Acute Leukemias

Specific small nucleolar RNA expression profiles in acute leukemia

Abstract

Apart from microRNAs, little is known about the regulation of expression of non-coding RNAs in cancer. We investigated whether small nucleolar RNAs (snoRNAs) accumulation displayed specific signatures in acute myeloblastic and acute lymphoblastic leukemias. Using microarrays and high-throughput quantitative PCR (qPCR), we demonstrate here that snoRNA expression patterns are negatively altered in leukemic cells compared with controls. Interestingly, a specific signature was found in acute promyelocytic leukemia (APL) with ectopic expression of SNORD112–114 snoRNAs located at the DLK1-DIO3 locus. In vitro experiments carried out on APL blasts demonstrate that transcription of these snoRNAs was lost under all-trans retinoic acid-mediated differentiation and induced by enforced expression of the PML–RARalpha fusion protein in negative leukemic cell lines. Further experiments revealed that the SNORD114-1 (14q(II-1)) variant promoted cell growth through cell cycle modulation; its expression was implicated in the G0/G1 to S phase transition mediated by the Rb/p16 pathways. This study thus reports three important observations: (1) snoRNA regulation is different in normal cells compared with cancer cells; (2) a relationship exists between a chromosomal translocation and expression of snoRNA loci; and (3) snoRNA expression can affect Rb/p16 cell cycle regulation. Taken together, these data strongly suggest that snoRNAs have a role in cancer development.

Introduction

In cancer cells much effort has been devoted to the understanding of microRNA expression.1, 2 Other classes of non-coding RNAs (ncRNAs) were originally considered products of constitutively expressed housekeeping genes or representing transcriptional noise so until now have been largely neglected. Small nucleolar RNAs (snoRNAs) are 60–300 nucleotide-long ncRNAs that are excised from the intron regions of pre-mRNAs, usually those encoding fundamental housekeeping proteins.3, 4, 5 There are two structurally distinct classes of snoRNAs, called the box C/D and H/ACA snoRNAs, which function mainly as guide RNAs in the site-specific 2′-O-methylation and pseudouridylation of rRNAs, respectively.3, 6, 7 However, a subgroup of so-called ‘orphan’ snoRNAs exists that lacks any apparent complementarity to cellular RNAs and therefore has unknown function. A few orphan snoRNAs, such as SNORD112 to SNORD116, are encoded by multiple, tandemly arranged intronic genes that display high sequence similarity.8, 9

Recently, new unexpected regulatory functions have been described for human snoRNAs. In terms of pathologies, the human HBII-52 cluster that expresses several variants of the SNORD115 snoRNA was implicated in a neurodevelopmental disorder, the Prader-Willi syndrome.10 Moreover, recent reports suggest that snoRNAs could be relevant to oncogenesis and solid cancer prognosis.11, 12, 13, 14 Thus, we hypothesized that, in cancer, expression of human snoRNAs is regulated and displays specific signatures. We used the acute myeloblastic leukemia (AML) model for two reasons: (1) it represents an interesting model of oncogenesis that is mainly based on blockade of cell differentiation and (2) chromosomal anomalies are frequently observed due to recurrent translocations forming oncogenic fusion proteins. As microRNA signatures can be specific to subsets of chimeric proteins in leukemia,15 we assumed that this was also the case for snoRNA profiles. We tested snoRNAs already present on microarrays together with high-throughput quantitative reverse transcription (RT-qPCR) on microfluidic chips.

Materials and methods

Patients and healthy donors

Fresh and thawed samples from AML or acute lymphoblastic leukemia (ALL) patients (Supplementary Table 1) and from healthy donors (bone marrow, CD3+, CD19+ and CD33+ cells) were obtained as described in Supplementary Methods.

RNA preparation and cDNA synthesis

Total RNA was prepared by extraction with TRIzol (Invitrogen, Grand Island, NY, USA) from mononuclear-sorted cells (CD33+, CD3+ and CD19+) and fresh or thawed samples (from AML and ALL patients). RNA integrity was evaluated using the RNA 6000 Nano Chip kit (Agilent Technologies, Massy, France). Only RNAs with a RIN>7.5 were used. RNAs were reverse transcribed into cDNA using 3 μM of random primer mix (Biolabs, Evry, France) and Superscript III reverse transcriptase (Invitrogen). For high-throughput qPCR, RNAs were polyadenylated by the Poly(A) Polymerase Tailing Kit (EPICENTRE Biotechnologies, Madison, WI, USA) before RT.

GeneChip microRNA arrays

Four hundred nanograms of total RNA were hybridized on GeneChip microRNA array (Affymetrix, High Wycombe, UK). Unlogged data were treated with the dChip software (Boston, MA, USA) with a specific cutoff (filter) to generate dendrograms. The cutoff value of 0.5 applied to microRNAs could not be used for snoRNAs because their expression was lower, thus, we used a cutoff value of 0.25. Rows (corresponding to RNAs of interest) were standardized by respective RNA means of expression and distant metric system applied is correlation. Sample clustering was carried out by the centroid method. Significance analysis of microarrays was used to analyze statistical significance of the results.

