K., Xiong, S., Tsukamoto, H. (2000) RAR and RXR expression by Kupffer cells. Exp. Mol. Pathol. 68, 13–20. Szanto, A., Narkar, V., Shen, Q., Uray, I. P., Davies, P. J., Nagy, L. (2004) Retinoid X receptors: X-ploring their (patho)physiological functions. Cell Death Differ. 11, Suppl 2, S126 –S143. Krezel, W., Dupe, V., Mark, M., Dierich, A., Kastner, P., Chambon, P. (1996) RXR gamma null mice are apparently normal and compound RXR alpha ⫹/⫺/RXR beta ⫺/⫺/RXR gamma ⫺/⫺ mutant mice are viable. Proc. Natl. Acad. Sci. USA 93, 9010 –9014. Kastner, P., Mark, M., Leid, M., Gansmuller, A., Chin, W., Grondona, J. M., Decimo, D., Krezel, W., Dierich, A., Chambon, P. (1996) Abnormal spermatogenesis in RXR beta mutant mice. Genes Dev. 10, 80 –92. Kastner, P., Grondona, J. M., Mark, M., Gansmuller, A., LeMeur, M., Decimo, D., Vonesch, J. L., Dolle, P., Chambon, P. (1994) Genetic analysis of RXR alpha developmental function: convergence of RXR and RAR signaling pathways in heart and eye morphogenesis. Cell 78, 987– 1003. Sucov, H. M., Dyson, E., Gumeringer, C. L., Price, J., Chien, K. R., Evans, R. M. (1994) RXR alpha mutant mice establish a genetic basis for vitamin A signaling in heart morphogenesis. Genes Dev. 8, 1007–1018. Chen, J., Kubalak, S. W., Chien, K. R. (1998) Ventricular musclerestricted targeting of the RXRalpha gene reveals a non-cell-autonomous requirement in cardiac chamber morphogenesis. Development 125, 1943– 1949. Kuhn, R., Schwenk, F., Aguet, M., Rajewsky, K. (1995) Inducible gene targeting in mice. Science 269, 1427–1429. Ausubel, F. M. (1994) Current protocols in molecular biology. John Wiley & Sons, New York. Lee, C. K., Raz, R., Gimeno, R., Gertner, R., Wistinghausen, B., Takeshita, K., DePinho, R. A., Levy, D. E. (2002) STAT3 is a negative regulator of granulopoiesis but is not required for G-CSF-dependent differentiation. Immunity 17, 63–72. Higuchi, M., O’Brien, D., Kumaravelu, P., Lenny, N., Yeoh, E. J., Downing, J. R. (2002) Expression of a conditional AML1-ETO oncogene bypasses embryonic lethality and establishes a murine model of human t(8;21) acute myeloid leukemia. Cancer Cell 1, 63–74. Morrison, S. J., Uchida, N., Weissman, I. L. (1995) The biology of hematopoietic stem cells. Annu. Rev. Cell Dev. Biol. 11, 35–71. Morrison, S. J., Weissman, I. L. (1994) The long-term repopulating subset of hematopoietic stem cells is deterministic and isolatable by phenotype. Immunity 1, 661– 673. Spangrude, G. J., Heimfeld, S., Weissman, I. L. (1988) Purification and characterization of mouse hematopoietic stem cells. Science 241, 58 – 62. Spangrude, G. J., Scollay, R. (1990) A simplified method for enrichment of mouse hematopoietic stem cells. Exp. Hematol. 18, 920 –926.

http://www.jleukbio.org

50. Spangrude, G. J., Smith, L., Uchida, N., Ikuta, K., Heimfeld, S., Friedman, J., Weissman, I. L. (1991) Mouse hematopoietic stem cells. Blood 78, 1395–1402. 51. Jacobsen, S. E., Fahlman, C., Blomhoff, H. K., Okkenhaug, C., Rusten, L. S., Smeland, E. B. (1994) All-trans- and 9-cis-retinoic acid: potent direct inhibitors of primitive murine hematopoietic progenitors in vitro. J. Exp. Med. 179, 1665–1670. 52. Rusten, L. S., Dybedal, I., Blomhoff, H. K., Blomhoff, R., Smeland, E. B., Jacobsen, S. E. (1996) The RAR-RXR as well as the RXR-RXR pathway is involved in signaling growth inhibition of human CD34⫹ erythroid progenitor cells. Blood 87, 1728 –1736. 53. Purton, L. E., Bernstein, I. D., Collins, S. J. (1999) All-trans retinoic acid delays the differentiation of primitive hematopoietic precursors (lin-ckit⫹Sca-1(⫹)) while enhancing the terminal maturation of committed granulocyte/monocyte progenitors. Blood 94, 483– 495. 54. Mikkola, H. K., Klintman, J., Yang, H., Hock, H., Schlaeger, T. M., Fujiwara, Y., Orkin, S. H. (2003) Haematopoietic stem cells retain longterm repopulating activity and multipotency in the absence of stem-cell leukaemia SCL/tal-1 gene. Nature 421, 547–551.

55. Horcher, M., Souabni, A., Busslinger, M. (2001) Pax5/BSAP maintains the identity of B cells in late B lymphopoiesis. Immunity 14, 779 –790. 56. He, J., Navarette, S., Mason, P. J., Bessler, M. (2002) Deletions in Dkc1, the gene mutated in X-linked dyskeratosis congenita, cause embryonic lethality in mice, but permit cell survival in adult tissues. Blood 100, 659a. 57. Okada, S., Nakauchi, H., Nagayoshi, K., Nishikawa, S., Miura, Y., Suda, T. (1992) In vivo and in vitro stem cell function of c-kit- and Sca-1positive murine hematopoietic cells. Blood 80, 3044 –3050. 58. Osawa, M., Nakamura, K., Nishi, N., Takahasi, N., Tokuomoto, Y., Inoue, H., Nakauchi, H. (1996) In vivo self-renewal of c-Kit⫹ Sca-1⫹ Lin(low/-) hemopoietic stem cells. J. Immunol. 156, 3207–3214. 59. Szanto, A., Nagy, L. (2005) Retinoids potentiate peroxisome proliferatoractivated receptor gamma action in differentiation, gene expression, and lipid metabolic processes in developing myeloid cells. Mol. Pharmacol. 67, 1935–1943. 60. Dawson, M. I., Elstner, E., Kizaki, M., Chen, D. L., Pakkala, S., Kerner, B., Koeffler, H. P. (1994) Myeloid differentiation mediated through retinoic acid receptor/retinoic X receptor (RXR) not RXR/RXR pathway. Blood 84, 446 – 452.

Ricote et al. Conditional gene targeting of RXR␣

861