FULL PAPER Internal Medicine

Quantitative Analysis of mRNA Transcripts of Hox, SHH, PTCH, Wnt, and Fzd in Canine Hematopoietic Progenitor Cells and Various in vitro Colonies Differentiated from the Cells

Kaori IDE1), Yuko GOTO-KOSHINO1), Yasuyuki MOMOI2), Yasuhito FUJINO1), Koichi OHNO1) and Hajime TSUJIMOTO1)*

1)Department of Veterinary Internal Medicine, Graduate School of Agricultural and Life Science, The University of Tokyo, 1–1–1 Yayoi, Bunkyo-ku, Tokyo 113–8657 and 2)Department of Veterinary Diagnostic Imaging, Faculty of Agriculture, Kagoshima University, 1–21–24 Korimoto, Kagoshima 890–0065, Japan

(Received 10 June 2008/Accepted 23 September 2008)

ABSTRACT. (Hox), Sonic hedgehog (SHH), and Wingless-type MMTV integration site family (Wnt) are known to modulate the self-renewal and expansion of hematopoietic progenitor/stem cells in humans and mice. Frizzled (Fzd) and Patched1 (PTCH1) rep- resent the receptors of Wnt and SHH, respectively. In this study, the amounts of mRNA transcripts of the genes associated with the self-renewal of hematopoietic stem cells, HoxB3, HoxB4, HoxA10, Wnt5a, Wnt2b, Fzd1, Fzd6, SHH, and PTCH1, were measured in canine unfractionated bone marrow cells, CD34-enriched cells, and various colony-forming units in culture (CFU-C). Partial cDNA sequences of these 9 canine genes were determined in this study. Quantitative real-time polymerase chain reaction was employed to indi- cate their relative amounts of mRNA transcripts. Amounts of mRNA transcripts of HoxB3, HoxA10, PTCH1, and Wnt5a genes in canine CD34-enriched cell fraction were significantly larger than those in the CD34-depleted cell fraction. Amounts of mRNA transcripts of HoxB3, HoxA10, PTCH1, Wnt5a, and Wnt2b genes in various CFU-C cells were significantly smaller than those in the seeded CD34- enriched cell fraction. These results suggested important roles of the products of these genes in self-renewal, expansion, and survival of hematopoietic progenitor cells in dogs as shown in humans and rodents. KEY WORDS: canine, CD34, Hox, SHH, Wnt. J. Vet. Med. Sci. 71(1): 69–77, 2009

Hematopoietic stem cells (HSCs) possess a capacity of methods that use these cytokines require a large amount of self-renewal and the ability to give rise to all types of blood expensive recombinant . More importantly, a num- cells. This tremendous potential for reconstituting the ber of recombinant cytokines that are widely available are hematopoietic system has allowed the development of the limited to humans and mice. The activities of cytokines, transplantation of the HSCs as a clinical strategy in humans particularly that of IL-3, are known to be species-specific, [14, 16, 39, 41, 44]. Conventional transplantations leave and their full effects can only be expected with recombi- engraftment to natural interactions between the graft and the nants of the same species [8, 17, 35]. Therefore, currently, host, requiring considerably large amounts of graft tissue. it is very difficult to obtain large amounts of hematopoietic Technologies that efficiently expand the HSCs ex vivo cytokines in order to expand HSCs in dogs. would free the limitations resulting from graft shortages. Some of the recent studies on HSC focused on the endog- Obtaining a sufficient amount of bone marrow graft con- enous transcription factors and exogenous signaling mole- sumes both time and effort. Since the ex vivo expansion of cules. Representatives of the former group are Homeobox HSCs remains difficult, these highly potential therapeutic (Hox) and BMI1 polycomb ring finger oncogene (BMI-1), options have not been applied in clinical practice so far. whereas those of the latter group are Wingless-type MMTV Although the clinical applications of bone marrow trans- integration site family (Wnt), Sonic hedgehog (SHH) pro- plantation combined with intensified chemotherapy have tein, bone morphogenic (BMP), and Notch ligands recently been initiated in veterinary medicine [11], identical [46]. In particular, Hox, Wnt, and SHH are known to play problems exist in this field as well. important roles in the self-renewal of the HSCs [3, 19, 21, In humans and mice, many efforts to expand HSCs have 40, 46, 53]. been made using various cytokine cocktails, resulting in Hox, SHH, and Wnt have been reported to be important some remarkable results [51]. Key cytokines used today are regulators during development [22, 27, 46, 47, 53, 55]. stem cell factor (SCF), FLT-ligand, interleukin (IL)-3, IL-6, There are 39 Hox genes known today, and they are classified and thrombopoietin (TPO) [5, 6, 37, 38, 46, 51]. However, into 4 groups, from A to D [21, 46]. In particular, HoxB3, B4, and A10 are known to be expressed in human CD34+ *CORRESPONDENCE TO: TSUJIMOTO, H., Department of Veterinary Internal Medicine, Graduate School of Agricultural and Life Sci- cells [20, 42, 49]. Hematopoietic progenitors retrovirally ences, the University of Tokyo, 1–1–1 Yayoi, Bunkyo, Tokyo transduced with HoxB4 have demonstrated up to a 1000- 113–8657, Japan. fold expansion of HSCs relative to control cells transduced e-mail: [email protected] with control vectors [1]. Moreover, the transplantation of 70 K. IDE ET AL. these HoxB4-transduced HSCs was shown to be repeatable the BMMCs by a magnetic-activated cell separation system without leukemia development [1, 43]. HoxB3, its (MACS) (Miltenyi Biotec GmbH, Gladbach, Germany) located upstream of HoxB4 gene, is suggested to play a sim- according to the manufacturer’s instructions with some ilar role as HoxB4 [4, 21, 42, 49]. Further, HoxA10 has modifications. Mouse anti-canine CD34 monoclonal anti- been suggested to have a major influence on the lineage body, clone 1H6 (BD Biosciences, San Diego, CA.), and rat commitments of