Molecular characterization of Chinese hamster cells mutants affected in adenosine kinase and showing novel genetic and biochemical characteristics.

BACKGROUND: Two isoforms of the enzyme adenosine kinase (AdK), which differ at their N-terminal ends, are found in mammalian cells. However, there is no information available regarding the unique functional aspects or regulation of these isoforms. RESULTS: We show that the two AdK isoforms differ only in their first exons and the promoter regions; hence they arise via differential splicing of their first exons with the other exons common to both isoforms. The expression of these isoforms also varied greatly in different rat tissues and cell lines with some tissues expressing both isoforms and others expressing only one of the isoforms. To gain insights into cellular functions of these isoforms, mutants resistant to toxic adenosine analogs formycin A and tubercidin were selected from Chinese hamster (CH) cell lines expressing either one or both isoforms. The AdK activity in most of these mutants was reduced to <5% of wild-type cells and they also showed large differences in the expression of the two isoforms. Thus, the genetic alterations in these mutants likely affected both regulatory and structural regions of AdK. We have characterized the molecular alterations in a number of these mutants. One of these mutants lacking AdK activity was affected in the conserved NxxE motif thereby providing evidence that this motif involved in the binding of Mg2+ and phosphate ions is essential for AdK function. Another mutant, FomR-4, exhibiting increased resistance to only C-adenosine analogs and whose resistance was expressed dominantly in cell-hybrids contained a single mutation leading to Ser191Phe alteration in AdK. We demonstrate that this mutation in AdK is sufficient to confer the novel genetic and biochemical characteristics of this mutant. The unusual genetic and biochemical characteristics of the FomR-4 mutant suggest that AdK in this mutant might be complexed with the enzyme AMP-kinase. Several other AdK mutants were altered in surface residues that likely affect its binding to the adenosine analogs and its interaction with other cellular proteins. CONCLUSIONS: These AdK mutants provide important insights as well as novel tools for understanding the cellular functions of the two isoforms and their regulation in mammalian cells.

Structure and protein-protein interaction studies on Chlamydia trachomatis protein CT670 (YscO Homolog).

Comparative genomic studies have identified many proteins that are found only in various Chlamydiae species and exhibit no significant sequence similarity to any protein in organisms that do not belong to this group. The CT670 protein of Chlamydia trachomatis is one of the proteins whose genes are in one of the type III secretion gene clusters but whose cellular functions are not known. CT670 shares several characteristics with the YscO protein of Yersinia pestis, including the neighboring genes, size, charge, and secondary structure, but the structures and/or functions of these proteins remain to be determined. Although a BLAST search with CT670 did not identify YscO as a related protein, our analysis indicated that these two proteins exhibit significant sequence similarity. In this paper, we report that the CT670 crystal, solved at a resolution of 2 A, consists of a single coiled coil containing just two long helices. Gel filtration and analytical ultracentrifugation studies showed that in solution CT670 exists in both monomeric and dimeric forms and that the monomer predominates at lower protein concentrations. We examined the interaction of CT670 with many type III secretion system-related proteins (viz., CT091, CT665, CT666, CT667, CT668, CT669, CT671, CT672, and CT673) by performing bacterial two-hybrid assays. In these experiments, CT670 was found to interact only with the CT671 protein (YscP homolog), whose gene is immediately downstream of ct670. A specific interaction between CT670 and CT671 was also observed when affinity chromatography pull-down experiments were performed. These results suggest that CT670 and CT671 are putative homologs of the YcoO and YscP proteins, respectively, and that they likely form a chaperone-effector pair.

Subcellular localization of adenosine kinase in mammalian cells: The long isoform of AdK is localized in the nucleus.

Two isoforms of adenosine kinase (AdK) have been identified in mammalian organisms with the long isoform (AdK-long) containing extra 20-21 amino acids at the N-terminus (NTS). The subcellular localizations of these isoforms are not known and they contain no identifiable targeting sequence. Immunofluorescence labeling of mammalian cells expressing either only AdK-long or both isoforms with AdK-specific antibody showed only nuclear labeling or both nucleus and cytoplasmic labeling, respectively. The AdK-long and -short isoforms fused at the C-terminus with c-myc epitope also localized in the nucleus and cytoplasm, respectively. Fusion of the AdK-long NTS to green fluorescent protein also resulted in its nuclear localization. AdK-long NTS contains a cluster of conserved amino acids (PKPKKLKVE). Replacement of KK in this sequence with either AA or AD abolished its nuclear localization capability, indicating that this cluster likely serves as a nuclear localization signal. AdK in nucleus is likely required for sustaining methylation reactions.

