Characterizing dye bias in microarray experiments.

MOTIVATION: Spot intensity serves as a proxy for gene expression in dual-label microarray experiments. Dye bias is defined as an intensity difference between samples labeled with different dyes attributable to the dyes instead of the gene expression in the samples. Dye bias that is not removed by array normalization can introduce bias into comparisons between samples of interest. But if the bias is consistent across samples for the same gene, it can be corrected by proper experimental design and analysis. If the dye bias is not consistent across samples for the same gene, but is different for different samples, then removing the bias becomes more problematic, perhaps indicating a technical limitation to the ability of fluorescent signals to accurately represent gene expression. Thus, it is important to characterize dye bias to determine: (1) whether it will be removed for all genes by array normalization, (2) whether it will not be removed by normalization but can be removed by proper experimental design and analysis and (3) whether dye bias correction is more problematic than either of these and is not easily removable. RESULTS: We analyzed two large (each >27 arrays) tissue culture experiments with extensive dye swap arrays to better characterize dye bias. Indirect, amino-allyl labeling was used in both experiments. We found that post-normalization dye bias that is consistent across samples does appear to exist for many genes, and that controlling and correcting for this type of dye bias in design and analysis is advisable. The extent of this type of dye bias remained unchanged under a wide range of normalization methods (median-centering, various loess normalizations) and statistical analysis techniques (parametric, rank based, permutation based, etc.). We also found dye bias related to the individual samples for a much smaller subset of genes. But these sample-specific dye biases appeared to have minimal impact on estimated gene-expression differences between the cell lines.

Conservation of a CD1 multigene family in the guinea pig.

CD1 is a family of cell-surface molecules capable of presenting microbial lipid Ags to specific T cells. Here we describe the CD1 gene family of the guinea pig (Cavia porcellus). Eight distinct cDNA clones corresponding to CD1 transcripts were isolated from a guinea pig thymocyte cDNA library and completely sequenced. The guinea pig CD1 proteins predicted by translation of the cDNAs included four that can be classified as homologues of human CD1b, three that were homologues of human CD1c, and a single CD1e homologue. These guinea pig CD1 protein sequences contain conserved amino acid residues and hydrophobic domains within the putative Ag binding pocket. A mAb specific for human CD1b cross-reacted with multiple guinea pig CD1 isoforms, thus allowing direct analysis of the structure and expression of at least a subset of guinea pig CD1 proteins. Cell-surface expression of CD1 was detected on cortical thymocytes, dermal dendritic cells in the skin, follicular dendritic cells of lymph nodes, and in the B cell regions within the lymph nodes and spleen. CD1 proteins were also detected on a subset of PBMCs consistent with expression on circulating B cells. This distribution of CD1 staining in guinea pig tissues was thus similar to that seen in other mammals. These data provide the foundation for the development of the guinea pig as an animal model to study the in vivo function of CD1.

Effect of promoters and signal sequences on the production of secreted HIV-1 gp120 protein in the baculovirus system.

We compared insect cell production levels of secreted HIV-1 gp120 glycoprotein encoded by five different baculovirus expression constructs. Combinations consisting of one of two baculovirus promoters (very late or hybrid late/very late) and one of three different signal sequences [human tissue plasminogen activator (tpa), human placental alkaline phosphatase (pap), or baculovirus envelope glycoprotein (gp67)] were constructed. Production of secreted gp120 from these constructs was analyzed in two enzyme-linked immunosorbent assay formats, one detecting the total amount of secreted gp120 protein and the other measuring the level of "active" gp120 (as defined by the ability to bind to CD4). We found that for all of the constructs, approximately 50 to 90% of the secreted gp120 protein was active. Furthermore, our results indicated that expression from either promoter yielded comparable production of secreted protein, despite the fact that transcription from the hybrid promoter begins at an earlier time. By contrast, the signal sequence had a much greater effect on the levels of secreted gp120: the tpa leader yielded the highest level of secreted protein, followed by the gp67 and pap sequences. This result suggests that transcription is not a limiting factor in the production of secreted gp120, but rather that downstream processing of the protein is more critical. Furthermore, these results confirm the notion that the "optimal" signal sequence is protein dependent and that an insect-derived signal sequence is not optimal in all cases.

Spontaneous cAMP-dependent derepression of gene expression in stationary phase plays a role in recombinant expression instability.

