Lysosomal α-D-mannosidase is an exoglycosidase involved in the ordered degradation of N-linked oligosaccharides. It is ubiquitously expressed, although the main transcript is more abundant in peripheral blood leucocytes. Here we report that α-D-mannosidase enzyme activity is very high in the promyelocytic leukaemia cell lines HL60 and NB4, as compared with other leukaemic cell lines or cells from different human sources. The MAN2B1 transcript level correlates with enzyme activity, indicating a transcriptional up-regulation of the α-D-mannosidase gene. The promoter was then characterized in HEK-293 cells (human embryonic kidney 293 cells) and HL60 cells; regulatory sequences crucial for its activity were determined by reporter gene assay in HEK-293 cells and located in the region −101/−71 with respect to the first ATG codon. Supershift assay demonstrated that Sp1 (specificity protein 1) bound to this sequence both in HEK-293 and HL60 cells. However, 5′-RACE (5′-rapid amplification of cDNA ends) indicated the use of multiple upstream TSSs (transcription start sites) in HL60 with respect to HEK-293 cells and gel shift analysis of the sequence −373/−269 demonstrated a specific binding by NF-κB (nuclear factor κB) transcription factor in HL60 but not in HEK-293 cells. We concluded that despite the α-D-mannosidase promoter showing typical features of housekeeping gene promoters, α-D-mannosidase transcription is specifically regulated in HL60 by NF-κB transcription factor.
Mannosidases are classified on the basis of their optimum pH in three main groups: acid or lysosomal α-D-mannosidases (optimum pH 4.0–4.5) that are relatively stable and activated by zinc ions (Zn2+); the Golgi-associated α-D-mannosidase (optimum pH 5.5–6.0), activated by cobalt ions (Co2+); and the endoplasmic-reticulum-associated α-D-mannosidase (optimum pH 7.0) that are less stable and probably linked to the membrane. On the basis of their substrate specificity they are also classified in two groups: class I mannosidases, which can hydrolyse α1,2-linkages, and class II mannosidases, which can hydrolyse α1,2-, α1,3- and α1,6-linkages [1–3]. Lysosomal α-D-mannosidase (EC 126.96.36.199) is an acidic exoglycosidase that catalyses the hydrolysis of α1,2-, α1,3- and α1,6-mannoside linkages during the ordered degradation of N-linked glycoproteins [4,5]. Deficiency of lysosomal α-D-mannosidase activity results in the lysosomal storage disorder α-mannosidosis [MIM (Mendelian Inheritance in Man) No. 248500], an autosomal-recessive disease described in humans, cattle and cats, characterized by accumulation of partially degraded oligosaccharides in the lysosomes [6–9].
Lysosomes and lysosomal glycohydrolases are known to be involved in cancer processes. Lysosomal enzymes, degrading glycoconjugates of connective tissue, may facilitate infiltration by malignant cells and release them to blood vessels  and may promote increased tumour cell shedding from primary tumours . Besides, there are results suggesting that modifications of cell glycoconjugates may be important in the progression of cancer [12–15]. Abnormal levels of α-D-mannosidase have been previously reported in peripheral cells from patients affected by Alzheimer's disease and in the same study it was demonstrated that this up-regulation correlates with Ras oncogene activation . Lysosomal α-D-mannosidase has also been studied in leukaemic lymphoid cells that were shown to have different α-D-mannosidase activity levels as a function of different immunophenotypes .
The gene encoding the human lysosomal α-D-mannosidase gene maps on chromosome 19cen-q13.1. Lysosomal α-D-mannosidase is expressed in all tissues, although its abundance varies ; the transcript is most abundant in peripheral blood leucocytes and barely detectable in brain. In addition, a minor transcript could also be detected in several tissues, but most prominently in spleen, thymus and leucocytes. The human α-D-mannosidase 5′-flanking region has been preliminarily characterized. No TATA sequences could be identified upstream of the first in-frame ATG codon but the region contains several GC-rich sequences with putative binding sites for several transcription factors [6,19]. In addition, the α-D-mannosidase promoter has been preliminarily analysed also in mouse , cattle  and cats , where it was also reported to contain GC-rich regions, which are again common in housekeeping gene promoters. However, considering the importance of the enzyme in the degradation of N-linked glycoproteins in normal and pathological conditions, we focused on transcription factors involved in α-D-mannosidase promoter regulation and molecular events on the basis of α-D-mannosidase expression.
