For the siRNA knockdown experiments, HeLa cells were cultured in Dulbecco’s modified Eagle’s medium (Invitrogen, Karlsruhe, Germany). Medium was supplemented with 10% fetal bovine serum (FBS), 1% glutamine and 1% antibiotic solution (penicillin and streptomycin; all from Gibco-BRL, Karlsruhe, Germany). Transfection of the cells was carried out in the absence of any antibiotics.

A negative control non-silencing siRNA with no known homology to mammalian genes (si-nons-Rho: 5′-UUC UCC GAA CGU GUC ACG Udtdt-3′; 5-prime labeled with rhodamine); and two different siRNAs specific to GliPR (si-GliPR-1, 5′-GGU GAA ACC AAC AGC CAG Udtdt-3′ and si-GliPR-2, 5′-GGA CUA UGA CUU CAA GAC Udtdt-3′) were used. All siRNAs were synthesized by Ambion (Darmstadt, Germany) and were purchased as annealed RNA-duplexes. Twenty-four hours before transfection, HeLa cells were plated in 24-well plates (Corning, Kaiserslautern, Germany) at 5×10⁴ cells/well in Dulbecco’s minimal essential medium containing 10% FBS with no antibiotics. Transfections were performed with Lipofectamine^® 2000 transfection reagent (Invitrogen) with siRNA at a final concentration of 20 nM according to the manufacturer’s recommendations. After incubating for 6 h, the lipid/siRNA complexes were removed and replaced with fresh medium. For further analysis, cells were removed from the culture dish by trypsinization with 0.25% trypsin/0.02% EDTA in PBS (Cambrex, Verviers, Belgium) at different time points after transfection. Transfection efficiency was analyzed by flow cytometry 24 h after transfection. Data were acquired and analyzed on a FACScan instrument with Cell Quest software (Becton-Dickinson, Heidelberg, Germany). Effects on cellular viability after siRNA treatment were measured using the cell proliferation reagent WST-1 (Roche, Penzberg, Germany) according to the manufacturer’s instructions.

RNA was extracted using the RNeasy Mini kit (Invitrogen) including treatment with RNase-free DNase I (Qiagen). Synthesis of cDNA was carried out using random hexamer primers and Superscript II RNaseH-reverse transcriptase (Invitrogen) according to the manufacturer’s specifications.

Real-time PCR was performed in duplicate reactions employing ABI PRISM 7700 (Applied Biosystems, Darmstadt, Germany) under standard conditions (50°C for 2 min, 95°C for 10 min and 40 cycles at 95°C for 15 sec and 60°C for 1 min). The 25 μl PCR included 2.5 μl cDNA, 1X TaqMan^® Universal PCR Master Mix (Applied Biosystems), 0.2 μM TaqMan^® probe, 0.2 μM forward primer and 0.2 μM reverse primer. Primers and probes were designed using Primer Express v.1.0 software (Applied Biosystems) and were synthesized by Thermo Hybaid (Ulm, Germany). In order to quantify GliPR1 and GAPDH cDNA, the following primers and probes were used: GliPR (sense, 5′-TGC CAG ACA AAG CAT GCG T-3′; antisense, 5′-GCT GTG TGT GAA TAA TTG GAG ACA A-3′; probe, 5′-FAM-TCA CAC TTG CTA CAA TAG CCT GGA TGG TTT C-3′-TAMRA) and GAPDH (sense, 5′-GAA GGT GAA GGT CGG AGT C-3′; antisense, 5′-GAA GAT GGT GAT GGG ATT TC-3′; probe, 5′-FAM-CAA GCT TCC CGT TCT CAG CC-3′-TAMRA). The probes were labeled with FAM at the 5′ end and TAMRA at the 3′ end. Copy numbers of the respective transcripts were calculated by plasmid standard curves, normalized by GAPDH housekeeping gene transcripts. Standard curves were obtained after amplification of log step dilutions between 10 to 10⁶ copy numbers of purified plasmids carrying the amplicons of human GliPR1 or GAPDH (12), respectively. The plasmid standard for the quantification of GliPR1 was prepared by inserting a PCR-generated fragment (sense, 5′-GGA TCC ATG CGT GTC ACA CTT GCT ACA ATA GC-3′ and antisense, 5′-GTC GAC TTA GTC CAA AAG AAC TAA ATT AGG GTA CTT GAG C-3′) into pCR2.1 (Invitrogen), which was amplified using HeLa cDNA as a template.