High-throughput qPCR

The Fluidigm high-throughput qPCR method (Biomark, Fluidigm, South San Francisco, CA, USA) was used as described previously16 and developed in Supplementary Methods. Primers used for amplification are described in Supplementary Table 2. Data were analyzed as for qPCR using 5S rRNA as the housekeeping gene (GAPDH and ACTIN were also tested). Non-supervised clustering was realized using the dChip software as for microarray analysis with a cutoff value of 0.85. The unpaired t-test was used for statistical analyses.

qPCR and data analysis

For real-time PCR, the 25 μl reaction contained 1xSYBR Green PCR Master Mix (Sigma-Aldrich, St Louis, MO, USA), with each primer at 0.3 μM, and 5 μl of cDNA (see above) diluted at 1:5. The primer efficiency (>85%) was checked before the experiments. The 2−ΔCT (threshold cycle) values were used for the analysis (5S rRNA was selected as the housekeeping gene).

In vitro differentiation of acute promyelocytic leukemia (APL) blast cells

APL blasts were cultured (Supplementary Methods) in the presence of 1 μM all-trans retinoic acid (ATRA) (Sigma, St Louis, MO, USA) diluted in ethanol, or in ethanol alone. After 3 and 5 days, the percentage of differentiated cells was assessed by cell morphology with May-Grünwald-Giemsa stained cytocentrifuge slides and by expression of CD11b as determined by flow cytometry. The monoclonal antibodies used for staining were as follows: CD11b-PE-Cy7 (BD Biosciences, Le pont de Claix, France) and CD33-PE-Cy5 (BD Biosciences) and isotype-matched control conjugates. Flow cytometry was performed using a BD-LSRII flow cytometer (BD Biosciences).

PML-RARalpha transient transfection into a KG-1 cell line

A cDNA encoding PML-RARalpha (kindly provided by Valerie Lallemand-Breitenbach; St Louis Hospital, Paris, France) was cloned into an expression vector pCDNA3 (Invitrogen). KG-1 cells (ATCC_CCL-246, Molsheim, France) were transfected with pCDNA3 or pCDNA3-P/R using nucleofection Amaxa Technology (Lonza, Basel, Switzerland) and Ingenio universal solution (Mirus Bio LLC, Madison, WI, USA). SnoRNA expression was monitored by RT-qPCR after 48 and 72 h of transfection. PML-RARalpha expression was monitored by RT-PCR using primer F (5′-IndexTermCAGTGTACGCCTTCTCCATCA-3′) and primer R (5′-IndexTermGCTTGTAGATGCGGGGTAGA-3′).

14q(II-1) overexpression and silencing

Sequences containing snoRNA (Figure 4b) were amplified by a Pfu-Ultra enzyme, sequenced and then cloned. The latter construct or empty vector were transiently transfected by Amaxa technology (Solution V, T16-program) in K562 cell line (ATCC_CCL-243). Cell growth was evaluated by Malassez cell counting or MTS ([3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt) proliferation assay (Promega, Madison, WI, USA). Overexpression of 14q(II-1) snoRNA was checked by qPCR using the primer described in Supplementary Table 2.

14q(II-1) silencing in K562 cell line was realized using siRNA (siRNA-14q(II-1), Sigma; siRNA-negative control, Eurogentec, Angers, France) transfection (25 nM as a final siRNA concentration) using Lipofectamine RNAiMax (Invitrogen). Effective targeting after 48 h was analyzed by qPCR as described above. Total Rb and p16 protein expressions were evaluated by western blotting using a mouse anti-human total Rb monoclonal antibody (Sc-102), mouse anti-human p16 (CDKN2A) antibody (Sc-9968) and anti-β-actin (MAB150L). Rb phosphorylation on Ser780 was analyzed using rabbit anti-human phosphor-ser780-Rb (Sc-9307).

Cell cycle analysis

Cells were washed with phosphate-buffered saline and fixed in cold 70% ethanol for 20 min. Cells were then washed twice with phosphate-buffered saline-0.1% bovine serum albumin and once with phosphate-buffered saline then labeled using propidium iodide staining for 30 min (Invitrogen). Cell cycle distribution was evaluated by fluorescence analysis on a FACScan cytometer (Beckton Dickinson Biosciences). Cell doublets were excluded and 20 000 events per condition were analyzed.

Results

Global downregulation of snoRNAs in ALL and AML

To characterize specific snoRNA expression patterns in AML, we first used GeneChip microRNA arrays that allow the simultaneous screening of almost all snoRNAs and microRNAs that are listed in the ENSEMBL and snoRNABase databases (http://www.ensembl.org and http://www-snorna.biotoul.fr/). SnoRNA expression levels were screened in a cohort of 12 AML samples and compared with 4 CD33+ myeloid-sorted cells from healthy donors. By using non-supervised clustering, we observed a global downregulation of snoRNAs compared with their non-neoplastic counterparts, thus distinguishing leukemic cells from normal cells (Figure 1a). To verify that these observations were robust, we used the same method to analyze ALL samples matched with CD3/CD19-sorted cells. This confirmed a global snoRNA downregulation that was the main characteristic differentiating healthy donors from cancer samples (Figure 1a). SnoRNA profiles from CD3+ and CD19+ control samples were also analyzed and demonstrated that the latter can be regarded as a single-control group (Supplementary Figure 1). To evaluate the accuracy of the snoRNA signatures, we also analyzed microRNA expression profiles and compared them to those reported in the literature. We successfully reproduced the microRNA signatures previously described for AML and ALL, confirming that our experimental design was robust (Figure 1b).17

Figure 1
figure1

snoRNA downregulation discriminates neoplastic from non-neoplastic cells. (a) In all, 12 AML samples, 10 ALL samples, 5 lymphoid and 4 myeloid controls were used to study snoRNA expression using GeneChip microRNA arrays. SnoRNAs from AML and ALL patients’ cells were quantified in enlarged cohorts by high-throughput qPCR and compared with those of CD33+ myeloid cells and CD3+ or CD19+ lymphoid cells from healthy donors. Non-supervised clustering by the dChip software was realized using a filter cutoff of 0.25 (for microarray) and 0.5 (for high-throughput qPCR). Each column represents a sample and each row represents a single snoRNA. The color displayed shows the logarithm of the expression changes, where varying shades of red and green indicate up and downregulation, respectively. The color key for sample labeling is at the top left. Patients who lacked the common translocations seen in these diseases were designated as ‘_NT’. (b) SAM analysis was used to find microRNAs that differentiate AML from ALL patients on GeneChip microRNA arrays. The unpaired t-test was used as the statistical test (*P<0.05; **P<0.01 and ***P<0.001). Each graph represents means±s.e.m.