α-D-Mannosidase enzyme activity was examined in a panel of human cell lines. We observed that very high levels of enzyme activity are peculiar of HL60 and NB4 promyelocytic leukaemia cell lines as compared with other human cells of different origin. Gene expression analysis demonstrated that the increased enzyme activity correlates with high levels of the MAN2B1 transcript, indicating a transcriptional up-regulation of the α-D-mannosidase gene. We then characterized α-D-mannosidase promoter in HEK (human embryonic kidney)-293 cells and HL60, showing that Sp1 (specificity protein 1) regulates α-D-mannosidase expression in both cell lines, while NF-κB (nuclear factor κB) specifically regulates α-D-mannosidase expression in HL60 cells. NF-κB is a key regulator of immune responses that is involved in oncogenic initiation and progression of haematologic malignancies, thus representing the connection between these two events.
MATERIALS AND METHODS
Promyelocytic leukaemia cell lines HL60 and NB4, AML (acute myeloid leukaemia) (derived from erythroleukaemia) cell line KG1, chronic myeloid leukaemia cell line K562 and breast cancer cell line MCF7 (all human) were cultured in RPMI 1640 medium supplemented with 10% (v/v) heat-inactivated foetal calf serum (37°C, 95% humidity, 5% CO2). HEK-293 and HUDE (human dermal fibroblasts) cells were grown in DMEM (Dulbecco's modified Eagle's medium) supplemented with 10% fetal bovine serum (37°C, 95% humidity, 5% CO2). Cell growth was determined by counting cell numbers in a haemocytometer. The viability of the cells was estimated by examining their ability to exclude Trypan Blue (0.1% in 0.9% NaCl).
Preparation of cell lysates and enzyme assays
Cell samples were washed twice with 0.9% NaCl (500 g for 10 min in a bench-top centrifuge) and suspended in 10 mM sodium phosphate (pH 6.0) buffer and 0.1% (v/v) Nonidet P40 detergent. After 1 h incubation on ice, they were harvested, sonicated using a 100 W ultrasonic disintegrator (Virtis) equipped with a microtip at 12 μm wave amplitude (four sonications, 15 s each), then centrifuged at 16000 g for 20 min. All procedures were carried out at 4°C. The supernatants were used as cell lysate. Levels of activity of the glycohydrolases α-D-mannosidase and β-D-hexosaminidase were measured using 3 mM solutions of the specific synthetic fluorogenic substrates 4MU-α-mann (4-methylumbelliferyl α-D-mannoside) and 4MU-GlcNAc (4-methylumbelliferyl β-N-acetylglucosaminide) respectively, dissolved in 0.1 M citrate/0.2 M phosphate buffer, pH 4.5, in 96-well black multiplates (Greiner). At the end of the reaction period, 0.290 ml of 0.4 M glycine/NaOH buffer, pH 10.4 were added. Fluorescence of the liberated 4-methylumbelliferone was measured on a Infinite F200 fluorimeter (Tecan) at 360 nm excitation, 450 nm emission. One unit is the amount of enzyme that hydrolyses 1 μmol of substrate/min at 37°C. The protein content was determined by the Bradford method [20a], using BSA as standard. Specific activity was expressed as enzyme units/mg of protein.
Q-PCR (quantitative PCR)
RNA was extracted from using PureLink Micro-to-Midi Total RNA Purification System (Invitrogen) according to the manufacturer's instructions. Then 1 μg of RNA was reverse-transcribed into cDNA by using random hexamers and SuperScript II Reverse Transcriptase (Invitrogen) according to the manufacturer's protocol. Approx. 10 ng of each cDNA was used as template for estimation of MAN2B1 gene expression by Q-PCR in a Stratagene Mx3000P thermocycler (Agilent Technologies) using SYBR Green technology. Primers were designed using Primer-BLAST software (http://www.ncbi.nlm.nih.gov/tools/primer-blast). Sequences for MAN2B1 amplification were the following: MAN2B1.for 5′-TGTTCGCCAGACCTTCTTCT and MAN2B1.rev 5′-ACCAAGGGTGTCTTCACCAG. Results were analysed using the ΔΔCt method, using ACTB (β-actin) genes as endogenous control. Sequences for ACTB amplification were the following: ACTB.for 5′-AGAAAATCTGGCACCACACC and ACTB.rev 5′-GGGGTGTTGAAGGTCTCAAA. ΔCt was calculated subtracting the average Ct value for β-actin to the average Ct valued for MAN2B1 gene for each cell line. ΔΔCt is the difference between ΔCt for HUDE and ΔCt for each cell line. The reported fold expression, expressed as RQ (relative quantity), was calculated by 2−ΔΔCt.
