Affymetrix Array Analysis Research Articles

Ion Transporters Involved in Acidification of the Resorption Lacuna in Osteoclasts

Osteoclasts possess a large amount of ion transporters, which participate in bone resorption; of these, the vacuolar-adenosine trisphosphatase (V-ATPase) and the chloride-proton antiporter ClC-7 acidify the resorption lacuna. However, whether other ion transporters participate in this process is currently not well understood. We used a battery of ion channel inhibitors, human osteoclasts, and their subcellular compartments to perform an unbiased analysis of the importance of the different ion transporters for acidification of the resorption lacuna in osteoclasts. CD14(+) monocytes from human peripheral blood were isolated, and mature osteoclasts were generated using RANKL and M-CSF. The human osteoclasts were (1) used for acridine orange assays for evaluation of lysosomal acidification, (2) used for bone resorption assays, (3) used for generation of osteoclasts membranes for acid influx experiments, or (4) lysed in trizol for mRNA isolation for Affymetrix array analysis. Inhibitors targeted toward most of the ion transporters showed low potency in the acidification-based assays, although some inhibitors, such as carbonic anhydrase II and the sodium-hydrogen exchanger (NHE) inhibitors, reduced resorption potently. In contrast, inhibitors targeted at V-ATPase and ClC-7 potently inhibited both acidification and resorption, as expected. We here show evidence that acidification of the resorption lacuna is mainly mediated by V-ATPase and ClC-7. Furthermore, a group of other ion transporters, including carbonic anhydrase II, the NHEs, and potassium-chloride cotransporters, are all involved in resorption but do not seem to directly be involved in acidification of the lysosomes.

Effect of bacterial vaginosis, Lactobacillus and Premarin estrogen replacement therapy on vaginal gene expression changes

The aim of the study was to investigate gene expression profiles of post-menopausal women receiving Premarin estrogen replacement therapy (ERT), compared to controls, and to examine any correlations between the bacterial vaginosis (BV) status of the subjects. Based upon an expected finding of a 50–60% difference between gene expression of host antimicrobials with alpha = 0.05 (2-sided), beta = 0.20 the calculation of 7 subjects per group, led to a sample size of 10 subjects receiving Premarin estrogen replacement therapy and 10 healthy, age-matched controls. Vaginal samples were collected at a single timepoint and processed for RNA recovery and Affymetrix array analysis, as well as Nugent scoring and denaturing gradient gel electrophoresis to identify bacteria. Lactobacillus iners was the most commonly detected species in the normal flora and this was confirmed with L. iners-specific PCR method. Vaginal swabs from 6 Premarin and 8 control vaginal samples provided a non-invasive means to analyze human gene expression. There was no significant up-regulation of cancer-associated gene expression in subject receiving Premarin ERT, but some evidence that the potentially protective innate immunity was reduced in patients with BV. Of those with a normal flora, there was a 2-fold down-regulation of carcinoma associated forkhead box A1 gene expression. BV was associated with 7-fold down-regulation of host antimicrobial colony stimulating factor, −9.83-fold for IL-1α, −8.33 for IL-1β and −3.63 for IL-6. This is the first study to use gene arrays to correlate changes in host expression response to estrogen replacement therapy and BV.

CONCOMITANT ACTIVATION OF KRAS AND EGFR SIGNALING IN MURINE PANCREATIC CARCINOGENESIS

Introduction: Kras and Egfr signaling pathways play key roles in pancreatic carcinogenesis. However, their respective role in initiation and progression of premalignant lesions is not completely known. While conditional pancreas-specific activation of KrasG12D in mice leads to PanIN progression and development of pancreatic ductal adenocarcinoma (PDAC), Tgfa overexpression induces acino-ductal metaplasia and PDAC at late stages. To analyze the effect of concomitant Kras and Egfr signaling, we crossed Tgfa overexpressing mice with p48-Cre;KrasG12D mice. Results: KrasG12D;Ela-Tgfa mice show an accelerated PanIN progression leading to metastasizing PDAC with a significantly reduced survival. Other key features include the induction of acinoductal metaplasia and development of intraductal tumors with striking resemblance to human IPMNs. In order to elucidate the signaling pathways involved in the formation of these tumors, we performed microarray analysis, western blots and Ras activity assays in the early carcinogenic process. Analysis of Ras activity 7 days postnatally showed a highly increased Ras activity in KrasG12D; Ela-Tgfa pancreata, compared to KrasG12D and Ela- Tgfa pancreata and a Ras signature in KrasG12D; Ela-Tgfa mice by Affymetrix array analysis. Western blot analysis revealed no increase in phosphorylated ERK1/2 and AKT but upregulation of phosphorylated STAT3 only in KrasG12D; Ela-Tgfa pancreata. Conclusions: Concomitant Kras and Egfr signaling leads to the induction of PanIN and IPMN preneoplastic lesions and acinoductal metaplasia. Egfr-signaling may increase Ras proteins and Ras signaling beyond the level of mutant KrasG12D alone possibly due to upregulation of other Ras proteins or Ras effectors. The formation of IPMN-like tumors in the pancreata of KrasG12D; Ela-Tgfa mice may involve STAT3 signaling.