To further confirm these results we used the Biomark digital PCR system (Fluidigm) for high-throughput expression studies on 62 selected snoRNAs, among which 46 significantly varied on microarrays. The remaining 16 snoRNAs, 5 guide and 11 orphan RNAs did not vary on the microarrays. In addition, we also tested four spliceosomal small nuclear RNAs (U2, U4, U6 and U12) and two snoRNAs (U8 and U3) that are synthesized from independent genes. The 5S ribosomal RNA was used to normalize snoRNA expression due to its stability and length. We extended our RT-qPCR experiments to 26 AML cases. Non-supervised clustering indicated that the snoRNA expression profiles of the AML samples were different from those of CD33+ myeloid cells (Figure 1a), CD34+ progenitors or total bone marrow (Supplementary figure 1). It is worth mentioning, however, that Flt3-ITD-mutated AML cosegregated with controls, suggesting that this group of leukemia is distinct with regard to snoRNA profiles (Figure 1a). To confirm the identity of snoRNA expression patterns obtained in ALL cases by microarray analysis, we also checked the delineation of ALL samples compared with ‘lymphoid controls’ in a larger cohort (25 ALL cases). In leukemic samples, 61% of snoRNAs tested were significantly underexpressed compared with control cells (Figure 1a; Supplementary Table 3). This is in keeping with the global underexpression reported for microRNAs in several types of cancers.1, 18, 19 Thus, RT-qPCR experiments validated the global downregulation of snoRNA expression previously noticed on microarrays (Figure 1a).

Specific snoRNA expression profiles identify APL cases

Although snoRNAs appeared to be globally downregulated in leukemia cells, few snoRNAs located into clusters are overexpressed in some AML cases. Their expression was associated or not to specific chromosomal translocations (Supplementary figure 2). Thus, through microarray analysis we noticed ectopic expression of a set of snoRNAs from the DLK1-DIO3 locus in APL samples carrying identical PML-RARalpha_bcr1 translocations (Figure 1a). DLK1-DIO3 snoRNAs are located within the introns of the maternally expressed gene 8 (Meg8), which carries 1, 9 and 31 highly related copies of SNORD112 (14q(0)), SNORD113 (14q(I)) and SNORD114 (14q(II)) snoRNA genes, respectively (Figure 2a).8, 20 As expected (because of their restricted expression pattern), they were not detected in controls (healthy CD33+ myeloid cells) or in PML/RARα-negative AML samples (Figures 2b and c). Using microarrays, we observed that only eight sequence variants of SNORD114, five variants of SNORD113 and SNORD112 were expressed in APL patients (Figure 2b). Of note, the level of expression of these particular snoRNAs varied significantly within the same cluster.

Figure 2
figure2

The DLK1-DIO3 locus is deregulated in APL samples. Inside the DLK1-DIO3 locus (a), three classes of ncRNA (long ncRNA, microRNA and clustered snoRNA) are represented. Owing to imprinting, the genes are either paternally (Pat) or maternally (Mat) expressed under the control of an intergenic differentially-methylated region (IG-DMR). The microRNA and snoRNA loci are represented by diagonally and vertically hatched boxes. (b) The signal intensities of DLK1-DIO3 snoRNAs in AML samples assayed on GeneChip microRNA arrays demonstrated that only a small set of snoRNA was expressed. Shaded boxes represent signal noise from the chip and _st and x_st designed two different probes for one snoRNA. Only the snoRNAs whose expression was significantly different from noise signal were represented on the graph. (c) The snoRNA signature obtained using microarrays was confirmed by qPCR in 8 APL (PR), 18 AML PML-RARalpha-negative (nPR) and 7 CD33+ myeloid controls. (d) We confirmed a specific DLK1-DIO3 microRNA profile of APL samples by GeneChip microRNA array. Expression of microRNAs in four APL and eight AML PML-RARα-negative samples was normalized to those of four CD33+ myeloid controls. Only significant changes in microRNA expression were presented on the graphs. The unpaired t-test was used to determine the statistical significance of the differences (*P<0.05; **P<0.01 and ***P<0.001). Each graph represents means±s.e.m.

DLK1-DIO3 snoRNA expression profiles were confirmed by RT-qPCR (Figure 2c). Owing to their high sequence homology, we selected only the snoRNAs for which we could design specific primers: SNORD112 (14q(0)), SNORD113-6 (14q(I-6)), SNORD113-7 (14q(I-7)), SNORD113-8 (14q(I-8)), SNORD113-9 (14q(I-9)) and SNORD114-1 (14q(II-1)). We first tested SNORD113-3 whose expression did not vary when assayed by microarrays. RT-qPCR confirmed the lack of expression indicating that, despite high sequence homology of SNORD113 snoRNAs, the PCR-based quantification was specific.