E. coli recombinant expression systems that utilize lac operon control elements to modulate gene expression are known to produce some amount of uninduced (leaky) gene expression. Previously, we showed that high levels of uninduced gene expression was a major cause of instability in the pET expression system. We show here that the pET system, in which the phage T7 RNA polymerase gene is expressed via lac operon control elements, exhibits leaky expression that increases markedly as cells grown in complex medium enter stationary phase. Moreover, we found that this phenomenon occurs with the chromosomal lac operon as well. Further investigation revealed that stationary phase leaky expression requires cyclic AMP, and that substantial leaky expression could be effected in log phase cells by adding cyclic AMP and acetate at pH6.0. Finally, a comparison of otherwise isogenic cya and wild-type hosts showed that expression stability and plasmid maintenance in the cya host is greatly enhanced, even when cells are passaged repeatedly in non-selection medium. These findings both provide a method to enhance the stability of lac-based recombinant expression systems, and suggest that derepression of the lac operon in the absence of inducer may be part of a general cellular response to nutrient limitation.

An expression vector encoding a lacZ reporter gene facilitates identification of stable, high-producing CHO cell clones.

We describe a vector and a streamlined procedure for isolating high-producing stable mammalian cell transfectants. The vector encodes the Escherichia coli lacZ gene as a reporter. We show that levels of beta-galactosidase activity, assayed in situ in clonal isolates, can be used to identify clones producing high levels of the protein of interest (in this case, soluble human CD4 protein).

Regulation of sCD4-183 gene expression from phage-T7-based vectors in Escherichia coli.

In this paper, we describe various parameters affecting the regulation of expression of the sCD4-183 gene, encoding the 183-amino-acid soluble human two-domain CD4 protein, from phage-T7-based pET vectors. We demonstrated that for the sCD4-183 protein, the highest protein yield was obtained using vector pET-9a, in which neither expression of the T7 RNA polymerase-encoding gene nor the target gene was tightly regulated. The highest overall protein yield was obtained from cells grown for 24 h in the absence of inducer, a strategy that may be generally useful for production of less toxic proteins. We also describe two modifications of the pET vector system that effectively minimized leaky (uninduced) expression and enhanced plasmid stability. These have potential use in the production of toxic proteins, or of non-toxic proteins produced in high-density cultures.

Accumulation of p53 tumor suppressor gene protein: an independent marker of prognosis in breast cancers.

BACKGROUND: Mutations of the tumor suppressor gene p53 have been identified in breast cancer cell lines, and some breast carcinomas are detectable by immunohistochemical assay because of p53 protein accumulation. PURPOSE: This study was designed to determine whether p53 protein accumulation in breast cancers correlates with p53 gene mutation, with survival, and with five pathobiologic factors associated with prognosis. METHODS: IgG1 monoclonal antibody to human p53 protein (PAb 1801) and immunohistochemical methods were used to detect p53 protein accumulation in archival formalin-fixed, paraffin-embedded, randomly selected carcinomas. We studied 295 invasive ductal carcinomas from the Massachusetts General Hospital; 151 were determined to be sporadic (not hereditary). We also studied 97 invasive ductal carcinomas--21 sporadic and 76 familial (hereditary)--from Creighton University. In addition, we examined 31 archival in situ carcinomas, 15 snap-frozen invasive ductal carcinomas, primary cell cultures from three benign breast tissue samples, and breast carcinoma cell lines MDA-MB-231 and MDA-MB-468. RESULTS: Nuclear p53 protein was observed in 16% of the 31 in situ carcinomas, 22% of the 172 sporadic carcinomas, 34% of the 50 tumors from patients with familial breast cancer, 52% of the 23 tumors from patients with the familial breast and ovarian cancer syndrome, and all three tumors from two patients with the Li-Fraumeni syndrome. There was complete concordance between p53 gene mutation and p53 protein accumulation in the 15 snap-frozen carcinomas and in both breast carcinoma cell lines. Statistically significant associations of p53 protein accumulation with estrogen receptor negativity and with high nuclear grade were found. There were statistically significant associations, independent of other prognostic factors, between p53 protein accumulation and metastasis-free and overall survival, for randomly accrued and for both sporadic and familial tumors. CONCLUSIONS: Immunohistochemically detected p53 protein accumulation was an independent marker of shortened survival and was seen more often in familial than in sporadic carcinomas. Our findings also suggest a correlation between p53 protein accumulation and p53 gene mutation.

Clonal analysis of bcr-abl rearrangement in T lymphocytes from patients with chronic myelogenous leukemia.