the hematopoietic cells [25, 48]. anti-mouse IgG microbeads (Miltenyi Biotec) were used for SHH is a member of the “hedgehog” family of proteins. the selection of the CD34-enriched cell fraction. Beside the Of the 5 members of the family, SHH has been most inten- MACS separation of CD34-enriched cell fraction, a CD34- sively studied in relation to the self-renewal of HSCs [3, 46]. depleted cell fraction was also collected. For flow cytome- Patched (PTCH) and Smoothed (SMOH) are known as try (FACSCaliburTM; BD Biosciences), the cells were receptors of SHH, and PTCH inhibits SMOH activation in a treated with biotin-conjugated mouse anti-canine CD34 normal state [31, 32, 55]. When SHH binds to PTCH, a monoclonal antibody and streptavidin-PerCP-Cy 5.5 conju- repressor against SMOH is released, leading to the activa- gate to detect canine CD34+ cells. Biotin-conjugated mouse tion of the downstream Gli family transcription factors, IgG1κ (BD Biosciences) was used as an isotype control. which results in the activation of target genes [3, 9]. SHH Data analysis was carried out with CellQuestTM (BD Bio- has been shown to induce HSC proliferation via BMP sig- sciences). The freshly isolated BMMCs, CD34-enriched nals in humans [3]. cell fraction, and CD34-depleted cell fraction were washed, Wnt is a group of secreted glycoproteins, and 19 families flash-frozen in liquid nitrogen, and stored at –80°C until of Wnt genes are known in humans and mice so far [34, 45, use. 47]. Similar to the hedgehog family, Wnt acts via autocrine Stromal cell primary culture: Stromal cells were cultured and paracrine pathways. Frizzled (Fzd) proteins act as Wnt according to a method previously reported [33]. Briefly, receptors. Of the 19 molecules of the family, Wnt2b, 5A, BMMCs were cultured in Iscove’s Modified Dulbecco’s and 10B are known to promote the proliferation of murine Medium (IMDM) (Sigma-Aldrich, St. Louis, MO) supple- HSCs [2, 30, 52]. mented with 20% horse serum (Gibco, Invitrogen, Carlsbad, Much evidence supports the possibility that these mole- CA) and 10–7 M of hydrocortisone (Solu-Cortef®; Pfizer cules–Hox, SHH, and Wnt–are associated with one another Japan Inc., Tokyo, Japan). After cultivation for 2–3 weeks, and work together to maintain normal hematopoiesis [10, a mixture of adherent cells with a fibroblastic and macroph- 40]. Unlike cytokines, Hox, SHH, and Wnt are highly con- age-like appearance was obtained. After removing the cul- served among species from flies to humans [15, 24, 46, 47, ture medium containing nonadherent cells, the adherent 54, 55]. However, studies regarding the self-renewal and stromal cells were harvested by digestion with 0.5% trypsin- expansion of the HSC are extremely limited in dogs. EDTA (Gibco). The final goal of this study is to utilize these novel factors CFU-C assay: CFU-Cs were obtained by culturing the for clinical application, particularly for HSC transplantation CD34-enriched cell fraction in methylcellulose CFU assay. in dogs. Knowing the profiles of the transcripts from these Prior to the experiment, phytohemagglutinin (PHA-P)-acti- molecules in HSCs in dogs was considered to be the first vated lymphocyte conditioned medium (PHA-LCM; PHA- essential step. In this study, the amounts of mRNA tran- P, Sigma-Aldrich) was prepared. Canine peripheral blood scripts derived from selected Hox, SHH, and Wnt genes mononuclear cells (2 × 106 cells/ml) were cultured in IMDM were measured in bone marrow-derived mononuclear cells supplemented with 15% fetal bovine serum (FBS) and 10 µg (BMMCs), CD34-enriched and CD34-depleted fractions, of PHA-P per ml for 1 week. The culture supernatant was bone marrow-derived stromal cells, and various CFU-Cs in collected, clarified by centrifugation, and then stored at dogs. –80°C until use. The CD34-enriched cell fraction (1.14 × 104 cells) suspended in 250 µl PHA-LCM was added to 2.5 MATERIALS AND METHODS ml of methylcellulose medium with human recombinant cytokine supplements (SCF, 50 ng/ml; GM-CSF, 20 ng/ml; Animals and cells: Bone marrow cells were aspirated IL-3, 20 ng/ml; IL-6, 20 ng/ml; G-CSF, 20 ng/ml; and EPO, from 4 healthy Beagle dogs under general anesthesia. These 3 U/ml) (Methocult GF+ 4435; Stem Cell Technologies dogs were 1~5 years old; 1 dog was male and 3 dogs were Inc., Vancouver, Canada). Then, 2.2 ml of the cell suspen- female. Sodium citrate was used as an anticoagulant for sion was dropped into 35-mm plates (Sumitomo Bakelite harvesting the bone marrow cells. This study was con- Co., Tokyo, Japan), and the plates were incubated at 37°C in ducted in accordance with the guidelines of the Animal Care a humidified atmosphere of 5% CO2 for 2 weeks. Colonies Committee of the Graduate School of Agricultural and Life (number of cells > 50) were morphologically differentiated Sciences, The University of Tokyo. and counted under an inverted microscope. Each colony Isolation of BMMCs and CD34-enriched cell fraction: type was confirmed by Wright-Giemsa stain. Under an BMMCs were isolated by performing density gradient cen- inverted microscope, the colonies were carefully picked up trifugation of the aspirated bone marrow cells with Ficoll- into RNase-free phosphate-buffered saline (PBS) (Sigma- Hypaque Plus (GE Healthcare UK Ltd., Buckinghamshire, Aldrich). To obtain a sufficient number of cells for an ade- UK). The CD34-enriched cell fraction was isolated from quate amount of RNA, approximately 5 colonies of the same GENE TRANSCRIPTS IN BONE MARROW PROGENITORS 71 type were collected into a single tube. Harvested stromal System; TaKaRa) were as follows: incubation at 95°C for cells and CFU-Cs were washed, flash-frozen in liquid nitro- 10 sec; followed by 45 cycles of 95°C for 5 sec and 60°C for gen, and