Several microRNA genes are located around the SNOR112-114 gene cluster in the DLK1-DIO3 locus and seem to have the same expression pattern as their neighboring snoRNAs.21, 22 To validate snoRNA results from microarrays in APL, we examined the expression of the microRNAs from the DLK1-DIO3 locus in APL and PML-RARalpha-negative cells using GeneChip microRNA arrays. When compared with AML samples, we confirmed the significant overexpression of hsa-miR-127, miR-370 and miR-154, as already reported in APL, as well as 11 other microRNAs (Figure 2d).15

PML-RARalpha impacts on DLK1-DIO3 snoRNA expression

To test whether snoRNA expression from the DLK1-DIO3 locus is controlled by PML-RARalpha, we treated leukemic blast cells from three APL patients in vitro with ATRA, which directly targets PML-RARalpha, leading to granulocytic differentiation.23 Blast differentiation was monitored by May-Grünwald-Giemsa staining and CD11b overexpression (Figures 3a and b). RT-qPCR revealed that expression of the SNORD112, SNORD113 and SNORD114 gene cluster in the DLK1-DIO3 locus decreased during differentiation (Figure 3c). Interestingly, levels of the six snoRNAs did not decrease at the same time. After 3 days of treatment, only SNORD113-8 (14q(I-8)) was significantly underexpressed. After 5 days, all snoRNAs were significantly decreased. These findings demonstrate that expression of orphan snoRNAs encoded within the DLK1-DIO3 locus is linked to PML-RARalpha. Moreover, these data corroborate the notion that within the 14q polycistronic snoRNA clusters expression of individual snoRNAs is regulated in a surprisingly independent manner. To further investigate whether the PML-RARα protein chimera influenced snoRNA expression in the DLK1-DIO3 locus, we transfected the human AML M1 (non-APL) cell line KG-1 with a plasmid that expressed PML-RARalpha_bcr1. At both 2 and 3 days after transfection, we verified expression of PML-RARalpha mRNA (Figure 3d) and measured the levels of SNORD112, SNORD113-6, SNORD113-7, SNORD113-8, SNORD113-9 and SNORD114-1 snoRNAs by RT-qPCR. In non-transfected control KG-1 cells, there was a low level of basal expression of all these snoRNAs (<2−12 relative expression compared with 5S rRNA). Transient expression of PML-RARalpha resulted in an upregulation of all the snoRNAs tested at 48 h post-transfection (Figure 3e). Similar data were obtained when we expressed PML-RARalpha in KG1a cells, an AML M0 (non-APL) cell line (data not shown).

Figure 3
figure3

DLK1-DIO3 snoRNA expression was lost by ATRA treatment and increased by PML-RARalpha. (a) Blasts from APL patients were treated with ATRA for 5 days. Blast differentiation was assayed by marker phenotype (CD33 and CD11b) just before treatment (day 0, D0) and at day 3 (D3) and day 5 (D5) by plotting the increase of CD33+ and CD11b+ cells on a density dot plot. (b) Differentiation was additionally evaluated by May-Grunwald-Giemsa staining of control and ATRA-treated blasts. Nuclear morphology and cytoplasm granularity changes in ATRA-treated blasts indicated blast differentiation. (c) SNORD112 (14q(0)), SNORD113-6 SNORD113-7, SNORD113-8 and SNORD113-9) (14q(I-6,-7,-8 and -9)) and SNORD114-1 (14q(II-1)) expressions were monitored at days 3 and 5 by qPCR. Expression of snoRNAs in ATRA-treated cells was normalized to those under control conditions and expression fold changes were represented on the graph. (d) KG-1 cells were transfected with a plasmid expressing PML-RARalpha. Expression of PML-RARalpha was checked by RT-PCR 48 and 72 h after transfection. (e) Expression of 14q(0), 14q(I-6,-7,-8 and -9) and (14q(II-1)) was monitored by qPCR until 48 h after transfection. A paired t-test was used as the statistical test (*P<0.05; **P<0.01; ***P<0.001). Each graph represents means±s.e.m.

The SNORD114-1 (14q(II-1)) variant is implicated in cell growth through cell cycle regulation

To test whether DLK1-DIO3 snoRNAs had a role in APL pathogenesis, we selected the SNORD114-1 (14q(II-1)) variant to analyze its impact on cell growth. This variant was chosen because among the snoRNAs it displayed the highest expression of the locus and because expression of this variant was sustained the longest following treatment with ATRA. The 14q(II-1) variant was cloned and transiently transfected into the K562 cell line, which represents the best model described for studying the DLK1-DIO3 locus (that is, this locus is transcriptionally active in these cells).24 To upregulate 14q(II-1) snoRNA, a vector was constructed that contained the intron hosting the snoRNA sequence and its flanking exons. RT-qPCR quantification of expression of the 14q(II-1) variant revealed a mean fold change of 57 in transfected cells (Figure 4a). We then sought to ensure that snoRNA processing from this construct was effective. RT-PCR was carried out with primers specific to the intronic lariat (the flanking sequences of snoRNA) or primers specific to the mature snoRNA sequence. Thus, we noticed a strong overexpression of the mature snoRNA sequence compared with that of the intron lariat (Figure 4b). Overexpression of the 14q(II-1) variant in the K562 cell line induced a significant increase in cell proliferation as detected with both an MTS cell proliferation assay (Figure 4c) and Malassez counting (Figure 4d). Conversely, snoRNA silencing carried out with 25 nM of siRNA 14q(II-1) in K562 cells caused a 50% reduction in snoRNA expression (Figure 4e) and a significant decrease in cell growth at 24 and 48 h (Figure 4f). We then looked at the cell cycle distribution in siRNA-treated cells and demonstrated that after 48h of treatment 14q(II-1) snoRNA targeting induced a significant increase of cells in G0/G1 phase and, conversely, a significant decrease in S phase (Figure 5a).