The MAN2B1 gene sequence was obtained from the NCBI (National Center for Biotechnology Information) under the Accession Number NG_008318, gi:195232772. Putative transcription factor binding sites within the human regulatory region were analysed by MatInspector software algorithms (Genomatix).
5′-RACE (5′-rapid amplification of cDNA ends)
Total RNA was isolated by the TRIzol® method according to the manufacturer's (Invitrogen) protocol and the 5′-RACE was carried out using the 5′-RACE System (Invitrogen). The α-D-mannosidase-specific oligonucleotide RAC1.rev 5′-CGAAGCC-CATCTGCGCAAACAGCG-3′ was used to prime the first-strand cDNA synthesis of 3 μg of total RNA. After the SNAP (soluble N-ethylmaleimide-sensitive fusion protein-attachment protein) column purification of cDNA, the cDNA was tailed with dCTP using terminal deoxynucleotidyl transferase. The subsequent first PCR was performed with the oligonucleotide AAP (abridged anchor primer) and with the nested primer RAC2.rev 5′-CTGTGGGGCATGTCTCGTATCC-3′, and the second PCR was performed with the primer AUAP (abridged universal amplification primer) and the nested primer MAN2B1.rev 5′-GAAGATCTTCCCACTCACCTCGTATCCCC-3′, in order to obtain specific amplification products. The identity of amplified product was confirmed initially by amplification with primers 5′-GGGGGGGGGGGGGGGGGGG-3′/5′-CCAT-GGGCGCCTACGCGCGG-3′ (approx. 110 bp) and 5′-GGG-GGGGGGGGGGGGGGGG-3′/5′-CCCGGCCTTTCCAGGGC-CG-3′ (approx. 150 bp) and then by sequencing.
Construction of luciferase reporter vectors
The promoter region of the human α-D-mannosidase gene was selected by searching the Human Genome Resources available at NCBI (National Center for Biotechnology Information). The nucleotide upstream of the translation start site was numbered −1. A reporter gene construct containing the 5′-flanking region −1378 [including the 5′-UTR (untranslated region)] was obtained by PCR amplification of the human genomic DNA using high-fidelity DNA polymerase Expand Long (Roche Diagnostics). Genomic DNA was prepared from human primary fibroblasts using QIAmp method (Qiagen). The PCR was performed with the reverse primer MAN2B1.rev, mapping +165 bp downstream of the translation initiation codon (5′-GAAGATCTTCCCACTCACCTCGTATCCCC) and the forward primer MAN2B1.for (5′-GGGGTACCCCAGTT-TTTCACTCTGGCTGTC), mapping at −1378 bp with respect to the first translation initiation codon of the α-D-mannosidase gene. PCR amplicon was subcloned into the KpnI and BglII restriction sites of the firefly luciferase expressing reporter gene vector pGL3 (Promega), thus obtaining the pGL3-HMAN construct −1378/+165 bp. A series of 5′ deletions were generated from this constructs with the following forward primers: 5′-GGGGTACCCCTTGAGCCACTGCGCCTGCCC (−654/+165 bp), 5′-GGGGTACCCCGTATAGGACTCTTCA-TCAGATCC (−451/+165 bp), GGGGTACCCCTGCTC-CTACCGGCATTAAG (−310/+165 bp), 5′-GGGGTACCC-CACCCCCGGGATTGCCCAG (−161/+165 bp), 5′-GGGTA-CCCCGGGTCTGGGGGCGGGG (−71/+165 bp). To analyse the reporter activity of (−161/+165 bp) construct we subcloned two fragments with the following forward and reverse primers: 5′-GGGGTACCCACCCCCGGGATTGCCCAG and 5′GAAGA-TCTTGTGAGCCTCCAGACCGCG (−161/−101 bp), 5′-GG-GGTACCCCCCTGGCCCCGCCCCCCAC and 5′-GAAGAT-CTTGTGAGCCTCCAGACCGCG (−101/−53 bp). Mutational modification of α-D-mannosidase promoter sequences was obtained with modified oligonucleotides GC-1.mut.for 5′-GGGGTACCCCGAACCCCACCCCG (−101/+165 GC-1.mut) and GC-2.mut.for 5′-GGGGTACCCCGCCCCCCAAA-CCG (−101/+165 GC-2.mut). Mutagenic exchange was confirmed by sequencing.