Inhibition of Epidermal Growth Factor Receptor Signaling Elevates 15-Hydroxyprostaglandin Dehydrogenase in Non–Small-Cell Lung Cancer

Evidence indicates that the induction of cyclooxygenase-2 (COX-2) and high prostaglandin E2 (PGE2) levels contribute to the pathogenesis of non-small-cell lung cancer (NSCLC). In addition to overproduction by COX-2, PGE2 concentrations also depend upon the levels of the PGE2 catabolic enzyme 15-hydroxyprostaglandin dehydrogenase (15-PGDH). We find a dramatic down-regulation of PGDH protein in NSCLC cell lines and in resected human tumors when compared with matched normal lung. Affymetrix array analysis of 10 normal lung tissue samples and 49 resected lung tumors revealed a much lower expression of PGDH transcripts in all NSCLC histologic groups. In addition, treatment with the epidermal growth factor receptor tyrosine kinase inhibitor (EGFR TKI) erlotinib increased the expression of 15-PGDH in a subset of NSCLC cell lines. This effect may be due in part to an inhibition of the extracellular signal-regulated kinase (ERK) pathway as treatment with mitogen-activated protein kinase kinase (MEK) inhibitor U0126 mimics the erlotinib results. We show by quantitative reverse transcription-PCR that the transcript levels of ZEB1 and Slug transcriptional repressors are dramatically reduced in a responsive cell line upon EGFR and MEK/ERK inhibition. In addition, the Slug protein, but not ZEB1, binds to the PGDH promoter and represses transcription. As these repressors function by recruiting histone deacetylases to promoters, it is likely that PGDH is repressed by an epigenetic mechanism involving histone deacetylation, resulting in increased PGE2 activity in tumors. This effect is reversible in a subset of NSCLC upon treatment with an EGFR TKI.

Engineered Expression of JAK2V617F Drastically Changes the Response of Progenitor Cells to Erythropoietin and Partially Mimics the Polycythemia Expression Signature.

Polycythemia vera (PV) is a myeloproliferative disorder, consisting of an abnormal increase of red blood cells. Over 90% of PV patients harbor the JAK2V617F mutation. To determine the distinct role that JAK2V617F plays in the pathogenesis of PV, we transduced human bone marrow CD34+ cells with either wild-type (WT) or V617F (MT) JAK2 constructs, incubated them in cytokine mixtures designed to promote full myeloid outgrowth (+EPO) or the same mixture lacking erythropoietin (EPO) (−EPO). After 5 days of growth, flow cytometry showed MT cells had a 3-fold increase in erythroid differentiation in the presence of EPO compared to WT cells and demonstrated EPO-independent erythroid development. To ascertain whether MT JAK2 yielded an abnormal EPO response, Affymetrix array analysis was performed on WT and MT CD34+ cells +/−EPO. When we compared WT-EPO cells to MT-EPO cells, 83 genes were significantly different in expression by 1.5x or more. Next we compared WT+EPO to WT-EPO (710 genes), MT+EPO to MT-EPO (1247 genes) and contrasted these groups. 332 genes found in the JAK2 WT response were missing in EPO-treated JAK2V617F cells; 869 genes regulated in JAK2V617F were absent in the WT dataset. A set of 378 genes, part of the normal EPO response, were retained in V617F signaling, but 51 of these were expressed at significantly different levels in the WT and MT conditions and were therefore added to the abnormal response to EPO, now 920 genes. To determine which experimental conditions best modeled PV we tested which set had the best ability to identify PV patients among a group of 9 PV and 8 control samples which we had subjected to Affymetrix expression profiling. The predictor genes were selected from (a) the 83 genes in regulated by MT JAK2 in the absence of EPO, (b) the 920 genes that were aberrantly regulted MT cells were treated with EPO, (c) and the 332 representing the normal EPO response missing in V617F EPO response. The best predictors were 12 genes identified from the 920 gene set that correctly predicted all 17 samples and performed perfectly on cross-validation at a p-value of 0.05. Genes selected from the other groups performed poorly. These analyses indicate that the abnormal EPO response in JAK2V617F cells is more indicative of PV than the normal EPO response and that JAK2 mutation is a gain of function mutation at the genetic level, with a transcriptional readout exceeding that of normal EPO signaling. We next compared the V617F EPO response to the expression profile of PV CD34+ cells and identified 210 in common out of 985 on the PV list (16%) showing a modest overlap. The largest ontological category in this set consisted of 40 genes and represents genes involved in cell growth and maintenance. Two additional well-represented categories, nucleic acid metabolism (26 genes) and protein metabolism (22 genes), are also involved in basic cellular functioning, indicating a profound state of dysfunction in PV cells harboring the V617F mutation. While myelopoiesis was similar between WT and MT cells, gene ontology analysis showed a deficit in immune response genes suggesting JAK2V617F may cause qualitative defects in myelopoiesis. Examination of these genes will illuminate alternative pathways, aberrantly regulated by V617FJAK2, involved in PV pathogenesis.