Figure 4
figure4

The 14q(II-1) snoRNA variant impacts on cell growth and cell cycle. (a) 14q(II-1) (SNORD114-1) was cloned and transfected into a K562 cell line. Its overexpression was verified by qPCR 48 h after transfection. (b) To overexpress the 14q(II-1) snoRNA, sequences containing the intron (in which the snoRNA sequence was present) and its flanking exon were cloned. We ensured that only the mature snoRNA sequence was overexpressed by RT-PCR for the snoRNA and intronic lariat sequences. ‘Blank’ corresponds to RT-PCR on the mature snoRNA sequence using cDNA template from DLK1-DIO3-negative cells. Cell growth of 14q(II-1) overexpressing cells was estimated using an MTS proliferation assay (c) and Malassez counting (d) and compared with cells transfected with the empty vector. Conversely, 14q(II-1) silencing by siRNA was realized and the effective targeting was checked by qPCR (e). (f) Cell growth was estimated 24 and 48 h after transfection by Malassez counting. Paired t-tests were used as a statistical test (*P<0.05; **P<0.01 and ***P<0.001). Each graph represents means±s.e.m.

Figure 5
figure5

14q(II-1) snoRNA variant silencing induces a decrease in cell growth via the Rb pathways. (a) 14q(II-1) snoRNA silencing by siRNA was realized in a K562 cell line. We first verified the decrease in cell growth 48 h after transfection, then evaluated cell cycle distribution of transfected cells by immunoperoxidase staining and flow cytometry analysis. Graph represented means±s.e.m. of cell frequency in each phase of the cell cycle. Cell doublets were excluded by the PE-A/PE-W method. A paired t-test was used as the statistical test (*P<0.05; **P<0.01 and ***P<0.001). (b) The increase of cells in G0/G1 phase was verified by the effect on total Rb protein expression on a western blot under silenced 14q(II-1) conditions. (c) Conversely, western blotting showed a decrease of total Rb protein expression in 14q(II-1) overexpressing cells. Phosphorylation on ser780 of the remaining Rb was demonstrated by western blot analysis. Each graph represents means±s.e.m.

The Rb/p16 pathways are involved in snoRNA-mediated cell growth

The effect of 14q(II-1) snoRNA on cell growth and the cell cycle prompted us to focus on what signaling pathways were potentially activated. Indeed, in K562 cells important cell cycle regulators such as p16(CDKN2A) and p53 are known to be inactivated.25, 26 Therefore, we investigated the impact of 14q(II-1) snoRNA silencing on the Rb pathway using western blot analysis and noticed a clear-cut overexpression of the corresponding protein (Figure 5b). This result is in keeping with those described above for cell cycle distribution. To validate these observations, we analyzed the effect of 14q(II-1) overexpression on Rb. As expected, we observed an underexpression of total Rb protein in 14q(II-1) overexpressing cells (Figure 5c). In addition, we noticed that the remaining Rb protein was hyperphosphorylated on serine-780 compared with that under control conditions (Figure 5c), suggesting that cell cycle activity was increased. To test whether this effect on Rb was also observed in APL patients, 14q(II-1) silencing by siRNA was performed using APL blasts from five patients. We confirmed an effective targeting of the 14q(II-1) variant at the RNA level (Figure 6a). Interestingly, a cell growth decrease of 30% was also observed after 48h (Figure 6b). We therefore analyzed the expression of total Rb protein in APL blasts transfected with siRNA-14q(II-1) or an siRNA-negative control. In three of these patients, we confirmed an increase in Rb expression after 14q(II-1) silencing (Figure 6c). In the two remaining patients, Rb expression was almost undetectable whereas p16 expression was detected. It has been previously reported that in cancer cells p16 is often deleted or mutated and Rb/p16 expressions are inversely regulated.27, 28, 29 In APL cells with a basal level of p16, siRNA-14q(II-1) fosters the expression of the protein (Figure 6c). Similar to the K562 cell line, we could not find p53 expression (using western blot analysis with two different antibodies) in APL patients’ cells. Overall, we demonstrated that 14q(II-1) snoRNA modulated cell growth through the Rb pathway in both a K562 cell line and in APL cells ex vivo. Interestingly, in some cases we observed that the p16 pathway could be an alternative regulator of the cell cycle.

Figure 6
figure6

14q(II-1) snoRNA effect on cell growth and Rb/p16 pathways is confirmed in APL blasts. (a) Blasts from APL patients were transfected with siRNA-14q(II-1) or siRNA-negative control (siRNA-negative). Effective targeting of the 14q(II-1) snoRNA was evaluated by qPCR and means±s.e.m. of expression fold change were represented on the graph. (b) Cell growth of APL blasts after 48 h of siRNA treatment was estimated using Malassez cell counting. Paired t-tests were used as the statistical test (*P<0.05; **P<0.01 and ***P<0.001). The impact of siRNA-14q(II-1) transfection on total Rb (c) or p16(CDKN2A) (d) protein expression was analyzed by western blot analysis.