Transfection and reporter gene assay
Approx. 3×105 HEK-293 cells were seeded in six-well tissue culture plates, cultured for 48 h and transfected using 4 μl of Lipofectamine™ Reagent (Invitrogen), in accordance with the manufacturer's protocol. Transfections were performed with 900 ng of pGL3 vector reporter constructs and 100 ng of pRL-SV40 (Promega), as transfection efficiency control. After incubation overnight, cells were washed and transfection medium was replaced for 24 h by serum-containing culture medium. Cells were then washed with PBS and harvested in reporter lysis buffer (Promega). After incubation for 30 min, cell extract were centrifuged at 12000 g for 30 min to pellet the cell debris and were subjected to luciferase assays. Quantification of firefly and Renilla luciferase activities was performed with the Dual Luciferase Reporter Assay System (Promega). The relative firefly luciferase activity was calculated by normalizing transfection efficiency.
Preparation of NE (nuclear extract)
Cells (1×106–1×107) were trypsinized and washed twice in ice-cold PBS. Cells were then lysed with 500 μl of lysis buffer [10 mM Hepes, pH 7.9, 1.5 mM MgCl2, 10 mM KCl, 0.5 mM DTT (dithiothreitol), including as protease inhibitors 100 μg/ml aprotinin, 5 μg/ml leupeptin, 1 μg/ml pepstatin and 0.5 mM PMSF] and kept on ice for 15 min. Nuclei were pelleted by centrifugation at 6500 g for 1 min, and the supernatants were discarded. Nuclei were resuspended in 300 μl of extraction buffer [10 mM Hepes, pH 7.9, 1.5 mM MgCl2, 420 mM NaCl, 0.2 mM EDTA, 25% (v/v) glycerol, including as protease inhibitors 100 μg/ml aprotinin, 5 μg/ml leupeptin, 1 μg/ml pepstatin and 0.5 mM PMSF] and kept on ice for 30 min. Debris were pelleted by centrifugation at 12000 g for 10 min at 4°C, and the supernatants were recovered. The protein concentrations were determined by Bio-Rad assay, and the samples were diluted to 1 μg/μl with extraction buffer. The nuclear proteins were stored at −70°C.
EMSA (electrophoretic mobility-shift assay)
α-D-Mannosidase promoter segments tested for binding were synthesized by PCR using oligonucleotides biotinylated at 5′ (MWH Biotech). Competitor segments were obtained by PCR amplification with oligonucleotides amplifying the same or a mutated sequence (entire or partial), but unmodified at 5′. PCR was purified with QIAquick PCR purification kit and quantified on 2% agarose gel. Gel-shift assays were performed using 5–10 fmol of biotinylated segment and 1–2 μg of NE per assay in a final volume of 20 μl. The presence of poly(dI-dC) prevented non-specific protein-DNA binding. Incubation for 20 min at room temperature (22–24°C) preceded electrophoretic separation on a native 4–6% polyacrylamide gel in 0.5 × TBE [Tris/borate/EDTA (1 × TBE = 45 mM Tris/borate and 1 mM EDTA)]. Probes were transferred on nylon membrane (Hybond) and incubated with streptavidin conjugated with horseradish peroxidase and developed with luminol. All reagents were purchased from LightShift Chemiluminescent EMSA Kit (Pierce). For supershift analysis, appropriate amounts of anti-Sp1 antibody or anti-NF-κB (p65) antibody (SantaCruz Biotechnology) were added to the samples and pre-incubated on ice for 20–30 min before biotinylated oligonucleotide PCRs were added, and standard incubation for 20 min at room temperature followed. The promoter fragment (−133/−53 bp) used to characterize the Sp1 binding site was amplified by PCR from genomic DNA with the following primers forward 5′-GCCTCTCTCCCGTGAGCCTC and reverse 5′-CCCCGCCCCCAGACCCGGG. The sub-fragments (−101/−53 bp) and (−133/−101 bp) were amplified with the following primers forward and reverse 5′-CCCTGGC-CCCGCCCCCCAC/5′-CCCCGCCCCCAGACCCGGG and 5′-GCCTCTCTCCCGTGAGCCTC/5′-GGGCGCGGTCTGGAG-GCTCAC respectively. The mutated competitor DNA probe GC-1.mut and GC-2.mut were obtained with modified oligonucleotides compGC-1.mut.for 5′-CCCCGAACCCCACCCCG and compGC-2.mut.for 5′-GGGGTACCCCGCCCCCCAAACCG respectively. The promoter fragments (−373/−269 bp) and (−268/−134 bp) used to characterize upstream sequences were amplified from genomic DNA with the following primers forward and reverse 5′-GACGGACACCCTGGATTCCC/5′-ACGGAATGCCTCTTAATGCCGG and 5′-CTTCATAGCCC-GGAGACGCC/5′-TGTAGCCCTGGGCAATCCCG respect-ively. The sub-fragments used to characterize the (−373/−269 bp) promoters sequences were (−301/−269 bp) amplified with primers forward and reverse 5′-CTTC-TGCTCCTACCGGCATTAAGA/5′-ACGGAATGCCTCTTAA-TGCCGG (−323/−293 bp), carrying a mutation in the underlined C/EBP (CCAAT/enhancer-binding protein) core motif, amplified with primers forward and reverse 5′-CG-GCCCCCTTAGAACCAGCACTTCTGCTCC/5′-GGAGCAG-AAGTGCTGGTTCTAAGGGGGCCG and (−373/−314 bp), carrying a mutation in the underlined NF-κB core motif, amplified with primers forward and reverse 5′-CTTC-TGCTCCTACCGGCATTAAGA/5′-GGGGGCCCCGTTGGG-GTAGCCGGTGTAG.