kept at –80°C until use. 30 sec during which the fluorescence data were collected; RNA extraction and cDNA synthesis: Total RNA samples followed by the last step to determine dissociation at 95°C were extracted from the BMMCs, CD34-enriched cell frac- for 15 sec, 60°C for 30 sec, and 95°C for 15 sec. Dissocia- tion, CD34-depleted cell fraction, bone marrow stromal tion curves and data from polyacrylamide gel electrophore- cells, and CFU-Cs by using the Illustra RNAspin Mini RNA sis (data not shown) of the products were used to confirm isolation kit (GE Healthcare UK Ltd., Buckinghamshire, that there was a single product for each gene. The threshold UK) in combination with DNase treatment. cDNAs were cycle (Ct) was determined by the second derivative method, synthesized with the SuperscriptTM III First-Strand Synthe- and the amount of cDNA derived from the mRNA transcript sis System (Invitrogen) according to the manufacturers’ of each gene was calculated as a relative value of the amount instructions and then kept at –20°C until use. Elimination of of HPRT cDNA, with a software attached to the thermal genomic DNA was confirmed by analyzing samples without cycler (TP800) (TaKaRa). reverse transcriptase treatment. Statistical analysis: All statistical analyses were per- Cloning of the target genes: Canine homologues of 9 formed using JMP computer software (SAS Institute Inc., genes–SHH, PTCH1, Wnt5a, Wnt2b, Fzd1, Fzd6, HoxB3, Cary, NC). Wilcoxon rank sum test was used for compari- HoxB4, and HoxA10–were analyzed. The sequence of son of data between CD34-enriched and CD34-depleted cell cDNA of the canine hypoxanthine phosphoribosyl trans- fractions. Dunnett’s test was employed for comparison of ferase (HPRT) gene–which was already registered (Gen- data between CD34-enriched cell fraction and various CFU- bank accession number, AY283372)–was used as internal Cs. control. All of the genes except HPRT were partially cloned in this study to determine the target sequences for quantita- RESULTS tive real-time polymerase chain reaction (RT-PCR). Vari- ous canine tissues were used as templates; they were as Sequences of partially cloned genes were all confirmed follows: uterus (HoxB4, SHH), spleen (HoxB3), BMMCs by BLAST search, and they were registered in the Genbank (PTCH1, Wnt5A, Fzd1, and Fzd6), bone marrow stromal with the accession numbers that are listed in Table 1. cells (Wnt2B), and small intestine (HoxA10). The poly- The percentage of CD34+ cells in the unfractionated merases used for the cloning were ExTaq® Hot Start Ver- BMMCs was 1.37 ± 0.73%. The percentages of CD34+ cells sion (TaKaRa Bio, Otsu, Japan) and LA TaqTM (TaKaRa were 82.2 ± 5.0% and 0.87 ± 0.43% in the CD34-enriched Bio). Primer pairs were designed using Primer3 software and CD34-depleted cell fractions, respectively. Amounts of (http://frodo.wi.mit.edu/). The primer pairs used for cloning mRNA transcripts of the genes examined were shown as - each gene, their melting temperatures (Tm), sizes of the ative values in comparison to that of the HPRT gene. The cloned sequences, and their Genbank accession numbers are amounts of mRNA transcripts of HoxB4, HoxB3, HoxA10, listed in Table 1. Conditions for the thermal cycler (Takara SHH, PTCH1, Wnt5a, Wnt2b, Fzd1, and Fzd6 genes in the PCR Thermal Cycler Dice® Gradient; TaKaRa Bio) were as CD34-enriched cell fraction were compared to those in the follows: denaturing at 95°C for 30 sec, followed by amplifi- CD34-depleted cell fraction (Fig. 1). Of the 9 genes exam- cation with 45 cycles of 98°C for 10 sec, each Tm for 30 sec ined, the amounts of mRNA of HoxB3, HoxA10, PTCH1, and 72°C for 1 min, then final extension at 72°C for 7 min. and Wnt5a genes in the CD34-enriched cell fraction were PCR products were electrophoresed in 2% agarose, and the significantly larger than those in the CD34-depleted cell single bands were directly ligated into pGEM®-T Easy plas- fraction. Notably, the amount of mRNA transcript of mid vector (Promega Corporation, Madison, WI), while HoxA10 gene in the CD34-depleted cell population was less nonspecific bands were extracted from the gel using than 10–3-fold quantity of the transcript in the CD34- QIAEXII (QIAGEN, Valencia, CA) prior to ligation. enriched cell population. The amount of HoxA10 mRNA Sequencing analyses of these PCR products were carried out transcript in unfractionated BMMCs was significantly with the BigDye® Terminator v3.1 Cycle Sequencing Kit smaller than that of the CD34-enriched cell fraction but was (Applied Biosystems, Foster City, CA) and Genetic Ana- within the detectable range. Unlike the HoxA10 gene, the lyzer 3130/3130×l (Applied Biosystems). Cloned amounts of mRNA transcripts of the other 8 genes were sim- sequences were searched for the dog genome [23] using ilar between the unfractionated BMMCs and the CD34- NCBI BLAST software. depleted cell fraction. Quantitative RT-PCR: Primer pairs for RT-PCR were Figure 2 shows the amounts of mRNA transcripts of designed based on the partially cloned sequences of the 9 HoxB4, HoxB3, HoxA10, SHH, PTCH1, Wnt5a, Wnt2b, genes. All the primers except those of the HPRT gene [36] Fzd1, and Fzd6 genes in the CD34-enriched cell fraction were designed in this study (Table 2). For RT-PCR, the and CFU-Cs derived from the culture of the CD34-enriched SYBR green system (SYBR® Premix Ex TaqTM; TaKaRa) cells. The amounts of mRNA transcripts of HoxB4 gene in was employed to detect double-stranded nucleic acids CFU-granulocyte/erythroid/monocyte/megakaryocyte formed during the PCR cycles. Conditions of the thermal (CFU-GEMM), CFU-granulocyte/monocyte (CFU-GM), cycler for real-time PCR (Thermal Cycler Dice® Real Time CFU-granulocyte (CFU-G), and burst-forming-unit- eryth- 72 K. IDE ET AL.