C-Myb Is an Essential Downstream Target for Homeobox Mediated Transformation of Hematopoietic Cells.

Overexpression of Hox genes is frequently associated with leukemogenic transformation of hematopoietic cells. In particular the abdominal-type HoxA genes together with their homeodomain containing dimerization partners of the TALE (three amino acid loop extension) family are well documented oncogenes in various hematologic malignancies. Paradigmatic examples are Hoxa9 and Meis1 that are coactivated by retroviral insertions in the BXH2 mouse leukemia model. In addition Hoxa9 is a fusion partner of NUP98 in acute myeloid leukemias with a t(7;11) and Meis1 as well as Hoxa9 are critical targets of MLL (mixed lineage leukemia) fusion proteins. Despite their important role in leukemogenesis little is known how they exert their transforming function. Here we used a novel system of transient transformation by an inducible oncoprotein to generate populations of primary hematopoietic cells that overexpress Hoxa9 in combination with Meis1. The RNA inventory of these cells was compared to control cells by Affymetrix array analysis. Interestingly the RNA of c-Myb (myeloblastosis leukemia virus oncogene) and of a known c-myb downstream target (Gstm1) was amongst the top ten genes upregulated by Hoxa9/Meis1. c-Myb overexpression was verified by Northern blot and quantitative RT-PCR. c-Myb levels also responded to the introduction of MLL-ENL, an oncoprotein that upregulates Hoxa9 and Meis1. Small interfering RNAs directed against c-Myb suppressed transformation by MLL-ENL but had no effect on transformation of hematopoietic precursors by E2A-HLF an oncogene that does not work by increasing Hoxa9/Meis1 levels. The anti c-Myb siRNA effect was abrogated by coexpression of a c-Myb derivative with a mutated siRNA target site. The coexpression of a dominant negative c-Myb mutant had a similar but weaker effect on MLL-ENL mediated transformation. Ectopic expression of c-Myb in hematopoietic precursor cells induced a mild differentiation block. However, c-Myb alone was not able to give a positive readout in a serial replating assay under conditions where the combined overexpression of Hoxa9 together with Meis1 was clearly transforming. In summary the results suggest that c-Myb is essential but not sufficient for the oncogenic activity of Hoxa9/Meis1.

Gene Expression Profiles of the Clinically Significant "Late" Stage Expanded Cytotoxic T Cells in Myeloma.