D-mannosidase enzyme activity and transcript level in leukaemia and control cells
A panel of cell lines was assayed for the ability to hydrolyse the substrate for acidic α-D-mannosidase. Results reported in Table 1 demonstrated that, among the human cell lines tested, HL60 (derived from a patient affected by promyelocytic leukaemia ) displayed a very high level of enzyme activity with respect to HUDE, comparable only with the enzyme activity level detected in NB4, a maturation inducible cell line with a t(15;17) marker also isolated from a human promyelocytic leukaemia . The human AML (evolved from erythroleukaemia ) KG1 cell line also displayed quite high levels of activity, whereas chronic myeloid leukaemia K562, acute T-cell leukaemia Jurkat, and cells not of haemopoietic origin, such as MCF7 (breast cancer) and HEK-293 cells, displayed an enzyme activity level just two or three times higher than that of HUDE. The observed increase was not due to a general up-regulation of lysosomal glycohydrolases, as in the same samples we also determined the enzyme activity of β-D-hexosaminidase (EC 188.8.131.52), a lysosomal enzyme removing terminal β-linked N-acetylglucosamine and N-acetylgalattosamine residues from glycoproteins, glycolipids and glycosaminoglycans . We observed that the highest levels of β-D-hexosaminidase activity did not match with the highest levels of α-D-mannosidase activity (Table 1).
|.||.||Specific activity (m-units/mg of protein) .|
|Sample .||Source .||α-D-Mannosidase .||β-D-Hexosaminidase .|
|K562||Chronic myeloid leukaemia in blast crisis||0.7±0.3||26.2±2.7|
|Jurkat||Acute T-cell leukaemia||0.7±0.1||27.9±3.5|
|HEK-293 cells||Embryonic kidney||1.3±02||28.2±6.0|
|.||.||Specific activity (m-units/mg of protein) .|
|Sample .||Source .||α-D-Mannosidase .||β-D-Hexosaminidase .|
|K562||Chronic myeloid leukaemia in blast crisis||0.7±0.3||26.2±2.7|
|Jurkat||Acute T-cell leukaemia||0.7±0.1||27.9±3.5|
|HEK-293 cells||Embryonic kidney||1.3±02||28.2±6.0|
To investigate the molecular basis underlying the different level of α-D-mannosidase enzyme activity, we analysed MAN2B1 gene expression by real-time Q-PCR (Figure 1). Results showed a good correlation between enzyme activity and transcript level, suggesting an up-regulation of MAN2B1 gene transcription in AML-derived cells, such as HL60, NB4 and KG1. HL60 displayed the highest level of α-D-mannosidase transcript, corresponding to an approx. 20-fold increase with respect to HUDE cells, showing the lowest level of transcript, and for this reason used as reference. Very high level of transcript were also showen by NB4 and quite a high level was shown by by KG1.