Discussion

SnoRNAs represent an abundant group of non-protein-coding RNAs with their main function being ribosomal RNA maturation.3 Recent new cellular functions have led us to hypothesize that snoRNA accumulation patterns could be modulated in cancer.10, 11, 12, 14, 30, 31, 32, 33, 34, 35 Acute leukemia is a good model to try to answer this question as it is characterized by restrictive oncogenic events leading to a blockade in differentiation.

Using microarrays and a high-throughput qPCR strategy, we showed that neoplastic samples display specific snoRNA accumulation patterns allowing us to differentiate between tumor and control samples. Thus, as already reported for microRNAs,1, 18, 19 global downregulation is a characteristic feature of snoRNAs in cancer cells. Despite a large body of data on microRNAs, there is no clear explanation for this phenomenon. For snoRNAs however, their location within the introns of housekeeping genes implies that downregulation does not result from a transcriptional default but rather from disbranching alterations of the splicing lariat.

Despite global downregulation of snoRNAs in cancer cells, specific signatures were noticed in tumor subtypes that corresponded to ectopic expression of clusters. In particular, a specific signature was noticed in APL cases corresponding to transcriptional activation of restricted chromosomal regions. Thus, DLK1-DIO3 snoRNAs are part of such a chromosomal region containing clustered intronic orphan snoRNAs produced from long nuclear-retained non-coding mRNA-like transcripts such as Meg8.8, 20 The Meg8 transcript has no known functions, and the lack of sequence conservation between species suggests that spliced Meg8 ncRNAs yield only intronic SNORD112, SNORD113 and SNORD114 snoRNAs. We made the interesting observation that only 8 out of 31 sequence variants of SNORD114 and 5 out 9 variants of SNORD113 and SNORD112 were expressed in APL patients. The regulatory mechanism of differential snoRNA expression from the Meg8 region remains unclear, although this region is believed to be transcribed from one promoter (this has not been confirmed experimentally).20 It is possible that the Meg8 region has multiple internal promoters rather than being a single transcription unit. Alternatively, expression of these snoRNAs may be regulated by differential splicing of their host introns, as removal and disbranching of introns are essential for snoRNA production. We sought to determine the impact of PML-RARalpha on DLK1-DIO3 snoRNA expression under ATRA treatment. This corroborated the notion that within these Meg8 polycistronic clusters expression of snoRNAs are either independent from each other or produced in specific subsets. PML-RARalpha seems to be implicated in DLK1-DIO3 snoRNA expression but we could not demonstrate a direct relationship. Regulatory elements that control gene and small ncRNA expression at this locus remain poorly characterized. Nevertheless, consistent with a role for PML-RARalpha in DLK1-DIO3 transcriptional activation, Martens et al.36 recently reported a binding site for PML-RARalpha close to the DLK1 gene using chip sequencing. As AML subtypes display specific methylation profiles, another hypothesis is that a global modification of imprinting by PML-RARalpha leads to a deregulation of the DLK1-DIO3 locus.37 Indeed, apart from DLK1, other elements in this locus display tumor suppressor effects in other models.24, 38, 39, 40 Thus, DLK1-DIO3 snoRNAs may be key players participating in APL leukemic cell proliferation in the context of late differentiation blockade. In line with this, we showed that an individual snoRNA, the SNORD114-1 (14q(II-1)) variant, could impact on cell growth through a negative regulation of the cell cycle and the Rb pathway. Low Rb protein levels have been previously noted in APL cells 41, 42 and the expression of DLK1-DIO3 snoRNAs could negatively regulate the Rb pathway through the control of total Rb protein. In addition, the remaining Rb protein is probably inactivated by direct binding to PML-RARalpha as already shown.43

In conclusion, we report herein a global downregulation of snoRNAs in acute leukemia cells when compared with non-neoplastic counterparts, similar to that reported previously for microRNAs. The hallmark signatures observed for APL support the hypothesis that genes encoding snoRNAs, similar to microRNAs, might be specifically regulated by chimeric transcription factors like PML-RARalpha, in a background of chromosome instability, and thus may have an important role in oncogenesis.

References

  1. 1

    Lu J, Getz G, Miska EA, Alvarez-Saavedra E, Lamb J, Peck D et alMicroRNA expression profiles classify human cancers. Nature 2005; 435: 834–838.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  2. 2

    Calin GA, Croce CM . MicroRNA signatures in human cancers. Nat Rev Cancer 2006; 6: 857–866.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  3. 3

    Kiss T . Small nucleolar RNAs: an abundant group of noncoding RNAs with diverse cellular functions. Cell 2002; 109: 145–148.

    CAS  Article  PubMed  Google Scholar 

  4. 4

    Kiss T, Filipowicz W . Exonucleolytic processing of small nucleolar RNAs from pre-mRNA introns. Genes Dev 1995; 9: 1411–1424.

    CAS  Article  PubMed  Google Scholar 

  5. 5

    Dieci G, Preti M, Montanini B . Eukaryotic snoRNAs: a paradigm for gene expression flexibility. Genomics 2009; 94: 83–88.

    CAS  Article  PubMed  Google Scholar 

  6. 6

    Decatur WA, Fournier MJ . rRNA modifications and ribosome function. Trends Biochem Sci 2002; 27: 344–351.

    CAS  Article  PubMed  Google Scholar 

  7. 7

    Decatur WA, Fournier MJ . RNA-guided nucleotide modification of ribosomal and other RNAs. J Biol Chem 2003; 278: 695–698.