The microarray analysis of the effect of GliPR1 suppression by siRNA was performed using the HG-U133 Plus 2.0 microarray of Affymetrix (Santa Clara, CA, USA) according to the manufacturer’s instructions (GeneChip^® Expression Analysis Manual). This chip contains 47,000 transcripts which represent 39,000 annotated genes. The data analysis was carried out according to established standards for Affymetric microarrays using GeneChip^® Operating software (GCOS; Affymetrix) and GeneSpring (Agilent Technologies).

Since GliPR1 has an N-terminal signal peptide that is important for its subcellular localization, EGFP was fused to the C-terminus of GliPR1. The plasmid p53-EGFP (Clontech, Heidelberg, Germany) was cut by restriction enzyme digestion with BamHI and SacII, in order to remove the p53 gene. The resulting gap was filled with the ORF of GliPR1 without the stop codon that had been amplified from HeLa mRNA by primers containing flanking BamHI and SacII restriction sites. The correct cDNA sequence of the GliPR1 insert was verified by sequencing.

Intracellular localization of GliPR1 was determined by confocal fluorescence microscopy. Therefore 1×10⁴ to 5×10⁴ adherently growing HeLa cells were seeded onto a slide overnight (diagnostic slide with adhesive epoxy layer; Roth, Karlsruhe, Germany) in culture dishes. Cells were gently washed in PBS (37°C), fixed for 10 min in 3% (v/v) formaldehyde in PBS at room temperature (RT) and permeabilized for 10 min in 0.1% (v/v) Triton X-100 in PBS. Prior to incubating with the antibodies, cells were blocked with Image-iT^® FX Signal Enhancer reagent (Molecular Probes, Eugene, OR, USA) for 30 min at RT and after washing were incubated with the primary antibody [diluted in 3% (w/v) BSA in PBS] for 90 min at RT. After washing in PBS, cells were incubated with the respective fluorophore-coupled secondary antibody for 90 min at RT. Then nuclei were stained with 4′,6-diamidino-2-phenylindol (DAPI; Molecular Probes) in dilution of 1:300 PBS for 5 min at RT followed by the overlaying of cells with cover medium (ProLong Mounting Medium; Molecular Probes). Slides were evaluated with the Axioplan II imaging microscope and AxioVision software (Zeiss, Göttingen, Germany) and displayed by Adobe Photoshop 6.01 software (Adobe Systems GmbH, Munich, Germany). Mouse IgG2a anti-protein disulfide isomerase (PDI) antibody (Acris Antibodies GmbH, Herford, Germany) was used as the primary antibody at a dilution of 1:200, and goat anti-mouse IgG (1:500) labelled with Alexa 594 (Molecular Probes, OR, USA; Invitrogen, Darmstadt, Germany) was employed as the secondary antibody for staining of the ER.

The LGA is a public platform which provides published gene expression datasets in the field of leukemia that can be used for further analyses in regard to specific genes of interest (13). The dataset published by Haferlach et al(14) (n=2,096 samples) was evaluated for differential GliPR1 expression among different leukemias. The same dataset was employed to compare GliPR1 expression between acute myeloid leukemia (AML) patients with different karyotypes (translocations) as well as between acute lymphoblastic leukemia (ALL) patients with different karyotypes. The dataset published by Verhaak et al(15) (n=461) was utilized to evaluate GliPR1 expression between the FAB classes of AML. Both datasets were generated on the Affymetrix HG-U133 Plus 2 platform that contains five probe sets for measuring the expression of GliPR1. Analyses were based on the mean of these probe sets. The Kruskal-Wallis test was applied to test for differential expression across several groups. The Wilcoxon test was applied to test for differences between a specific leukemia subgroup and the reference group (16).

HeLa cells were transfected with siRNAs specific for GliPR1 or a non-silencing siRNA, which was 5-prime-labeled with rhodamine. Flow cytometric analysis of cells transfected with the non-silencing siRNA 24 h post transfection revealed transfection efficiencies on average of ~90%. Forty-eight hours after transfection, the relative levels of GliPR1 mRNA transcripts were decreased by more than 90%, compared to HeLa cells transfected with the non-silencing control siRNA as measured by quantitative PCR (Fig. 1A). Viability and the proliferation rate of HeLa cells transfected with siRNAs against GliPR1 or with the non-silencing siRNA remained unchanged as determined by the WST-1 cell proliferation assay (Fig. 1B).

In conclusion, GliPR1-directed siRNAs reduced the expression of GliPR1 effectively in HeLa cells without affecting cell viability in general.

In order to examine the effect of GliPR1 knockdown on cellular gene expression, a microarray analysis was performed to identify cellular target genes of GliPR1.