Gene expression analysis by Q-PCR
Regulatory sequences within the human α-
To define the DNA regions involved in α-D-mannosidase basal expression, we constructed 5′ deletion mutants of the human α-D-mannosidase gene promoter in the reporter vector pGL3 Basic, then we transfected these constructs into the easily transfectable cell line HEK-293 to quantify luciferase activity (Figure 2A). No significant reporter activity was detected with the reporter construct containing the sequence −71/+165 with respect to the first translation start site of α-D-mannosidase gene, in comparison with luciferase activity obtained with the pGL3 Basic vector alone as control. However, a strong luciferase activity was detected when cells were transfected with the plasmid pGL3-MAN2B1(−161/+165 bp). No significant increase of luciferase activity was observed when longer constructs were transfected. Thus, the 5′ deletion data seemed to locate promoter activity to positions between −71 and −161. To verify this assumption, region −161/+165 was divided in two shorter segments that were also cloned in pGL3 Basic vector. Highest activity was obtained with the construct containing segment −101/−53 that showed 100% of the activity of the segment −161/+165. The segment −161/−101 showed very low activity and so it was confirmed for the adjacent sequence −71/+165 (Figure 2B). It was concluded that the promoter activity of the human MAN2B1 gene is mainly restricted within the region ranging from position −101 to −71, in agreement with previous observations . However, additional sequences for modulation and/or regulation in the −161 to −101 region may need further investigation, as the reporter activity of this construct was significantly higher than control.
Promoter active segments of α-
D-mannosidase gene promoter
Effect of mutations on human MAN2B1 promoter activity
Sequence analysis of the human MAN2B1 promoter using MatInspector revealed in the segment −101/−71 with the strongest promoter activity two overlapping GC boxes, putative binding sites for Sp1 transcription factor. These sequences were examined by mutational analysis in combination with reporter gene assays in HEK-293 cells, to assess their significance in determining promoter activity (Figure 3). Two bases were exchanged within motifs by oligonucleotide-based mutagenesis to provide −101/+165 segments GC-1.mut and GC-2.mut respectively. Both GC-1.mut and GC-2.mut were strongly affected, as promoter activity was decreased to 42 and 60% respectively. To detect functional linkages, double mutants were created with the two overlapping GC boxes. Double mutant GC-1.mut/GC-2.mut showed reporter activity in the range of the mutation of GC-1.mut alone. In summary, results indicate that GC boxes, namely the GC-1 box, are of fundamental importance in determining the activity of α-D-mannosidase promoter in HEK-293 cells, even if there is no evidence of additional cooperation between GC-1 box and GC-2 box, both potentially recognized by Sp1.
Identification of factors relevant for MAN2B1 gene promoter activity
Proteins that bind to promoter active segments −101/−53 in HEK-293 cells and HL60 NEs
To examine the binding capability of proteins present in HEK-293 cells NE, the region −133/−53 was employed in EMSA (Figure 4). When incubated with NE, strong protein binding occurred with segment −133/−53. When the region was divided into two segments, binding was confirmed with the −101/−53 segment but not with the −133/−101 segment, as it was expected from the effect of mutations on human MAN2B1 promoter activity (Figure 4B). Competition with unbiotinylated fragment in which the consensus motifs were mutated (Figure 4C) confirmed that the GC-1 box is preferentially bound, as compared with the GC-2 box; preincubation with oligonucleotides containing the GC-1 mutated Sp1 binding site did not inhibit the formation of complexes (lane 4), whereas preincubation with oligonucleotides containing the GC-2 mutated Sp1 binding site inhibited the formation of complexes (lane 5), as well as the preincubation with unmutated oligonucleotides (lane 3). These results were in agreement with the effect of mutations on human MAN2B1 promoter activity. As analysis of promoter suggested that GC boxes were potential binding sites for Sp1 transcription factor, we carried out supershift analysis in the presence of anti-Sp1 antibodies and complexes binding to the segment −101/−53 containing the Sp1 consensus sequence showed up-shifts in the presence of the anti-Sp1 antibody (Figure 4D), thus confirming Sp1 binding.
Binding of proteins present in the HEK-293 cells and HL60 NE to the α-
D-mannosidase gene promoter segment −136/−53
We then performed the same EMSA assay with HL60 NE, to determine whether this transcription factor was also able to bind to this sequence in these cells. A clear binding with the −101/−53 segment was also detected (Figure 4E). In addition, when we carried out supershift analysis in the presence of anti-Sp1 antibodies, complexes binding to the segment −101/−53 containing the Sp1 consensus sequence showed up-shifts in the presence of the anti-Sp1 antibody (Figure 4F), thus showing that Sp1 binding to this sequence occurred also in HL60 cells. However, the supershift in HEK-293 cells is close to 30% of the signal, whereas in HL60 cells it is less than 5%. This suggested a less important role of Sp1 in promoting α-D-mannosidase gene transcription in HL60 cells.