Table 1. Primer pairs used for cloning and Genbank accession numbers and the sizes of the cloned sequences Target genes Primer sequences (5’→3’) Tm Sequence Genbank (°C) sizes (bp) accession numbers Canine HoxB3 Fw CTTCATGAACGCCTTGCACTC 60 474 EU332993 Rev CACAGGTGTGTTAATTTGGGC Canine HoxB4 Fw CCAGAAATTAATGGCTATGAG 60 109 EU332992 Rev TGGTCGCTGGGTAGGTAATC Canine HoxA10 Fw AGAGCAGCAAAGCCTCGCCC 59 298 EU332994 Rev AAATTGGCTGTGAGCTCCCGGATCC Canine SHH Fw GCTTCGACTGGGTGTACTACG 60 213 EU332997 Rev GTGAGGAAGTCGCTGTAGAGC Canine PTCH1 Fw GCGGATATCCTGGTTTGAGA 55 308 EU332998 Rev AGGACGGGTTGAAGACAATG Canine Wnt5a Fw CAAGGTGGGCGACGCCCTGAAGGAG 57 153 EU332995 Rev CGTCTGCACGGTCTTGAACTGGTCGTA Canine Wnt2b Fw TAAGTGCCATGGCGTAAGTG 57 311 EU332996 Rev GCCACAGCACATGATTTCAC Canine Fzd1 Fw AGCCGAGAAAGTATGGCTGA 58 678 EU332999 Rev GGGAACTTCTCGCACTTGAG Canine Fzd6 Fw CACTGTGCCCCGATGTATGA 60 413 EU333000 Rev ATGCCTTGGACACCAAAATC

Table 2. Primer pairs used for the real-time PCR showed no statistical differences between any of the cell populations. Target genes Primer sequences (5’→3’) In stromal cells derived from BMMCs, the amounts of Canine HoxB3 Fw TCAGGGTAGAATCCAAGAAG mRNA transcripts of the HoxB3, HoxB4, HoxA10, PTCH1, Rev CACAGGTGTGTTAATTTGGG Wnt5a and Wnt2b genes were significantly smaller than Canine HoxB4 Fw CCAGAAATTAATGGCTATGAG those of the CD34-enriched cell fractions. Compared to the Rev TGGTCGCTGGGTAGGTAATC fresh BMMCs, the amounts of mRNA transcript of the Canine HoxA10 Fw AGAGCAGCAAAGCTCGCCC Rev GCGTCTGGTGCTTGGTGTAA examined genes except Fzd6 in the stromal cells were Canine SHH Fw AAGGCGCACATCCACTGTT smaller, yet without statistical differences. Rev CCACGGAATTCTCTGCTTTC Canine PTCH1 Fw GCGGATATCCTGGTTGAGA DISCUSSION Rev CAAGCCAGTGGACACATTGA Canine Wnt5a Fw CTCAAGGACAGAAGAAACTG In this study, the amounts of mRNA transcripts of Rev TTGCGCCTTCTCCAATGTAC Canine Wnt2b Fw CCTTGTCTACTTTGACAACT HoxB3, HoxB4, HoxA10, SHH, PTCH1, Wnt5a, Wnt2b, Rev GAACCTGCAGCCTTGTCCAA Fzd1, and Fzd6 genes were quantitatively analyzed in Canine Fzd1 Fw ATGAACAAGTTCGGCTTCCA canine bone marrow cells: BMMCs, CD34-enriched cell Rev GGGAACTTCTCGCACTTGAG fraction, CD34-depleted cell fraction, stromal cells, and Canine Fzd6 Fw GATGGCCTGAGGAACTTGAA CFU-Cs. The three Hox genes, SHH gene, and two Wnt Rev ATGCCTTGGACACCAAAATC genes were selected based on the previous studies in humans and rodents in which their important roles in self-renewal, maintenance, and survival of HSCs had been reported [3, roid (BFU-E) were significantly smaller than that in the 19, 21, 40, 46, 53]. PTCH1 and Fzd genes were analyzed seeded CD34-enriched cell fraction. Similarly, the amounts because they code the receptors of SHH and Wnt, respec- of mRNA transcripts of HoxB3 and HoxA10 genes in all of tively [32, 47]. the CFU-Cs examined were significantly smaller than those The amounts of mRNA transcripts of three Hox genes in the seeded CD34-enriched cell fractions. The amounts of were shown to be larger in canine CD34-enriched cell frac- mRNA transcripts of SHH gene in CFU-GEMM and CFU- tion than the differentiated CFU-Cs in this study . The GM were smaller than that in the CD34-enriched cell frac- HoxB4 gene was initially the top candidate strongly tion. The amounts of mRNA transcripts of PTCH1 gene in expressed in the CD34-enriched cell fraction; however, no all of the CFU-Cs examined were significantly smaller than significant difference in the amount of its mRNA transcript that in the CD34-enriched cell fraction. Similar to the was observed between the CD34-enriched and CD34- results observed in the HoxB3, HoxA10, and PTCH1 genes, depleted fractions in this study. Instead, the amounts of the amounts of mRNA transcripts of Wnt5a and Wnt2b mRNA transcripts of HoxB3 and HoxA10 genes were signif- genes in all of the CFU-Cs examined were significantly icantly larger in the CD34-enriched cell fraction than in smaller than those in the CD34-enriched cell fractions. The unfractionated BMMCs, CD34-depleted cells, stromal cells, amounts of mRNA transcripts of the Fzd1 and Fzd6 genes and CFU-Cs. Relatively large amounts of the mRNA tran- GENE TRANSCRIPTS IN BONE MARROW PROGENITORS 73

Fig. 1. Amounts of mRNA transcripts of HoxB4, HoxB3, HoxA10, SHH, PTCH1, Wnt5a, Wnt2b, Fzd1, and Fzd6 genes determined by quantitative RT-PCR relative to that of an internal reference gene, HPRT. Values are the mean of the 4 dogs with bars representing standard deviations. Plus (+) and minus (–) represent CD34-enriched fraction and CD34-depleted fraction, respectively. Wilcoxon rank sum test was used for the statistical analysis. Asterisks (*) show significant differences (P<0.05). scripts of these three Hox genes in the CD34-enriched cell tively high in fetal endodermal tissues [26]. A relatively fraction are suggestive of their important roles in the small amount of mRNA transcript of SHH gene detected in hematopoietic progenitors as described in human and canine CD34-enriched cell fraction in this study was consis- rodents. Moreover, with respect to HoxB3 and HoxB4, tent with the previous results shown in humans, which is which are genes adjacently located on the same chromo- suggestive of a similar phenomenon in both humans and some, similar results were obtained in this study. In a study dogs. Another possibility is that SHH may be produced and on human cells, expressions were shown to be