Different stages of differentiation of CD8 T cells based on the expression of costimulatory receptors - CD28 and CD27 has been previously described (Nature Medicine 8:379, 2002). Specifically, CD8 T cells that are CD28−CD27− are designated as the “Late” stage, while those that are CD27+ CD28− belong to the “Intermediate” stage. We have previously demonstrated that in a total of 39 total patients with multiple myeloma that had expanded T cell receptor Vβ (TCRVβ) clones, 78% of patients had clonally expanded CD8+CD57+TCRVβ+ T cells of the “Late” subset while the remaining 22% were of the “Intermediate” subset. It was also demonstrated that patients whose expanded T cells expressed a “Late” T cell phenotype had a significantly improved survival over those with an “Intermediate” T cell phenotype (Blood 102:927a, 2003). The survival of patients with an “Intermediate” phenotype was not significantly better than those who had no detectable T cell clones. It was suggested that the expanded CD8+CD57+ cells clones with late memory/effector phenotype are associated with good prognosis in patients with myeloma. Hence gene expression profiles of these clinically significant late stage T cells will be very useful for the understanding of the anti-tumor mechanisms of these cells. Our laboratory has previously demonstrated that only the CD57+ CD8+ cells in patients with myeloma are monoclonal or biclonal, while the CD57− counterparts are polyclonal. Based on this finding, we performed an Affymetrix array analysis on these “Late” expanded T cell clones and “Intermediate” T cell clones from each of 2 patients. Eventually we aim to select for individual TCRVβ clone rather than from total CD8+CD57+. We have optimised the procedure for RNA extraction of small number of sorted cells (n=1000), RNA amplification and subsequent labelling for Affymetrix array analysis using U133Plus 2.0 GeneChips. Patient's peripheral blood mononuclear cells were sorted using 6-color flow sorting with the phenotypes of DAPI−CD3+CD4−CD57+CD28−CD27+ (Intermediate) and DAPI−CD3+CD4−CD57+CD28−CD27− (Late). The RNA extracted from these two populations were then individually amplified and labeled for the array analysis. We implemented an algorithm called Rank Product (RP) to calculate the most significantly up-regulated genes in “Late” stage T cells in both patients' samples. The top 2 of the Rank Product value list were: i.) KLCRs, killer cell lectin like receptor members that were associated with increased cell lysis and ii.) IGFBP7, insulin-like growth factor binding protein 7 that were associated with tumor suppression activity. Granulysin, a protein present in cytotoxic granules of cytotoxic T lymphocytes, was shown to be highly up-regulated in one of the patients but not in the other. Furthermore, we used Gene Spring software (Silicon Genetics) as an alternative data analytical approach. KLCRs were also found to be significantly up-regulated by more than 2 fold in both patients' samples. To further validate our observation, more patients will be included in the future study and the up-regulation of these genes will be verified by real time PCR. (This work is supported by the Cancer Institute NSW)

Editorial: It's back –http://www.pigment.org

Editorial: It's back –http://www.pigment.org

Preliminary array analysis reveals novel genes regulated by ovarian steroids in the monkey raphe region

We hypothesize that ovarian hormones may improve serotonin neuron survival. We sought the effect of estradiol (E) and progesterone (P) on novel gene expression in the macaque dorsal raphe region with Affymetrix array analysis. Nine spayed rhesus macaques were treated with either placebo, E or E+P via Silastic implant for 1 month prior to euthanasia (n=3 per treatment). RNA was extracted from a small block of midbrain containing the dorsal raphe and examined on an Agilent Bioanalyzer. The RNA from each monkey was labeled and hybridized to an Affymetrix HG_U95AV Human GeneChip Array. After filtering and sorting, 25 named genes remained that were regulated by E, and 24 named genes remained that were regulated by supplemental P. These genes further sorted into functional categories that would promote neuronal plasticity, transmitter synthesis, and trafficking, as well as reduce apoptosis. The relative abundance of four pivotal genes was examined in all nine animals with quantitative RT-PCR and normalized by glyceraldehyde 3-phosphate dehydrogenase (GAPDH). E+/-P caused a significant threefold reduction in JNK-1 (a pro-apoptosis gene, p<0.007); and a significant sixfold decrease in kynurenine mono-oxygenase (produces neurotoxic quinolones, p<0.05). GABA-A receptor (alpha3 subunit; benzodiazepine site) and E2F1 (interferes with cytokine signaling) were unaffected by E, but increased sevenfold (p<0.02) and fourfold (p<0.009), respectively, upon treatment with P. In summary, subsets of genes related to tissue remodeling or apoptosis were up- or down-regulated by E and P in a tissue block containing the dorsal raphe. These changes could promote cellular resilience in the region where serotonin neurons originate.