5′-RACE and proteins binding to the promoter active segment −373/−269 in HEK-293 cells and HL60 NEs
To further investigate the α-D-mannosidase gene expression, we analysed α-D-mannosidase transcription initiation by 5′-RACE. Results showed in Figure 5 demonstrate that transcription initiation takes place at different sites in HL60 with respect to HEK-293 cells. In HEK-293 cells the TSS (transcription start site) was located at −20 position with respect to the first ATG codon, in agreement with previous observations , and so these cells have a shorter 5′-UTR region in comparison with HL60. On the other hand, HL60 apparently used two TSSs with the same frequency that were located at −78 and −111 positions with respect to the first ATG. These results suggested that in HL60 cells MAN2B1 gene initiation of transcription is possibly regulated by additional upstream transcription factors.
5′-RACE analysis of TSSs
We therefore examined protein binding to sequences that are located upstream the −133/−53 region previously investigated. Specifically, regions −268/−134 and −373/−269 were employed in EMSA. The region −268/−134 showed no protein binding either with HEK-293 cells or HL60 NEs (results not shown). Results obtained with −373/−269 are instead reported in Figure 6(A) and show that when the biotinylated probe was incubated with HL60 NE, protein binding occurred within the segment, while when the probe was incubated with HEK-293 cells NE, no binding was detected.
Binding of proteins present in the HEK-293 cells and HL60 NE to the α-
D-mannosidase gene promoter segment −373/−269
We then analysed transcription factors binding to the promoter segment −373/−269 by bioinformatic analysis. Detailed sequence analysis of MAN2B1 gene promoter using MatInspector revealed potential binding sites for NF-κB and C/EBP transcription factors within the region −373/−269 (Figure 6B). Both hits were interesting, as specific mutations in the gene encoding the C/EBP family transcription factor C/EBPα are frequently associated with AML (acute myeloid leukaemia) , while the NF-κB signalling pathway is a key regulator of immune responses .
We therefore employed the promoter active segment −373/−269 in EMSA, performing a competition experiment with the entire promoter active segment −373/−269 and with the overlapping segments −328/−295, including both NF-κB and C/EBP consensus sites but mutated in the C/EBP core motif, and −373/−314, excluding the C/EBP consensus site and mutated in the NF-κB binding motif. Results shown in Figure 6(C) confirmed that the promoter fragment not mutated in the NF-κB core motif but mutated in the C/EBP binding site was still able to compete for protein binding (lanes 5 and 6), whereas the fragment mutated in the NF-κB core binding motif was unable to compete (lanes 7 and 8), indicating that a correct NF-κB binding site was necessary for competition. We therefore performed supershift analysis in the presence of anti-NF-κB antibodies and complexes binding to the segment −373/−269 showed up-shift (Figure 6D), thus confirming that NF-κB transcription factor binds to this promoter segment.
Mannosidases are a complex enzyme system involved in both the biosynthesis and catabolism of N-linked glycoproteins. Among these, one of the more frequently studied is lysosomal α-D-mannosidase, which is highly diffused throughout tissues . Due to the role played by this acidic glycohydrolase in remodelling carbohydrate moieties of glycoproteins in normal and pathological conditions (i.e. cancer progression), we have investigated the molecular basis of α-D-mannosidase overexpression in a leukaemia cell model. Here we provide evidence that human lysosomal α-D-mannosidase expression increases to a different extent in leukaemic cell lines of both myeloid or lymphoid origin, in a solid tumour-derived cell line (MCF7, from breast cancer) or in an embryo-derived cell line (HEK-293 cells), when compared with HUDE. Even if an up-regulation of a lysosomal glycohydrolase activity in cancerous cells is not a surprising result because similar observations have been previously reported [27,28], the most significant finding is that α-D-mannosidase shows about a 70-fold increase in HL60 and about a 35-fold increase in NB4 promyelocytic leukaemia cell lines, with respect to dermal fibroblasts. Analysis of gene expression demonstrated that the increased level of enzyme activity correlates very well with transcriptional up-regulation. This indicates that high levels of α-D-mannosidase transcript are on the basis of high levels of enzyme activity observed in AML cells, and particularly in HL60 and NB4 promyelocytic cells. Interestingly, HL60 are at the myeloblast-promyelocyte (AML-M2) stage of development and NB4 cells are at the promyelocyte (AML-M3) stage of development , while KG1 cell line, also displaying a remarkable increase in enzyme activity and transcript level, is at the early stage of myeloid differentiation . Of consequence, very high levels of α-D-mannosidase enzyme activity and transcript appear to be a characteristic of cell lines from AML subtypes with blasts at early stage of differentiation.