secreted at a distant site and is received by PTCH1-express- similar between genes close to each other by position rather ing bone marrow cells. This was suggested because the than those between genes with paralogous relationships amount of PTCH1 mRNA transcript was significantly larger [50]. Although paralogs were not analyzed in this study, the in the CD34-enriched cell fraction than any of the other cell very similar pattern observed in canine HoxB3 and HoxB4 groups in this study. In humans, PTCH1 is expressed in nor- might reflect this consistency. In other human studies, the mal bone marrow cells, while it is not detectable in differen- expressions of various Hox genes were analyzed in hemato- tiated peripheral blood mononuclear cells [10]. poietic progenitors [13, 42]. In that study, the order of the The amount of Wnt5a mRNA transcript in the CD34- expression level was HoxA, HoxB, and HoxC; further, no enriched cell fraction was significantly larger than that in all HoxD expression was detected [13, 21, 42]. In this study, of the other cells groups; this was similar to the results with the amount of mRNA transcript of HoxA10 gene was com- respect to HoxB3, HoxA10, and PTCH1 genes. On the other parable to those of HoxB3 and HoxB4. The most distinct hand, the amount of Wnt2b transcript in the CD34-enriched difference between the CD34-enriched and CD34-depleted cell fraction was significantly larger than that in CFU- cell fractions was observed in the amount of the HoxA10 GEMM, CFU-GM, CFU-M, CFU-G, BFU-E, and stromal gene transcript in the present study. In human cells, HoxA10 cells. These results were partly consistent with previous expression is known to be strictly limited to early-stage pro- studies; both Wnt5a and Wnt2b genes were expressed in genitors and is downregulated in differentiated cells such as human bone marrow CD34+ cells, both adult and fetal, as neutrophils and monocytes [20]. Moreover, high HoxA10 well as in bone marrow stromal cells [19, 52]. When the expression was reported in human CD34+ cells together genes of Wnt receptors were analyzed, the results on the with a marked reduction in CD34– cells [20, 42]. The cur- Fzd1 and Fzd6 genes were different from those on most of rent results obtained in canine cells were consistent with the other genes examined in this study. The amounts of mRNA results in these reports. transcripts of Fzd1 and Fzd6 genes between the CD34- Although the amounts of SHH gene transcript in enriched cell fraction and various CFU-Cs did not show sig- BMMCs, CD34-enriched cell fraction, CD34-depleted cell nificant difference in this study. Especially, the amounts of fraction, CFU-M, CFU-G, and BFU-E were above the lower Fzd6 transcript in CFU-Cs were similar to that in the pri- detection limit, the amounts were generally small in this mary CD34-enriched cell fraction. Continuous expression study. In previous studies, SHH treatment induced HSC of Fzd6 through the lineage commission suggests a certain proliferation in human cells [3]; however, its expression was relationship of the Wnt signals with lineage differentiations low or nondetectable in adult tissues whereas it was - in dogs. 74 K. IDE ET AL.

Fig. 2. Amounts of mRNA transcripts of Hox (A), SHH and PTCH (B), and Wnt and Fzd (C) genes in the primary CD34-enriched cell fraction and CFU-Cs determined by quantitative RT-PCR relative to that of an internal reference gene, HPRT. Values are the mean of the 4 dogs with bars representing standard deviations. Dunnett’s test was used for the statistical analysis. Asterisks (*) show significant differences (P<0.05), and those without bars connecting particular samples represent that the cell group with asterisk has data significantly different from all of the other cell groups.

In dogs, CD34 is currently the most commonly used CD34-enriched cell fraction than those in the CD34- marker to detect or isolate early hematopoietic progenitors depleted cell fraction, significant differences were observed [28, 29]. While the amounts of mRNA transcripts of all in the HoxB3, HoxA10, PTCH1, and Wnt5a genes. Yet, the genes examined except for SHH gene were larger in the differences between these 2 populations were not as large as GENE TRANSCRIPTS IN BONE MARROW PROGENITORS 75 expected prior to this study; particularly in the amount of ducted using canine cells. There has been a report of retro- HoxB4 transcript. Previous studies in normal human bone viral induction of the human HoxB4 gene into a canine marrow cells showed that the expression levels of HoxB3, CD34+ cell resulting in immortalization of the cells [56]. HoxB4, and HoxA10 were significantly higher, at least by The report also showed different outcomes with the cells 10-fold, in CD34-enriched cell fractions compared to from different species such as humans, suggesting species CD34-depletd cell fractions [18, 20, 42]. One of the possi- compatibilities between the transgene and the cells. Other ble explanations is the limitation of MACS sorting. The molecules such as SHH and Wnt in dogs should be investi- purity of CD34-enriched cell faction was 82.2 ± 5.0%, and gated in the future. However, the fact that they are highly the percentage of CD34+ cells in the CD34-depleted cell conserved between species promises a potential use for such fraction was 0.87 ± 0.43%, nearly the same as BMMCs. technologies in the future of