The cytogenetic hallmark of chronic myelogenous leukemia (CML) is the Philadelphia chromosome (Ph1), which reflects a chromosomal translocation t(9;22) and a rearrangement of the ABL and bcr genes. This marker is found in all cells arising from the same malignant precursor cell and can be detected in CML cells of the myeloid, monocytic, erythroid, and B-lymphocyte lineage. It is, however, controversial as to whether T lymphocytes of CML patients carry this gene rearrangement. An answer to this question would clarify whether the translocation in CML occurs in a pluripotent hematopoietic stem cell or in a precursor cell already committed to certain lineages, but not the T-cell lineage. To address this question, we established T-cell clones from peripheral venous blood cells of four patients with CML and screened these clones for bcr-abl fusion transcripts by means of polymerase chain reaction and Southern blot analysis. In four T-cell clones of three of these patients, the bcr-abl transcript could be detected. None of 12 T-cell clones of the fourth patient disclosed detectable bcr-abl amplification product. Both CD4+ as well as CD8+ clones displayed fused bcr-abl sequences. These data imply that in CML some but not all T lymphocytes may originate from the Ph1-positive stem cell.

Single-base mutational analysis of cancer and genetic diseases using membrane bound modified oligonucleotides.

A convenient format for the detection of PCR amplified sequences is the hybridization of the PCR products to oligonucleotide probes which are immobilized on a solid phase. We describe a new method for site-specific attachment of such probe oligonucleotides to nylon membranes. The method is based on the formation of an amide bond between carboxyl groups present on the membranes and amino-linkers situated on the 5' end of the oligonucleotides. The covalent attachment is via a carbodiimide mediated condensation. The single, 5' end attachment of the oligonucleotides to the membrane surface leaves the probe free to interact with complementary sequences, thus increasing the hybridization efficiency relative to methods where heat or ultraviolet light is used for non-specific fixation. Using biotinylated PCR products in hybridization reactions along with a non-radioactive chemiluminescent detection system, high efficiency hybridization is obtained as well as a very good signal to noise ratio. The method has been applied successfully to the detection of RAS point mutations, cystic fibrosis deletion and point mutations and others. The sensitivity, simplicity and reproducibility of this method make it an ideal tool for the diagnosis of infectious and genetic diseases, as well as analysis of mutations in neoplasias, HLA typing and other areas.

Bladder cancer. Human leukocyte antigen II, interleukin-6, and interleukin-6 receptor expression determined by the polymerase chain reaction.

The prediction of tumour biology rarely rests upon a single characteristic of the malignancy. The analysis of a single gene can complement standard histologic evaluation. The investigation of new parameters as well as the routine clinical analysis of gene expression is often limited because of the small amount of tissue available. This is particularly true of de novo human bladder cancers because they are generally small or handled in such a way as to hinder the analysis of multiple different parameters. Analysis of expressed mRNA by the polymerase chain reaction (RNA/PCR) is a method that allows the development of a profile of bladder cancer gene expression. The authors report the use of the RNA/PCR method to examine in bladder cancer the expression of the human leukocyte antigen (HLA) class II gene family (HLA-DR, DQ, and DP) as well as interleukin-6 (IL-6) and the interleukin-6 receptor (IL-6R). All de novo transitional cell carcinomas, one squamous carcinoma, and two transitional cell carcinoma cell lines expressed the majority of HLA class II genes. All samples expressed IL-6R RNA whereas production of IL-6 message was limited to one of the cell lines and to the high-grade bladder cancers. These results were combined with stage, grade, and DNA content to develop a profile of the cancers examined. Although an improved predictive index based on gene expression analysis by RNA/PCR has not been realized, a broader survey of human tumors for expression of these genes and others is likely to refine the classification of bladder cancer.

Molecular relapse in chronic myelogenous leukemia patients after bone marrow transplantation detected by polymerase chain reaction.