    CAS  Article  PubMed  Google Scholar 

  8. 8

    Cavaille J, Buiting K, Kiefmann M, Lalande M, Brannan CI, Horsthemke B et alIdentification of brain-specific and imprinted small nucleolar RNA genes exhibiting an unusual genomic organization. Proc Natl Acad Sci USA 2000; 97: 14311–14316.

    CAS  Article  PubMed  Google Scholar 

  9. 9

    Runte M, Huttenhofer A, Gross S, Kiefmann M, Horsthemke B, Buiting K . The IC-SNURF-SNRPN transcript serves as a host for multiple small nucleolar RNA species and as an antisense RNA for UBE3A. Hum Mol Genet 2001; 10: 2687–2700.

    CAS  Article  PubMed  Google Scholar 

  10. 10

    Gallagher RC, Pils B, Albalwi M, Francke U . Evidence for the role of PWCR1/HBII-85 C/D box small nucleolar RNAs in Prader-Willi syndrome. Am J Hum Genet 2002; 71: 669–678.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  11. 11

    Gee HE, Buffa FM, Camps C, Ramachandran A, Leek R, Taylor M et alThe small-nucleolar RNAs commonly used for microRNA normalisation correlate with tumour pathology and prognosis. Br J Cancer 2011; 104: 1168–1177.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  12. 12

    Dong XY, Guo P, Boyd J, Sun X, Li Q, Zhou W et alImplication of snoRNA U50 in human breast cancer. J Genet Genomics 2009; 36: 447–454.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  13. 13

    Dong XY, Rodriguez C, Guo P, Sun X, Talbot JT, Zhou W et alSnoRNA U50 is a candidate tumor-suppressor gene at 6q14.3 with a mutation associated with clinically significant prostate cancer. Hum Mol Genet 2008; 17: 1031–1042.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  14. 14

    Mei YP, Liao JP, Shen JP, Yu L, Liu BL, Liu L et alSmall nucleolar RNA 42 acts as an oncogene in lung tumorigenesis. Oncogene 2011; e-pub ahead of print 10 October 2011.

  15. 15

    Dixon-McIver A, East P, Mein CA, Cazier JB, Molloy G, Chaplin T et alDistinctive patterns of microRNA expression associated with karyotype in acute myeloid leukaemia. PLoS One 2008; 3: e2141.

    Article  PubMed  PubMed Central  Google Scholar 

  16. 16

    Narsinh KH, Sun N, Sanchez-Freire V, Lee AS, Almeida P, Hu S et alSingle cell transcriptional profiling reveals heterogeneity of human induced pluripotent stem cells. J Clin Invest 2011; 121: 1217–1221.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  17. 17

    Marcucci G, Radmacher MD, Maharry K, Mrozek K, Ruppert AS, Paschka P et alMicroRNA expression in cytogenetically normal acute myeloid leukemia. N Engl J Med 2008; 358: 1919–1928.

    CAS  Article  PubMed  Google Scholar 

  18. 18

    Garzon R, Volinia S, Liu CG, Fernandez-Cymering C, Palumbo T, Pichiorri F et alMicroRNA signatures associated with cytogenetics and prognosis in acute myeloid leukemia. Blood 2008; 111: 3183–3189.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  19. 19

    Gaur A, Jewell DA, Liang Y, Ridzon D, Moore JH, Chen C et alCharacterization of microRNA expression levels and their biological correlates in human cancer cell lines. Cancer Res 2007; 67: 2456–2468.

    CAS  Article  Google Scholar 

  20. 20

    Cavaille J, Seitz H, Paulsen M, Ferguson-Smith AC, Bachellerie JP . Identification of tandemly-repeated C/D snoRNA genes at the imprinted human 14q32 domain reminiscent of those at the Prader-Willi/Angelman syndrome region. Hum Mol Genet 2002; 11: 1527–1538.

    CAS  Article  PubMed  Google Scholar 

  21. 21

    Seitz H, Royo H, Bortolin ML, Lin SP, Ferguson-Smith AC, Cavaille J . A large imprinted microRNA gene cluster at the mouse Dlk1-Gtl2 domain. Genome Res 2004; 14: 1741–1748.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  22. 22

    Tierling S, Dalbert S, Schoppenhorst S, Tsai CE, Oliger S, Ferguson-Smith AC et alHigh-resolution map and imprinting analysis of the Gtl2-Dnchc1 domain on mouse chromosome 12. Genomics 2006; 87: 225–235.

    CAS  Article  PubMed  Google Scholar 

  23. 23

    Dulaney AM, Murgatroyd RJ . Use of trans-retinoic acid in the treatment of acute promyelocytic leukemia. Ann Pharmacother 1993; 27: 211–214.

    CAS  Article  PubMed  Google Scholar 

  24. 24

    Sakajiri S, O'Kelly J, Yin D, Miller CW, Hofmann WK, Oshimi K et alDlk1 in normal and abnormal hematopoiesis. Leukemia 2005; 19: 1404–1410.

    CAS  Article  Google Scholar 

  25. 25

    Lubbert M, Miller CW, Crawford L, Koeffler HP . p53 in chronic myelogenous leukemia. Study of mechanisms of differential expression. J Exp Med 1988; 167: 873–886.

    CAS  Article  PubMed  Google Scholar 

  26. 26

    Rui HB, Su JZ . Co-transfection of p16(INK4a) and p53 genes into the K562 cell line inhibits cell proliferation. Haematologica 2002; 7: 136–142.

    Google Scholar 

  27. 27

    Butler AP, Trono D, Coletta LD, Beard R, Fraijo R, Kazianis S et alRegulation of CDKN2A/B and retinoblastoma genes in xiphophorus melanoma. Comp Biochem Physiol C Toxicol Pharmacol 2007; 145: 145–155.