veterinary medicine. This indicates that the CD34-enriched fraction contained approximately 80 times more CD34+ cells than the CD34- REFERENCES depleted cell fraction; however, the differences in the amounts of mRNA transcripts did not exceed 10-fold. In 1. Antonchuk, J., Sauvageau, G. and Humphries, R. K. 2002. humans and mice, CD34 is no longer the landmark for stem HOXB4-induced expansion of adult hematopoietic stem cells cells; CD34 is known to be expressed in multiple progeni- ex vivo. Cell 109: 39–45. tors including the stem cells, and combined markers are 2. Austin, T. W., Solar, G. P., Ziegler, F. C., Liem, L. and Mat- thews, W. 1997. A role for the Wnt gene family in hematopoie- available for the isolation of more primitive HSCs particu- sis: expansion of multilineage progenitor cells. Blood 89: larly in mice and humans. In addition, HoxB4 is known to 3624–3635. be expressed later than HoxB3 within the primitive hemato- 3. Bhardwaj, G., Murdoch, B., Wu, D., Baker, D. P., Williams, K. poietic progenitors [12]. If this fact was also true for dogs, P., Chadwick, K., Ling, L. E., Karanu, F. N. and Bhatia, M. the CD34-enriched cell fraction used in this study may have 2001. Sonic hedgehog induces the proliferation of primitive been a population that had not fully expressed HoxB4. Nev- human hematopoietic cells via BMP regulation. Nat. Immunol. ertheless, CD34 is a marker of primitive hematopoietic pro- 2: 172–180. genitors, and current results suggest that recently recognized 4. Bjornsson, J. M., Larsson, N., Brun, A. C., Magnusson, M., factors such as Hox, SHH, and Wnt may play some roles Andersson, E., Lundstrom, P., Larsson, J., Repetowska, E., particularly in these cells. Ehinger, M., Humphries, R. K. and Karlsson, S. 2003. Reduced proliferative capacity of hematopoietic stem cells deficient in In the present study, the mRNA transcripts of 9 kinds of Hoxb3 and Hoxb4. Mol. Cell Biol. 23: 3872–3883. genes possibly associated with the self-renewal of hemato- 5. Bodine, D. M., Karlsson, S. and Nienhuis, A. W. 1989. Combi- poietic stem cells were analyzed for their amounts relative nation of interleukins 3 and 6 preserves stem cell function in to that of an internal reference (housekeeper) gene, HPRT, culture and enhances retrovirus-mediated gene transfer into in canine bone marrow-derived cells. Since HPRT gene has hematopoietic stem cells. Proc. Natl. Acad. Sci. U.S.A. 86: been used as a suitable internal reference gene for the quan- 8897–8901. titative assay of the mRNA transcripts of Hox family and 6. Bodine, D. M., Orlic, D., Birkett, N. C., Seidel, N. E. and other related genes [7, 25, 48] in human and murine bone Zsebo, K. M. 1992. Stem cell factor increases colony-forming marrow-derived cells, we chose HPRT gene as a normalizer unit–spleen number in vitro in synergy with interleukin-6, and in this study. Moreover, it has been shown that HPRT gene in vivo in Sl/Sld mice as a single factor. Blood 79: 913–919. 7. Brun, A. C., Bjornsson, J. M., Magnusson, M., Larsson, N., can be one of the most stable internal control genes for the Leveen, P., Ehinger, M., Nilsson, E. and Karlsson, S. 2004. quantitative RT-PCR assay for the canine bone marrow Hoxb4-deficient mice undergo normal hematopoietic develop- samples [36]. As indicated in the same recent report, use of ment but exhibit a mild proliferation defect in hematopoietic multiple internal control genes was recommended for the stem cells. Blood 103: 4126–4133. quantitative RT-PCR analysis for tissues from dogs as well 8. Burger, H., van Leen, R. W., Dorssers, L. C., Persoon, N. L., as humans and rodents [36]. In the present study, the Lemson, P. J. and Wagemaker, G. 1990. Species specificity of amounts of transcripts of various genes were estimated as human interleukin-3 demonstrated by cloning and expression relative values in comparison to the amount of transcript of of the homologous rhesus monkey (Macaca mulatta) gene. a single internal reference gene, HPRT, in canine bone mar- Blood 76: 2229–2234. row-derived cells. Although HPRT gene can be considered 9. Cohen, M. M., Jr. 2003. The hedgehog signaling network. Am. J. Med. Genet. A 123: 5–28. as a suitable internal reference gene for the quantitative RT- 10. Detmer, K., Walker, A. N., Jenkins, T. M., Steele, T. A. and PCR in bone marrow cells, further strict consideration to use Dannawi, H. 2000. Erythroid differentiation in vitro is blocked multiple internal references genes would be needed. by cyclopamine, an inhibitor of hedgehog signaling. Blood In the future, we must endeavor to utilize these promising Cells Mol. Dis. 26: 360–372. factors in expanding HSCs for transplantation and/or tissue 11. Frimberger, A. E., Moore, A. S., Rassnick, K. M., Cotter, S. engineering as well as use them as growth factors, which M., O'Sullivan, J. L. and Quesenberry, P. J. 2006. A combina- may eventually substitute cytokines. Investigations are tion chemotherapy protocol with dose intensification and autol- underway in humans in various fields spanning HSC expan- ogous bone marrow transplant (VELCAP-HDC) for canine sion to regenerative medicine. For these goals to be lymphoma. J. Vet. Intern. Med. 20: 355–364. achieved in dogs, functional researches must also be con- 12. Giampaolo, A., Pelosi, E., Valtieri, M., Montesoro, E., Ster- petti, P., Samoggia, P., Camagna, A., Mastroberardino, G., 76 K. IDE ET AL.