Relapse of chronic myelogenous leukemia after bone marrow transplantation can be detected by using clinical, cytogenetic, or molecular tools. A modification of the polymerase chain reaction can be used in patients to detect low levels of the BCR-ABL-encoded mRNA transcript, a specific marker for chronic myelogenous leukemia. Early detection of relapse after bone marrow transplantation could potentially alter treatment decisions. We prospectively evaluated 19 patients for evidence of molecular relapse, cytogenetic relapse, and clinical relapse after bone marrow transplantation. We used the polymerase chain reaction to detect residual BCR-ABL mRNA in patients followed up to 45 months after treatment (median, 15 months; range, 6-45 months) and found 4 patients with BCR-ABL mRNA expression following bone marrow transplantation. In 2 patients BCR-ABL mRNA was detected in all samples, and both have developed cytogenetic relapse. In 1 patient BCR-ABL mRNA was detected transiently during the first month after transplant but was undetectable thereafter. The fourth patient had BCR-ABL mRNA 6 months after bone marrow transplantation but not in prior samples. Fifteen patients did not express detectable BCR-ABL mRNA. All 19 patients remain in clinical remission. In this prospective study of chronic myelogenous leukemia patients treated with bone marrow transplantation, molecular relapse preceded cytogenetic relapse in those patients who persistently express BCR-ABL mRNA. We recommend using standard clinical and cytogenetic testing to make patient care decisions until further follow-up determines the clinical outcome of those patients with residual BCR-ABL mRNA transcripts detected by polymerase chain reaction.

Molecular biology of macrophage colony-stimulating factor.

In this chapter we have described one of the more complex hemopoietic factors, M-CSF. The single-copy M-CSF gene is almost 21 kb in length and is arranged into 10 exons and 9 introns. Expression of the gene at the RNA level is heterogeneous, and several species of M-CSF mRNA have been found in human and murine cells and tissues. In human cells the different mRNAs arise from alternative splicing of the nuclear RNA precursor in both coding and noncoding regions. This results in mRNAs encoding two distinct M-CSF proteins, 256 and 554 amino acids in length. In murine cells only a 552-amino-acid form has been found thus far. All forms of M-CSF have a 32-amino-acid signal peptide and a 23-amino-acid hydrophobic region near the carboxy-terminus, which resembles a transmembrane domain. A large portion of the carboxy-terminal end, including the hydrophobic region, is not found in the mature protein. Thus, the primary translation product of M-CSF is a prepropolypeptide, with processing occurring at both amino- and carboxy-terminal ends. The exact size of the mature protein is still somewhat in doubt, but deletion mutagenesis from the carboxy-terminal end indicates that the protein may be as small as 150 amino acids and still be functional. Site-directed mutagenesis has also shown that the first seven cysteines in the mature molecule are probably necessary for biological activity, whereas the next two cysteine residues are not. In spite of the heavy glycosylation found in the native protein, removal of the N-linked glycosylation signals does not seem to affect activity to any great degree. The M-CSF gene and its receptor, C-FMS, are tightly linked on the long arm of chromosome 5, a unique finding in the ligand/receptor field. This region also contains the genes for GM-CSF, IL-3, ECGF, and the receptor for PDGF. A similar situation may exist on chromosome 11 of the mouse. The close linkage of these factors and receptors is the probable cause for the disorders of hemopoiesis that arise when deletions occur in this area. The preceding discussion has shown how quickly the area of M-CSF molecular biology has advanced in the past 2-3 years. A great deal of effort is now being directed toward expressing M-CSF at high levels in a variety of prokaryotic and eukaryotic systems.(ABSTRACT TRUNCATED AT 400 WORDS)

P185BCR-ABL in two patients with late appearing Philadelphia chromosome-positive acute nonlymphocytic leukemia.

Two patients with acute nonlymphocytic leukemia (ANLL) who had normal karyotypes at diagnosis and developed the Philadelphia (Ph) translocation during leukemia relapse are described in this report. Patient 1 relapsed with Ph-positive acute leukemia, FAB classification M1. The Ig heavy chain locus and T cell receptor gamma and beta genes of relapse cells from this patient were all found to be germline configuration confirming the diagnosis of M1 acute leukemia. Patient 2 displayed a complex karyotypic evolution leading to Ph-positive M4 relapse. Ph-positive relapse specimens from both patients expressed P185BCR-ABL protein and RNA gene products that were identified serologically and by polymerase chain amplification of the BCR-ABL RNA junction. In vitro derived myeloid cell lines from relapse M1 leukemia cells of patient 1 also expressed the P185BCR-ABL protein. In two described patients, late appearance of the Ph translocation that encodes P185BCR-ABL coincided with relapse of acute leukemia. We conclude that P185BCR-ABL may be a strong indicator of Ph-positive acute leukemias.

Canine and bovine ras family expression detected and discriminated by use of polymerase chain reaction.

Cellular proto-oncogenes are highly conserved genes thought to be critical in cell growth and differentiation. In this study, we used human sequence designed oligonucleotide primers to detect and discriminate c-Ha-ras-1, c-Ki-ras-2 and c-N-ras genes of dogs and cows by polymerase chain reaction (PCR) amplification of genomic DNA (DNA/PCR). Further, we have applied PCR for analysis of expressed mRNA transcribed from the RAS genes (RNA/PCR).