Gene regulatory mechanisms underlying α-D-mannosidase expression were investigated. Bioinformatic analysis confirmed that α-D-mannosidase promoter has typical features of housekeeping gene promoters, such as the absence of a TATA-box and the presence of GC-boxes. Transfection of deletion constructs in HEK-293 cells showed that the promoter is regulated by a short sequence located at position −101/−71 with respect to the first ATG. Besides, the significant activity of the −161/−101 segment and 4-fold signal above the control of the −101/+165 construct suggest additional sequences for modulation/regulation in the −161/−101 region that should be further investigated. Similar findings have been made for the promoters of other lysosomal enzymes, such as human β-hexosaminidase α- and β-subunit genes  and human N-acetylgalactosamine-6-sulfatase gene , where promoter activity was obtained with as little as less than 100 bp of 5′-flanking region.
The occurrence of two overlapping GC-boxes with Sp1 consensus sequences in the regulatory region suggested that this element could be possibly the main regulatory factor controlling expression of human α-D-mannosidase. The major role of Sp1 in the regulation of α-D-mannosidase expression was confirmed by mutational analysis, gel shift and supershift assay both in HEK-293 cells and HL60. Sp1 is a ubiquitous activator of numerous TATA-containing and TATA-less promoters within the human genome . It is a well-characterized sequence-specific transcription factors, and it has numerous functions in the transcription of many cellular and viral genes harbouring GC boxes in their promoters [34,35]. Nevertheless, it was evident that other transcription factors have to be responsible of α-D-mannosidase higher level of transcript in HL60 with respect to HEK-293 cells.
5′-RACE analysis showed that HL60 cells have also an α-D-mannosidase transcript with a longer 5′-UTR, due to the preferential use of TSSs that are more distant from the first ATG, suggesting that a longer transcript may be either more transcribed or more stable, leading to the higher level of transcript observed by Q-PCR. Analysis of protein binding showed that a specific binding can be detected in the region −373/−269 with HL60 but not with HEK-293 cells NEs, and bioinformatic analysis suggested that NF-κB and C/EBP sequences within the −373/−269 region may be regulatory elements controlling expression of human α-D-mannosidase in HL60 cells. The C/EBP family of transcription factors regulates the expression of numerous myeloid genes  and C/EBPα mutations are frequently associated with AML . NF-κB has been shown to play a pivotal role in haemopoiesis  and is involved in oncogenic initiation and progression of haematologic malignancies [38,39]. Gel shift and supershift analysis clearly demonstrated that this sequence is bound by NF-κB transcription factor. Actually NF-κB is a collection of dimers from transcription factors playing a central role in immune responses, development, cell growth and survival. Deregulated functions of NF-κB contribute to the development of cancer [40,41] and immune-related disorders . Of interest, it was previously reported that NF-κB is constitutively activated in human AML blasts, as demonstrated by its increasing ability to bind specific consensus sequences [43–45]. Our results indicate that the activation of NF-κB leads to an increased expression of lysosomal α-D-mannosidase.
In general, the term ‘housekeeping’ implies permanent expression with little regulation, but highly tissue specific regulation has been previously shown to occur with this type of promoter . Our results show that despite α-D-mannosidase is also considered an housekeeping gene, its expression is differently regulated in HL60 cells, as compared with a control cell line, such as HEK-293 cells. Aberrant glycosylation is a hallmark of many pathologies and in particular it occurs in essentially all types of experimental and human cancers, and many glycosyl epitopes constitute tumour-associated antigens . α-D-Mannosidase is responsible of the degradation of N-linked oligosaccharides, so the presence of high levels of α-D-mannosidase in AML cells, namely HL60 and NB4, could be related to an abnormal degradation of glycosylated proteins and/or alteration of cell adhesion mediated by cell glycoconjugates that should be further investigated. Furthermore, overall results suggest further investigations on lysosomal glycohydrolases as target genes of transcription factors involved in oncogenesis such as NF-κB.
acute myeloid leukaemia
electrophoretic mobility-shift assay
human embryonic kidney
human dermal fibroblast(s)
National Center for Biotechnology Information
nuclear factor κB
5′-rapid amplification of cDNA ends
specificity protein 1
transcription start site
Lorena Urbanelli designed research, performed experiments, analysed the data and co-wrote the paper; Alessandro Magini, Luisa Ercolani, Francesco Trivelli, Alice Polchi and Brunella Tancini performed experiments; Carla Emiliani supervised experiments and co-wrote the paper.
This work was supported by the Fondazione Cassa di Risparmio di Perugia [grant number 2008.021.375 (to C.E.)].
Present address: Department of Biochemistry, Biology and Genetics, Polytechnic University of Marche, Ancona, Italy.