    Article  PubMed  Google Scholar 

  28. 28

    Mizuarai S, Machida T, Kobayashi T, Komatani H, Itadani H, Kotani H . Expression ratio of CCND1 to CDKN2A mRNA predicts RB1 status of cultured cancer cell lines and clinical tumor samples. Mol Cancer 2011; 10: 31.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  29. 29

    Guo SX, Taki T, Ohnishi H, Piao HY, Tabuchi K, Bessho F et alHypermethylation of p16 and p15 genes and RB protein expression in acute leukemia. Leuk Res 2000; 24: 39–46.

    CAS  Article  PubMed  Google Scholar 

  30. 30

    Brameier M, Herwig A, Reinhardt R, Walter L, Gruber J . Human box C/D snoRNAs with miRNA like functions: expanding the range of regulatory RNAs. Nucleic Acids Res 2010; 39: 675–686.

    Article  PubMed  PubMed Central  Google Scholar 

  31. 31

    Ender C, Krek A, Friedlander MR, Beitzinger M, Weinmann L, Chen W et alA human snoRNA with microRNA-like functions. Mol Cell 2008; 32: 519–528.

    CAS  Article  PubMed  Google Scholar 

  32. 32

    Kishore S, Khanna A, Zhang ZY, Hui JY, Balwierz PJ, Stefan M et alThe snoRNA MBII-52 (SNORD 115) is processed into smaller RNAs and regulates alternative splicing. Human Molecular Genetics 2010; 19: 1153–1164.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  33. 33

    Kishore S, Stamm S . The snoRNA HBII-52 regulates alternative splicing of the serotonin receptor 2C. Science 2006; 311: 230–232.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  34. 34

    Mitchell JR, Cheng J, Collins K . A box H/ACA small nucleolar RNA-like domain at the human telomerase RNA 3′ end. Mol Cell Biol 1999; 19: 567–576.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  35. 35

    Ono M, Scott MS, Yamada K, Avolio F, Barton GJ, Lamond AI . Identification of human miRNA precursors that resemble box C/D snoRNAs. Nucleic Acids Res 2011; 39: 3879–3891.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  36. 36

    Martens JH, Brinkman AB, Simmer F, Francoijs KJ, Nebbioso A, Ferrara F et alPML-RARalpha/RXR alters the epigenetic landscape in acute promyelocytic leukemia. Cancer Cell 2010; 17: 173–185.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  37. 37

    Figueroa ME, Lugthart S, Li Y, Erpelinck-Verschueren C, Deng X, Christos PJ et alDNA methylation signatures identify biologically distinct subtypes in acute myeloid leukemia. Cancer Cell 2010; 17: 13–27.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  38. 38

    Saito Y, Liang G, Egger G, Friedman JM, Chuang JC, Coetzee GA et alSpecific activation of microRNA-127 with downregulation of the proto-oncogene BCL6 by chromatin-modifying drugs in human cancer cells. Cancer Cell 2006; 9: 435–443.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  39. 39

    Meng F, Wehbe-Janek H, Henson R, Smith H, Patel T . Epigenetic regulation of microRNA-370 by interleukin-6 in malignant human cholangiocytes. Oncogene 2008; 27: 378–386.

    CAS  Article  Google Scholar 

  40. 40

    Benetatos L, Vartholomatos G, Hatzimichael E . MEG3 imprinted gene contribution in tumorigenesis. Int J Cancer 2011; 129: 773–779.

    CAS  Article  PubMed  Google Scholar 

  41. 41

    Kornblau SM, Xu HJ, Zhang W, Hu SX, Beran M, Smith TL et alLevels of retinoblastoma protein expression in newly diagnosed acute myelogenous leukemia. Blood 1994; 84: 256–261.

    CAS  PubMed  Google Scholar 

  42. 42

    Paggi MG, de Fabritiis P, Bonetto F, Amadio L, Santarelli G, Spadea A et alThe retinoblastoma gene product in acute myeloid leukemia: a possible involvement in promyelocytic leukemia. Cancer Res 1995; 55: 4552–4556.

    CAS  PubMed  Google Scholar 

  43. 43

    Khan MM, Nomura T, Kim H, Kaul SC, Wadhwa R, Zhong S et alPML-RARalpha alleviates the transcriptional repression mediated by tumor suppressor Rb. J Biol Chem 2001; 276: 43491–43494.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We are grateful to the patients who participated in this study and their referring physicians. We thank the Genotoul platform, Toulouse, France (Naïs Prade and Jean-José Maoret). We also thank Valerie Lallemand-Breitenbach, José Ángel Martínez Climent and Pauline Gravelle for their helpful comments on this work. This study was supported by grants from the Institut Universitaire de France, the Association pour la Recherche sur le Cancer (ARC) and the CITTIL program (Cooperación de Investigación Transpirenaica en la Terapia Innovadora de la Leucemia).

Author information

Affiliations

Authors

Corresponding author

Correspondence to P Brousset.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Supplementary Information accompanies the paper on the Leukemia website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Valleron, W., Laprevotte, E., Gautier, EF. et al. Specific small nucleolar RNA expression profiles in acute leukemia. Leukemia 26, 2052–2060 (2012). https://doi.org/10.1038/leu.2012.111

Download citation

Keywords

  • SnoRNA
  • acute leukemia
  • PML-RARalpha
  • DLK1-DIO3

Further reading

Search

Quick links