Baculovirus expression of functional P210 BCR-ABL oncogene product.

The chronic myelogenous leukemia-associated P210 BCR-ABL oncogene protein product has been produced using the baculovirus expression system. High-level expression of the P210 BCR-ABL protein required the removal of GC rich 5' non-coding sequences. P210 BCR-ABL synthesized in insect cells is an active tyrosine protein kinase indistinguishable from P210 BCR-ABL isolated from human cells. Both proteins utilize angiotensin II as a phosphate acceptor in vitro with a Km for ATP of approximately 1.5 microM. P210 BCR-ABL produced in insect cells undergoes autophosphorylation in vitro and in vivo. Gel filtration of P210 BCR-ABL reveals that the protein elutes as a high molecular weight complex of about 800 kD. Approximately 4 to 5 mg of P210 BCR-ABL is produced in one liter of infected insect cells. Following cell disruption and a three-step ion exchange and gel filtration purification procedure, 0.4 mg of soluble P210 BCR-ABL is obtained per liter of suspension culture. An alternative procedure employing detergent extraction and immunoaffinity chromatography gave higher yields and purity from smaller amounts of infected cell extracts. The availability of intact, soluble and enzymatically active P210 BCR-ABL represents a significant advance for studying the biochemical and biophysical properties of the ABL oncogene family of proteins.

Diagnosis of chronic myeloid and acute lymphocytic leukemias by detection of leukemia-specific mRNA sequences amplified in vitro.

The Philadelphia chromosome is present in more than 95% of chronic myeloid leukemia patients and 13% of acute lymphocytic leukemia patients. The Philadelphia translocation, t(9;22), fuses the BCR and ABL genes resulting in the expression of leukemia-specific, chimeric BCR-ABL messenger RNAs. To facilitate diagnosis of these leukemias, we have developed a method of amplifying and detecting only the unique mRNA sequences, using an extension of the polymerase chain reaction technique. Diagnosis of chronic myeloid and acute lymphocytic leukemias by this procedure is rapid, much more sensitive than existing protocols, and independent of the presence or absence of an identifiable Philadelphia chromosome.

Human CSF-1: gene structure and alternative splicing of mRNA precursors.

Bone marrow progenitor cells differentiate into mononuclear phagocytes in the presence of colony stimulating factor-1 (CSF-1). Characterization of the human CSF-1 gene shows that it contains 10 exons and 9 introns, which span 20 kb. Analysis of multiple CSF-1 transcripts demonstrates that alternate use of exon 6 splice acceptor sites and 3' noncoding sequence exons occurs. These alternatively spliced transcripts can encode either a 224 or a 522 amino acid CSF-1. Implications of differential splicing for the production and function of CSF-1 are discussed.

Assignment of CSF-1 to 5q33.1: evidence for clustering of genes regulating hematopoiesis and for their involvement in the deletion of the long arm of chromosome 5 in myeloid disorders.

The CSF-1 gene encodes a hematopoietic colony-stimulating factor (CSF) that promotes growth, differentiation, and survival of mononuclear phagocytes. By using somatic cell hybrids and in situ hybridization, we localized this gene to human chromosome 5 at bands q31 to q35, a chromosomal region that is frequently deleted [del(5q)] in patients with myeloid disorders. By in situ hybridization, the CSF-1 gene was found to be deleted in the 5q- chromosome of a patient with refractory anemia who had a del(5)(q15q33.3) and in that of a second patient with acute nonlymphocytic leukemia de novo who had a similar distal breakpoint [del(5)(q13q33.3)]. The gene was present in the deleted chromosome of a third patient, with therapy-related acute nonlymphocytic leukemia, who had a more proximal breakpoint in band q33 [del(5)(q22q33.1)]. Hybridization of the CSF-1 probe to metaphase cells of a fourth patient, with acute nonlymphocytic leukemia de novo, who had a rearrangement of chromosomes 5 and 21 [ins(21;5)(q22;q31.3q33.1)] resulted in labeling of the breakpoint junctions of both rearranged chromosomes; this suggested that CSF-1 is located at 5q33.1. Thus, a small segment of chromosome 5 contains GM-CSF (the gene encoding the granulocyte-macrophage CSF), CSF-1, and FMS, which encodes the CSF-1 receptor, in that order from the centromere; this cluster of genes may be involved in the altered hematopoiesis associated with a deletion of 5q.