Gabbianelli, M., Testa, U. and et al. 1995. HOXB gene expres- 26. Marigo, V., Roberts, D. J., Lee, S. M., Tsukurov, O., Levi, T., sion and function in differentiating purified hematopoietic pro- Gastier, J. M., Epstein, D. J., Gilbert, D. J., Copeland, N. G., genitors. Stem. Cells 13 (Suppl. 1): 90–105. Seidman, C. E., Jenkins, N. A., Seidman, J. G., McMahon, A. 13. Giampaolo, A., Sterpetti, P., Bulgarini, D., Samoggia, P., P. and Tabin, C. 1995. Cloning, expression, and chromosomal Pelosi, E., Valtieri, M. and Peschle, C. 1994. Key functional location of SHH and IHH: two human homologues of the role and lineage-specific expression of selected HOXB genes Drosophila segment polarity gene hedgehog. Genomics 28: in purified hematopoietic progenitor differentiation. Blood 84: 44–51. 3637–3647. 27. McGinnis, W. and Krumlauf, R. 1992. Homeobox genes and 14. Gianni, A. M., Bregni, M., Siena, S., Brambilla, C., Di Nicola, axial patterning. Cell 68: 283–302. M., Lombardi, F., Gandola, L., Tarella, C., Pileri, A., Ravag- 28. McSweeney, P. A., Rouleau, K. A., Storb, R., Bolles, L., Wal- nani, F., Valagussa, P. and Bonadonna, G. 1997. High-dose lace, P. M., Beauchamp, M., Krizanac-Bengez, L., Moore, P., chemotherapy and autologous bone marrow transplantation Sale, G., Sandmaier, B., de Revel, T., Appelbaum, F. R. and compared with MACOP-B in aggressive B-cell lymphoma. Nash, R. A. 1996. Canine CD34: cloning of the cDNA and New Engl. J. Med. 336: 1290–1297. evaluation of an antiserum to recombinant protein. Blood 88: 15. Goodrich, L. V., Johnson, R. L., Milenkovic, L., McMahon, J. 1992–2003. A. and Scott, M. P. 1996. Conservation of the hedgehog/ 29. McSweeney, P. A., Rouleau, K. A., Wallace, P. M., Bruno, B., patched signaling pathway from flies to mice: induction of a Andrews, R. G., Krizanac-Bengez, L., Sandmaier, B. M., mouse patched gene by Hedgehog. Genes. Dev. 10: 301–312. Storb, R., Wayner, E. and Nash, R. A. 1998. Characterization 16. Haioun, C., Lepage, E., Gisselbrecht, C., Bastion, Y., Coiffier, of monoclonal antibodies that recognize canine CD34. Blood B., Brice, P., Bosly, A., Dupriez, B., Nouvel, C., Tilly, H., Led- 91: 1977–1986. erlin, P., Biron, P., Briere, J., Gaulard, P. and Reyes, F. 1997. 30. Murdoch, B., Chadwick, K., Martin, M., Shojaei, F., Shah, K. Benefit of autologous bone marrow transplantation over V., Gallacher, L., Moon, R. T. and Bhatia, M. 2003. Wnt-5A sequential chemotherapy in poor-risk aggressive non- augments repopulating capacity and primitive hematopoietic Hodgkin’s lymphoma: updated results of the prospective study development of human blood stem cells in vivo. Proc. Natl. LNH87-2. Groupe d’Etude des Lymphomes de l’Adulte. J. Acad. Sci. U.S.A. 100: 3422–3427. Clin. Oncol. 15: 1131–1137. 31. Murone, M., Rosenthal, A. and de Sauvage, F. J. 1999. Hedge- 17. Kaushansky, K., Shoemaker, S. G., Broudy, V. C., Lin, N. L., hog signal transduction: from flies to vertebrates. Exp. Cell Matous, J. V., Alderman, E. M., Aghajanian, J. D., Szklut, P. Res. 253: 25–33. J., VanDyke, R. E., Pearce, M. K. and Abrams, J. S. 1992. 32. Murone, M., Rosenthal, A. and de Sauvage, F. J. 1999. Sonic Structure-function relationships of interleukin-3. An analysis hedgehog signaling by the patched-smoothened com- based on the function and binding characteristics of a series of plex. Curr. Biol. 9: 76–84. interspecies chimera of gibbon and murine interleukin-3. J. 33. Neuner, E., Schumm, M., Schneider, E. M., Guenther, W., Clin. Invest. 90: 1879–1888. Kremmer, E., Vogl, C., Buttner, M., Thierfelder, S. and Kolb, 18. Kawagoe, H., Humphries, R. K., Blair, A., Sutherland, H. J. H. J. 1998. Studies on canine bone marrow long-term culture: and Hogge, D. E. 1999. Expression of HOX genes, HOX effect of stem cell factor. Vet. Immunol. Immunopathol. 61: 1– cofactors, and MLL in phenotypically and functionally defined 16. subpopulations of leukemic and normal human hematopoietic 34. Nusse, R. and Varmus, H. E. 1982. Many tumors induced by cells. Leukemia 13: 687–698. the mouse mammary tumor virus contain a provirus integrated 19. Khan, N. I. and Bendall, L. J. 2006. Role of WNT signaling in in the same region of the host genome. Cell 31: 99–109. normal and malignant hematopoiesis. Histol. Histopathol. 21: 35. Ogilvie, G. K. 2000. Clinical Use of Hemopoietic Growth Fac- 761–774. tors, pp. 884–888. In: Schalm’s Veterinary Hematology, 5th 20. Lawrence, H. J., Sauvageau, G., Ahmadi, N., Lopez, A. R., ed. (Feldman, B. V., Zinkl, J. G. and Jain N. C. eds.), Lippin- LeBeau, M. M., Link, M., Humphries, K. and Largman, C. scott Williams & Wilkins. 1995. Stage- and lineage-specific expression of the HOXA10 36. Peters, I. R., Peeters, D., Helps, C. R. and Day, M. J. 2007. homeobox gene in normal and leukemic hematopoietic cells. Development and application of multiple internal reference Exp. Hematol. 23: 1160–1166. (housekeeper) gene assays for accurate normalisation of canine 21. Lawrence, H. J., Sauvageau, G., Humphries, R. K. and Larg- gene expression studies. Vet. Immunol. Immunopathol. 117: man, C. 1996. The role of HOX homeobox genes in normal 55–66. and leukemic hematopoiesis. Stem. Cells 14: 281–291. 37. Petzer, A. L., Hogge, D. E., Landsdorp, P. M., Reid, D. S. and 22. Levine, M. and Hoey, T. 1988. Homeobox proteins as Eaves, C. J. 1996. Self-renewal of primitive human hematopoi- sequence-specific transcription factors. Cell 55: 537–540. etic cells (long-term-culture-initiating cells) in vitro and their 23. Lindblad-Toh, K., Wade, C. M., Mikkelsen, T. S. and et al. expansion in defined medium. Proc. Natl. Acad. Sci. U.S.A. 93: 2005. Genome sequence, comparative analysis and haplotype 1470–1474. structure of the domestic dog. Nature 438: 803–819. 