The microarray analysis showed a similar suppression of GliPR1 expression to 20% confirming the results of the quantitative PCR.

The knockdown of GliPR1 revealed 262 differentially expressed genes including 40 induced genes and 222 suppressed genes (≥3-fold increase or ≤0.3-fold decrease) as listed in Table I.

GliPR1-associated differentially expressed genes were grouped according to the Gene Ontology (GO) Databank categories. The GO Databank is structured hierarchically according to 3 different criteria: subcellular localization, type of biological process and molecular function. Each of the differentially regulated genes was annotated to one or more specific GO categories. Subsequently, it was evaluated, which GO categories showed an overrepresentation of the differentially expressed genes compared to a theoretical random annotation. Table II lists the relevant overrepresented GO categories with their differentially expressed genes. All of the genes in these overrepresented GO categories were found repressed except for one gene (MTM1) in the category dephosphorylation. The overrepresented GO categories included G protein signaling pathways (7 downregulated genes), regulation of cyclin-dependent protein kinase activity (4 downregulated genes), ER to Golgi vesicle-mediated transport (3 downregulated genes), axon guidance (10 downregulated genes) and dephosphorylation (1 upregulated gene; 6 downregulated genes). The factors by which these genes were downregulated or upregulated are included in Table I. The genes with a more marked downregulation by a factor ≤1/5 were APC, EGFR, DNAJC6, GNB5, APBB2 and phosphodiesterase 1A (PDE1A).

In order to ascertain a possible function of GliPR1, its subcellular localization was studied. Confocal immune fluorescence microscopy was employed to determine the subcellular localization of GliPR1. Therefore, GliPR1 was expressed as an EGFP fusion protein emitting green fluorescence (Fig. 2). The nuclei were contrasted with DAPI dye showing no nuclear localization of GliPR1. The ER was stained with a specific antibody against PDI and subsequently with a secondary antibody labeled with fluorophore Alexa-594. Exogenous expression of GliPR1 prompted apoptosis in ~25–35% of HeLa cells after 24 h according to Annexin V testing (data not shown). The confocal microscopy of cells without apoptotic morphology showed a diffuse EGFP staining in the cytosole with intensified staining around the nuclei (Fig. 2A). The PDI staining (marker of ER) in red showed a similar pattern. When the 3 fluorophores were analyzed by confocal microscopy simultaneously, the red and green fluorescence became mostly yellow indicating a co-localization of GliPR1 with the ER. Another cell fraction revealed a vesicular pattern of green staining that was not superimposable with the red fluorescence (Fig. 2B). This pattern suggests a distribution of GliPR1 in cytoplasmic vesicles.

We questioned whether GliPR1 expression is associated with specific leukemia types. The analysis was performed in silico. A dataset based on a microarray analysis of bone marrow samples of 2,096 hematological patients and controls in total (14) was employed. Bone marrow samples (n=542) of patients with AML prior to therapy revealed a significantly increased GliPR1 expression level compared to 73 subjects without leukemia and healthy bone marrow (P<0.001) (Fig. 3A). In contrast, 134 bone marrow samples of ALL at diagnosis prior to therapy had a significantly decreased level of GliPR1 expression compared with the control group (P<0.001). The subsets of patients with chronic lymphocytic leukemia (CLL) and chronic myeloid leukemia (CML) showed a slight decrease in GliPR1 expression (fold-changes, >0.7), although formally significantly different from the control group (CLL, P=0.008; CML, P=0.001). GliPR1 expression in the subset of patients with myelodysplastic syndrome (MDS) was not significantly different from the normal controls (P=0.136).

Furthermore, we investigated whether GliPR1 expression levels were associated with specific subgroups of AML. The dataset of Verhaak et al(15) revealed clear GliPR1 expression differences between different FAB types (P<0.001) (Fig. 3B). FAB types M4, M4E and M5 were characterized by high expression values compared to the other FAB groups. We employed the Haferlach et al(14) dataset again to study the effect of AML-specific chromosomal abnormalities on GliPR1 expression. Particularly, AML patients with translocation t(15;17) or inversion inv(16) showed an increased expression compared to the non-leukemia controls (P<0.001) (Fig. 3C), which is in line with the findings from the Verhaak et al(15) dataset.

Comparison of ALL samples (Fig. 3D) that had abnormal karyotypes with non-leukemia controls using the Haferlach et al(14) dataset revealed a significant decrease in GliPR1 expression in ALL samples with translocation t(1;19) (P<0.001) and in ALL samples with translocation t(12;21) (P<0.001). Hyperdiploid ALL samples did not show altered GliPR1 expression (14–16).