Transcriptional Profiling During the Early Differentiation of Granulocyte and Monocyte Progenitors Controlled by Conditional Versions of the E2a–Pbx1 Oncoprotein

The E2a–Pbxl oncoprotein of human pre-B cell leukemia prevents differentiation and maintains continued cell division in cultured myeloid progenitors. Previously, estrogen-dependent forms of E2a–Pbxl were generated that immortalized neutrophil (ECoM-G cells) or monocyte (ECoM-M cells) progenitors and that permitted their terminal differentiation upon estrogen withdrawal. Here, representational difference analysis (RDA) and Affymetrix array analysis are used to identify changes in gene expression that accompany the early differentiation of these cells. The promoters of these genes, whose expression changes upon E2a–Pbxl inactivation, integrate the biochemical mechanism through which E2a–Pbx1 arrests differentiation and maintains cell division. Inactivation of E2a–Pbxl caused the 10- to 80-fold up regulation of a small subset of myeloid differentiation genes (MRP8, Cnlp, NB1, Bactenecin, YM1, Stefin 1, Lipocortin, Lactoferrin, gp91 phox and Ly6-G) and a 10-fold down regulation of the TLE1 corepressor gene, as well as of a group of genes expressed in dividing cells (c-Myc, Nucleophosmin, Spermidine synthase, NOP56, Hnrpa1). Transcription of 97% of cellular genes, including 300 other transcription factor genes (21 Hox genes) and other myeloid genes, varied less than 3-fold, with most varying less than 50%. Therefore, E2a–Pbxl prevents transcription and maintains the cell cycle by a specific rather than a global transcriptional mechanism. Monocyte progenitors were distinguished by persistent expression of IRF8 and of a category of other genes characterized as “interferon-stimulated” (ISG15, ISG20, Ifit1, Ifi202a, Ifi203, Ifi204, Ifi204-related, IRF7 and Ly6-E.1), as well as by the upregulation of the Lrg21 bZip transcription factor gene during late differentiation. The synchronous expression of stage-specific and cell cycle genes regulated by E2a–Pbx1 in these cell lines comprises a model system in which analysis of their promoters can be used as a starting point to backtrack to the transcriptional mechanisms of oncogenesis by E2a–Pbx1.

Connexin26 regulates the expression of angiogenesis-related genes in human breast tumor cells by both GJIC-dependent and -independent mechanisms.

We previously reported that over-expression of connexins in mammary tumor cells retarded tumor growth in vivo in the absence of appreciable gap junction formation, highlighting a possible connexin-linked, but gap junctional intercellular communication (GJIC)-independent mechanism. In the current study, we engineered GJIC-deficient MDA-MB-435 human breast tumor cells to express a chimeric Cx26 where the green fluorescent protein was fused to the amino-terminal of Cx26 (GFP-Cx26). Characterization of this chimeric protein revealed that GFP-Cx26 assembled into non-functional gap junction-like clusters that were impermeable to Lucifer Yellow. In contrast, expression of wild-type Cx26 or Cx26 tagged at the carboxy terminal with yellow fluorescent protein, efficiently rescued GJIC in these tumor cells. Interestingly, by screening 96 tumor-related genes, we observed that the expression of Cx26 or GFP-Cx26 in the tumor cells up-regulated both the transcription and the translation of thrombospondin-1 (TSP-1), an anti-angiogenic molecule. Affymetrix array analysis extended the list of Cx26 or GFP-Cx26 regulated genes by ten candidates including connective tissue growth factor (CTGF), another angiogenesis-related gene. CTGF mRNA and protein levels were found to be down-regulated by both Cx26 and GFP-Cx26. Thus, our data indicates that Cx26 regulates angiogenesis-related molecules by mechanisms that are both GJIC-dependent and -independent.

The FabR (YijC) Transcription Factor Regulates Unsaturated Fatty Acid Biosynthesis in Escherichia coli

Unsaturated fatty acid biosynthesis is a vital facet of Escherichia coli physiology and requires the expression of two genes, fabA and fabB, in the type II fatty acid synthase system. This study links the FabR (YijC) transcription factor to the regulation of unsaturated fatty acid content through the regulation of fabB gene expression. The yijC (fabR) gene was deleted by replacement with a selectable cassette, and the resulting strains (fabR::kan) possessed significantly elevated levels of unsaturated fatty acids, particularly cis-vaccenate, in their membrane phospholipids. The altered fatty acid composition was observed in the fabR::kan fabF1 double mutant pinpointing fabB as the condensing enzyme responsible for the increased cis-vaccenate production. The fabR::kan strains had 4- to 8-fold higher levels of fabB and a 2- to 3-fold increase in fabA transcripts as judged by Northern blotting, Affymetrix array analysis, and real-time PCR. FabR did not regulate the enzymes of fatty acid beta-oxidation. The elevated level of fabB mRNA was reflected by higher condensing enzyme activity in fabR::kan fabF1 double mutants. Thus, FabR functions as a repressor that potently controls the expression of the fabB gene, which in turn, modulates the physical properties of the membrane by altering the level of unsaturated fatty acid production.