38. Petzer, A. L., Zandstra, P. W., Piret, J. M. and Eaves, C. J. 24. Magli, M. C., Barba, P., Celetti, A., De Vita, G., Cillo, C. and 1996. Differential cytokine effects on primitive (CD34+CD38–) Boncinelli, E. 1991. Coordinate regulation of HOX genes in human hematopoietic cells: novel responses to Flt3-ligand and human hematopoietic cells. Proc Natl Acad Sci U.S.A. 88: thrombopoietin. J. Exp. Med. 183: 2551–2558. 6348–6352. 39. Philip, T., Guglielmi, C., Hagenbeek, A., Somers, R., Van der 25. Magnusson, M., Brun, A. C., Miyake, N., Larsson, J., Ehinger, Lelie, H., Bron, D., Sonneveld, P., Gisselbrecht, C., Cahn, J. M., Bjornsson, J. M., Wutz, A., Sigvardsson, M. and Karlsson, Y., Harousseau, J. L., Coiffier, B., Biron, P., Mandelli, F. and S. 2007. HOXA10 is a critical regulator for hematopoietic stem Chauvin, F. 1995. Autologous bone marrow transplantation as cells and erythroid/megakaryocyte development. Blood 109: compared with salvage chemotherapy in relapses of chemo- 3687–3696. therapy-sensitive non-Hodgkin’s lymphoma. New Engl. J. GENE TRANSCRIPTS IN BONE MARROW PROGENITORS 77

Med. 333: 1540–1545. 21–30. 40. Reya, T., Duncan, A. W., Ailles, L., Domen, J., Scherer, D. C., 48. Taghon, T., Stolz, F., De Smedt, M., Cnockaert, M., Verhas- Willert, K., Hintz, L., Nusse, R. and Weissman, I. L. 2003. A selt, B., Plum, J. and Leclercq, G. 2002. HOX-A10 regulates role for Wnt signalling in self-renewal of haematopoietic stem hematopoietic lineage commitment: evidence for a monocyte- cells. Nature 423: 409–414. specific . Blood 99: 1197–1204. 41. Santini, G., Salvagno, L., Leoni, P., Chisesi, T., De Souza, C., 49. Taghon, T., Thys, K., De Smedt, M., Weerkamp, F., Staal, F. Sertoli, M. R., Rubagotti, A., Congiu, A. M., Centurioni, R., J., Plum, J. and Leclercq, G. 2003. Homeobox gene expression Olivieri, A., Tedeschi, L., Vespignani, M., Nati, S., Soracco, profile in human hematopoietic multipotent stem cells and T- M., Porcellini, A., Contu, A., Guarnaccia, C., Pescosta, N., cell progenitors: implications for human T-cell development. Majolino, I., Spriano, M., Vimercati, R., Rossi, E., Zambaldi, Leukemia 17: 1157–1163. G., Mangoni, L. and Rizzoli, V. 1998. VACOP-B versus 50. Takahashi, Y., Hamada, J., Murakawa, K., Takada, M., Tada, VACOP-B plus autologous bone marrow transplantation for M., Nogami, I., Hayashi, N., Nakamori, S., Monden, M., Miya- advanced diffuse non-Hodgkin's lymphoma: results of a pro- moto, M., Katoh, H. and Moriuchi, T. 2004. Expression pro- spective randomized trial by the non-Hodgkin’s Lymphoma files of 39 HOX genes in normal human adult organs and Cooperative Study Group. J. Clin. Oncol. 16: 2796–2802. anaplastic thyroid cancer cell lines by quantitative real-time 42. Sauvageau, G., Lansdorp, P. M., Eaves, C. J., Hogge, D. E., RT-PCR system. Exp. Cell Res. 293: 144–153. Dragowska, W. H., Reid, D. S., Largman, C., Lawrence, H. J. 51. Ueda, T., Tsuji, K., Yoshino, H., Ebihara, Y., Yagasaki, H., and Humphries, R. K. 1994. Differential expression of Hisakawa, H., Mitsui, T., Manabe, A., Tanaka, R., Kobayashi, homeobox genes in functionally distinct CD34+ subpopulations K., Ito, M., Yasukawa, K. and Nakahata, T. 2000. Expansion of of human bone marrow cells. Proc. Natl. Acad. Sci. U.S.A. 91: human NOD/SCID-repopulating cells by stem cell factor, 12223–12227. Flk2/Flt3 ligand, thrombopoietin, IL-6, and soluble IL-6 recep- 43. Sauvageau, G., Thorsteinsdottir, U., Eaves, C. J., Lawrence, H. tor. J. Clin. Invest. 105: 1013–1021. J., Largman, C., Lansdorp, P. M. and Humphries, R. K. 1995. 52. Van Den Berg, D. J., Sharma, A. K., Bruno, E. and Hoffman, Overexpression of HOXB4 in hematopoietic cells causes the R. 1998. Role of members of the Wnt gene family in human selective expansion of more primitive populations in vitro and hematopoiesis. Blood 92: 3189–3202. in vivo. Genes. Dev. 9: 1753–1765. 53. van Oostveen, J., Bijl, J., Raaphorst, F., Walboomers, J. and 44. Shipp, M. A., Abeloff, M. D., Antman, K. H., Carroll, G., Meijer, C. 1999. The role of homeobox genes in normal Hagenbeek, A., Loeffler, M., Montserrat, E., Radford, J. A., hematopoiesis and hematological malignancies. Leukemia 13: Salles, G., Schmitz, N., Symann, M., Armitage, J. O., Philip, T. 1675–1690. and Coiffier, B. 1999. International Consensus Conference on 54. Wetmore, C. 2003. Sonic hedgehog in normal and neoplastic High-Dose Therapy with Hematopoietic Stem Cell Transplan- proliferation: insight gained from human tumors and animal tation in Aggressive Non-Hodgkin's Lymphomas: report of the models. Curr. Opin. Genet. Dev. 13: 34–42. jury. J. Clin. Oncol. 17: 423–429. 55. Wicking, C., Smyth, I. and Bale, A. 1999. The hedgehog sig- 45. Sidow, A. 1992. Diversification of the Wnt gene family on the nalling pathway in tumorigenesis and development. Oncogene ancestral lineage of vertebrates. Proc. Natl. Acad. Sci. U.S.A. 18: 7844–7851. 89: 5098–5102. 56. Zhang, X. B., Schwartz, J. L., Humphries, R. K. and Kiem, H. 46. Sorrentino, B. P. 2004. Clinical strategies for expansion of hae- P. 2007. Effects of HOXB4 overexpression on ex vivo expan- matopoietic stem cells. Nat. Rev. Immunol. 4: 878–888. sion and immortalization of hematopoietic cells from different 47. Staal, F. J. and Clevers, H. C. 2005. WNT signalling and hae- species. Stem. Cells 25: 2074–2081. matopoiesis: a WNT-WNT situation. Nat. Rev. Immunol. 5: