Identification of differentially expressed genes in hepatocellular carcinoma and metastatic liver tumors by oligonucleotide expression profiling.

BACKGROUND: The characterization of differentially expressed genes between cancerous and normal tissues is an important step in the understanding of tumorigenesis. Global gene expression profiling with microarrays has now offered a powerful tool to measure the changes of thousands of genes in any carcinoma tissues in an effort to identify these key disease-related genes. To compare the gene expression of a primary liver carcinoma, metastatic carcinoma to the liver, and normal liver, the authors analyzed tissue from six primary hepatocellular carcinomas (HCCs), five colorectal adenocarcinoma metastases to the liver, and eight normal livers. METHODS: Samples were processed from total RNA to fragmented cRNA and hybridized onto Affymetrix GeneChip(R) expression arrays. Analyses were performed to determine the consensus pattern of gene expression for primary liver carcinoma, metastatic liver carcinoma, and normal liver tissue and their changes in expression level. RESULTS: In hepatocellular carcinoma, 842 genes were overexpressed, and 393 genes were underexpressed in comparison with genes of normal liver tissue. Of note, 7 of the 20 most increased identified known genes previously have been associated with liver carcinoma or other types of cancers. The 13 additional identified genes until now have not previously shown strong association with cancers. Furthermore, the authors identified 42 genes and 24 expressed sequence tags that are expressed at a significant level in both HCC and metastastic tumors, presenting a list of marker genes indicative of cancerous liver tissue. CONCLUSIONS: In this study, genes that can be involved in the production of and maintenance of hepatic carcinomas were identified. These data offer new insight into genes that are potentially important in the pathogenesis of liver carcinoma, as well as additional targets for new strategies for cancer therapy and treatment.

Cloning of dimethylglycine dehydrogenase and a new human inborn error of metabolism, dimethylglycine dehydrogenase deficiency.

Dimethylglycine dehydrogenase (DMGDH) (E.C. number 1.5.99.2) is a mitochondrial matrix enzyme involved in the metabolism of choline, converting dimethylglycine to sarcosine. Sarcosine is then transformed to glycine by sarcosine dehydrogenase (E.C. number 1.5.99.1). Both enzymes use flavin adenine dinucleotide and folate in their reaction mechanisms. We have identified a 38-year-old man who has a lifelong condition of fishlike body odor and chronic muscle fatigue, accompanied by elevated levels of the muscle form of creatine kinase in serum. Biochemical analysis of the patient's serum and urine, using (1)H-nuclear magnetic resonance NMR spectroscopy, revealed that his levels of dimethylglycine were much higher than control values. The cDNA and the genomic DNA for human DMGDH (hDMGDH) were then cloned, and a homozygous A-->G substitution (326 A-->G) was identified in both the cDNA and genomic DNA of the patient. This mutation changes a His to an Arg (H109R). Expression analysis of the mutant cDNA indicates that this mutation inactivates the enzyme. We therefore confirm that the patient described here represents the first reported case of a new inborn error of metabolism, DMGDH deficiency.

Structure and analysis of the human dimethylglycine dehydrogenase gene.

Dimethylglycine dehydrogenase (DMGDH; E.C. 1.5.99.2) is an enzyme involved in the catabolism of choline, catalyzing the oxidative demethylation of dimethylglycine (DMG) to form sarcosine. Subsequently, sarcosine dehydrogenase (SDH; E.C. 1.5.99.1) converts sarcosine to glycine via a similar reaction. Both enzymes are found as monomers in the mitochondrial matrix, and both contain 1 mol of covalently bound flavin adenine dinucleotide. DMGDH and SDH also utilize a noncovalently bound folate coenzyme that receives the "1-carbon" groups that are removed by DMGDH and SDH, forming "active formaldehyde." We have recently described a new inborn error of metabolism of DMGDH characterized by an unusual fish-like body odor. To augment our study of this new disorder, we have isolated two human genomic clones that together contain 16 exons of coding sequence for the hDMGDH gene. Fluorescent in situ hybridization analysis of the hDMGDH gene indicates that it is found on chromosome 5q12.2-q12.3. In addition, several polymorphisms have been identified in the hDMGDH cDNA sequence. Population analysis of two Ser/Pro polymorphisms found 367 amino acids apart reveals a skew of alleles, with the haplotypes Ser/Pro or Pro/Ser (79%) overrepresented compared to the number of Ser/Ser or Pro/Pro alleles observed. Possible functional consequences of these findings are discussed. Characterization of the gene structure for hDMGDH will aid in the study of patients with inherited defects of this enzyme.

Molecular basis of hyperargininemia: structure-function consequences of mutations in human liver arginase.

Hyperargininemia is a rare autosomal recessive disorder that results from a deficiency of hepatic type I arginase. At the genetic level, this deficiency in arginase activity is a consequence of random point mutations throughout the gene that lead to premature termination of the protein or to substitution mutations. Given the high degree of sequence homology between human liver and rat liver enzymes, we have mapped both patient and nonpatient mutations of the human enzyme onto the structure of the rat liver enzyme to rationalize the molecular basis for the low activities of these mutant arginases. Mutations identified in hyperargininemia patients affect the structure and function of the enzyme by compromising active-site residues, packing interactions in the protein scaffolding, and/or quaternary structure by destabilizing the assembly of the arginase trimer.

The human arginases and arginase deficiency.

Arginase is the final enzyme in the urea cycle. Its deficiency is the least frequently described disorder of this cycle. It results primarily in elevated blood arginine, and less frequently in either persistent or acute elevations in blood ammonia. This appears to be due to a second arginase locus, expressed primarily in the kidney, which can be recruited to compensate, in part, for the deficiency of liver arginase. The liver arginase gene structure permitted study of the molecular pathology of patients with the disorder and the results of these studies and the inferences about the protein structure are presented. The conserved regions among all arginases allowed the cloning of AII, the second arginase isoform. It has been localized to the mitochondrion and is thought to be involved in ornithine biosynthesis. It shares the major conserved protein sequences, and structural features of liver arginase gene are also conserved. When AI and AII from various species are compared, it appears that the two diverged some time prior to the evolution of amphibians. The evidence for the role of AII in nitric oxide and polyamine metabolism is presented and this appears consonant with the data on the tissue distribution.

Cloning and characterization of the mouse and rat type II arginase genes.

Two forms of arginase, both catalyzing the hydrolysis of arginine to ornithine and urea, are found in animals ranging from amphibians to mammals. In humans, inherited deficiency of hepatic or type I arginase results in hyperargininemia, a syndrome characterized by periodic episodes of hyperammonemia, spasticity, and neurological deterioration. In these patients, a second extrahepatic or type II arginase activity is significantly increased, an induction that may partially compensate for the lack of AI activity and apparently mitigates some of the clinical effects of the condition. Cloning and characterization of the human AII cDNA was recently accomplished. The cloning, sequencing, and partial characterization of the mouse and rat AII cDNAs are reported herein. The DNA sequences predicted polypeptides of 354 amino acids, including a N-terminal mitochondrial import signal. Sequence homology to the human type II arginase, arginase activity data, and immunoprecipitation with an anti-AII antibody confirm the identity of these cloned genes as rodent extrahepatic type II arginases.

Cloning and characterization of the human type II arginase gene.

There are two forms of arginase in humans, both catalyzing the hydrolysis of arginine to ornithine and urea. Recent studies in animal models and in Type I arginase-deficient patients suggest that Type II arginase is inducible and may play an important role in the regulation of extra-urea cycle arginine metabolism by modulating cellular arginine concentrations. We PCR amplified and cloned the human Type II arginase gene, the first nonliver arginase gene reported in mammals. While sequence homology to Type I arginase, arginase activity data, and immunoprecipitation with an anti-AII antibody confirm the identity of this gene, Northern blot analysis demonstrates its differential expression in the brain, prostate, and kidney. Type II arginase may be an important part of the arginine regulatory system affecting nitric oxide synthase, arginine decarboxylase, kyotorphin synthase, and arginine-glycine transaminase activities and polyamine and proline biosynthesis.

Delivery of cytosolic liver arginase into the mitochondrial matrix space: a possible novel site for gene replacement therapy.

As a toxic metabolic byproduct in mammals, excess ammonia is converted into urea by a series of five enzymatic reactions in the liver that constitute the urea cycle. A portion of this cycle takes place in the mitochondria, while the remainder is cytosolic. Liver arginase (L-arginine ureahydrolase, A1) is the fifth enzyme of the cycle, catalyzing the hydrolysis of arginine to ornithine and urea within the cytosol. Patients deficient in this enzyme exhibit hyperargininemia with episodic hyperammonemia and long-term effects of mental retardation and spasticity. However, the hyperammonemic effects are not so catastrophic in arginase deficiency as compared to other urea cycle defects. Earlier studies have suggested that this is due to the mitigating effect of a second isozyme of arginase (AII) expressed predominantly in the kidney and localized within the mitochondria. In order to explore the curious dual evolution of these two isozymes, and the ways in which the intriguing, aspects of AII physiology might be exploited for gene replacement therapy of AI deficiency, the cloned cDNA for human AI was inserted into an expression vector downstream from the mitochondrial targeting leader sequence for the mitochondrial enzyme ornithine transcarbamylase and transfected into a variety of recipient cell types. AI expression in the target cells was confirmed by northern blot analysis, and competition and immunoprecipitation studies showed successful translocation of the exogenous AI enzyme into the transfected cell mitochondria. Stability studies demonstrated that the translocated enzyme had a longer half-life than either native cytosolic AI or mitochondrial AII. Incubation of the transfected cells with increasing amounts of arginine produced enhanced levels of mitochondrial AI activity, a substrate-induced effect that we have previously seen with native AII but never AI. Along with exploring the basic biological questions of regulation and subcellular localization in this unique dual-enzyme system, these results suggest that the mitochondrial matrix space may be a preferred site for delivery of enzymes in gene replacement therapy.

Loss of function mutations in conserved regions of the human arginase I gene.

We have utilized SSCP analysis to identify disease-causing mutations in a cohort with arginase deficiency. Each of the patient's mutations was reconstructed in vitro by site-directed mutagenesis to determine the effect of the mutations on enzyme activity. In addition we identified six areas of cross-species homology in the arginase protein, four containing conserved histidine residues thought to be important to Mn(2+)-dependent enzyme function. Mapping patient mutations in relationship to the conserved regions indicates that substitution mutations within the conserved regions and randomly occurring microdeletions and nonsense mutations have a significant effect on enzymatic function. In vitro mutagenesis was utilized to create nonpatient substitution mutations in the conserved histidine residues to verify their importance to arginase activity. As expected, replacement of histidine residues with other amino acids dramatically reduces arginase activity levels in our bacterial expression system.

Functional and molecular analysis of liver arginase promoter sequences from man and Macaca fascicularis.

Functional and DNA binding analyses were used to investigate transcriptional regulation of liver arginase, a mammalian urea cycle enzyme with marked tissue specificity. Reporter constructs containing the proximal 111 bp of the gene from man and Macaca fascicularis showed over sixfold background activity in HepG2 hepatoma cells, which express significant levels of liver arginase, and 12-fold background activity in minimally expressing HEK cells. Longer constructs, active in both cell lines, showed greater activity in the liver cell line. The constructs showed no activity in arginase-negative NIH 3T3 fibroblasts. A 54-bp dyad insert present in the human sequence and absent in M. fascicularis did not affect function. DNA binding analyses localized multiple liver-specific complexes as well as complexes shared among cell types. Little binding was evident in fibroblast extracts. Despite liver-specific binding, there was no evidence of a strong liver-specific enhancer. HEK and NIH 3T3 nuclear extracts showed strikingly different patterns of DNA binding. These studies demonstrate that molecular regulation of liver arginase transcription is complex and that control mechanisms differ among tissue types.

pUC18rspL: a plasmid vector for indirect selection and direct sequencing of promoter-bearing DNA fragments.

Plasmid pUC18rspL is a 3.788-kilobase pair vector, derived from pUC18, pKK232-8 and pHSG664, which identifies promoter-bearing DNA fragments functional in StrA E. coli by activation of a promoterless streptomycin-sensitive gene cartridge (rspL). Expression of the plasmid-borne rspL gene leads to sensitivity dominance and death of the cell. Promoter-bearing DNA fragments can be cloned within a synthetic polylinker containing 12 unique restriction nuclease target sequences. After transformation of StrA E. coli TB1, ampicillin-resistant transformants are replica plated on medium containing ampicillin and streptomycin to identify promoters cloned in their functional orientation. These elements can be sequenced without additional subcloning steps from the M13 universal forward primer hybridization site located 5' of the polylinker in the pUC18 contribution. Transcriptional terminators are cloned 3' of the rspL gene to maintain a balance between transcription and replication when high signal strength promoters such as the E. coli Tac promoter are analyzed. pUC18rspL is used to clone transcriptional promoters from the extreme thermophile T. aquaticus. Promoter signal strength can be estimated by determining the extent of sensitivity dominance conferred to the host.