Mycobacterial adenosine kinase is not a typical adenosine kinase.

Adenosine kinase (AK) is only found in eukaryotes. Recently, a Mycobacterium tuberculosis (MTub) protein exhibiting greater sequence similarity to ribokinases (RK) was identified as AK. We have expressed AKs from MTub, human and Chinese hamster (CH) cells in Escherichia coli and also AK from human and MTub in AK-deficient CH cells. While both E. coli and CH cells expressing mammalian AKs efficiently metabolized various adenosine analogs, those expressing MTub-AK were completely inactive. The AK activity of the MTub protein was very low (50-fold lower than E. coli RK) and it was not stimulated by phosphate or inhibited by several AK inhibitors. These results raise questions over MTub protein's true function and whether it functions as AK in cells.

Conserved inserts in the Hsp60 (GroEL) and Hsp70 (DnaK) proteins are essential for cellular growth.

The Hsp60 and Hsp70 chaperones contain a number of conserved inserts that are restricted to particular phyla of bacteria. A one aa insert in the E. coli GroEL and a 21-23 insert in the DnaK proteins are specific for most Gram-negative bacteria. Two other inserts in DnaK are limited to certain groups of proteobacteria. The requirement of these inserts for cellular growth was examined by carrying out complementation studies with temperature-sensitive (T(s)) mutants of E. coli groEL or dnaK. Our results demonstrate that deletion or most changes in these inserts completely abolished the complementation ability of the mutant proteins. Studies with GroEL and DnaK from some other species that either lacked or contained these inserts also indicated that these inserts are essential for growth of E. coli. The DnaK from some bacteria contains a two aa insert that is not found in E. coli. Introduction of this insert into the E. coli DnaK also led to its inactivation, indicating that these inserts are specific for different groups. We postulate that these conserved inserts that are localized in loop regions on protein surfaces, are involved in some ancillary functions that are essential for the groups of bacteria where they are found.

Unusual cellular disposition of the mitochondrial molecular chaperones Hsp60, Hsp70 and Hsp10.

A number of molecular chaperones in mammalian cells are localized in mitochondria and they are presumed to function mainly within this organelle. However, there is now compelling evidence that these chaperones are also localized at a variety of other sites/compartments in cells where they perform important functions. These proteins include: (i) the major chaperonin Hsp60 (or P1), which was identified in mammalian cells as a protein altered in mutants resistant to microtubule inhibitors and is involved in numerous functions at the cell surface and in other compartments; (ii) the Hspl0 or Cpn10 protein, which is a co-chaperone for Hsp60 in protein folding but also serves as an early pregnancy factor in maternal serum; and (iii) the mHsp70 protein, which plays a central role in mitochondrial protein import but is also important for cellular senescence (mortalin) and antigen presentation processes. The presence of these mitochondrial chaperones at specific extramitochondrial locations greatly broadens the range of functions that they can carry out in cells. However, these observations also raise important questions regarding the mechanisms by which these proteins reach these extramitochondrial locations. My paper will review some work in this area and discuss the significance of these results.

Identification and characterization of human ribokinase and comparison of its properties with E. coli ribokinase and human adenosine kinase.

The gene responsible for ribokinase (RK) in human/eukaryotic cells has not yet been identified/characterized. Blast searches with E. coli RK have identified a human protein showing significant similarity to the bacterial RK. The cDNA for this protein was expressed in E. coli and the recombinant protein efficiently phosphorylated ribose to ribose-5-phosphate using ATP, confirming its identity as RK. In contrast to ribose, the enzyme exhibited very little to no phosphorylation of D-arabinose, D-xylose, D-fructose and D-galactose. The catalytic activity of human RK was dependent upon the presence of inorganic phosphate, as observed previously for E. coli RK and mammalian adenosine kinases (AK). A number of activators and inhibitors of human AK, produced very similar effects on the human and E. coli RKs, indicating that the catalytic mechanism of RK is very similar to that of the AKs.

Identification and biochemical studies on novel non-nucleoside inhibitors of the enzyme adenosine kinase.

The enzyme adenosine kinase (AK) plays a key role in the regulation of intracellular and extracellular concentration of adenosine (Ado), which exhibits potent hormonal activity in cardiovascular, nervous and immune systems. In view of the pharmacological effects of Ado, there is much interest in identifying inhibitors of AK, which can augment its tissue-protective effects. In this study, we have screened 1040 compounds from a chemical library of putative kinase inhibitors for their effect on purified human recombinant AK. These studies have identified 8 novel, non-nucleoside AK inhibitors. Four of these compounds (viz. 2-tert-butyl-4H-benzo[1,2,4]thiadiazine-3-thione (2759-0749); N-(5,6-diphenyl-furo[2,3-d]pyrimidin-4-yl)-propionamide (3998-0118); 3-[5,6-Bis-(4-methoxy-phenyl)-furo[2,3-d]pyrimidin-4-ylamino]-propan-1-ol (4072-2732); and 2-[2-(3,4-dihydroxy-phenyl)-5-phenyl-1H-imidazol-4-yl]-fluoren-9-one (8008-6198)), which inhibited human AK in a concentration-dependent manner in a low micromolar range (IC(50) = 0.38 approximately 1.98 microM) were further studied. Kinetic and structural studies on these compounds provide evidence that inhibition of AK by these compounds was competitive with respect to Ado and non-competitive for ATP. All of these compounds also inhibited uptake of Ado and its metabolism in cultured mammalian cells at comparable concentrations indicating their efficient cellular penetrability. These AK inhibitors, whose chemical structures differ significantly from all previously known inhibitors, provide useful lead compounds for identification of more potent but less toxic AK inhibitors that may prove useful for therapeutic purposes.

Subcellular localization of fumarase in mammalian cells and tissues.

Fumarase, a mitochondrial matrix protein, is previously indicated to be present in substantial amounts in the cytosol as well. However, recent studies show that newly synthesized human fumarase is efficiently imported into mitochondria with no detectable amount in the cytosol. To clarify its subcellular localization, the subcellular distribution of fumarase in mammalian cells/tissues was examined by a number of different methods. Cell fractionation using either a mitochondria fraction kit or extraction with low concentrations of digitonin, detected no fumarase in a 100,000 g supernatant fraction. Immunofluorescence labeling with an affinity-purified antibody to fumarase and an antibody to the mitochondrial Hsp60 protein showed identical labeling pattern with labeling seen mainly in mitochondria. Detailed studies were performed using high-resolution immunogold electron microscopy to determine the subcellular localization of fumarase in rat tissues, embedded in LR White resin. In thin sections from kidney, liver, heart, adrenal gland and anterior pituitary, strong and specific labeling due to fumarase antibody was only detected in mitochondria. However, in the pancreatic acinar cells, in addition to mitochondria, highly significant labeling was also observed in the zymogen granules and endoplasmic reticulum. The observed labeling in all cases was completely abolished upon omission of the primary antibody indicating that it was specific. In a western blot of purified zymogen granules, a fumarase-antibody cross-reactive protein of the same molecular mass as seen in the mitochondria was present. These results provide evidence that fumarase in mammalian cells/tissues is mainly localized in mitochondria and significant amounts of this protein are not present in the cytosol. However, these studies also reveal that in certain tissues, in addition to mitochondria, this protein is also present at specific extramitochondrial sites. Although the cellular function of fumarase at these extramitochondrial locations is not known, the appearance/localization of fumarase outside mitochondria may help explain how mutations in this mitochondrial protein can give rise to a number of different types of cancers.

Mitochondrial import of human and yeast fumarase in live mammalian cells: retrograde translocation of the yeast enzyme is mainly caused by its poor targeting sequence.

Studies on yeast fumarase provide the main evidence for dual localization of a protein in mitochondria and cytosol by means of retrograde translocation. We have examined the subcellular targeting of yeast and human fumarase in live cells to identify factors responsible for this. The cDNAs for mature yeast or human fumarase were fused to the gene for enhanced green fluorescent protein (eGFP) and they contained, at their N-terminus, a mitochondrial targeting sequence (MTS) derived from either yeast fumarase, human fumarase, or cytochrome c oxidase subunit VIII (COX) protein. Two nuclear localization sequences (2x NLS) were also added to these constructs to facilitate detection of any cytosolic protein by its targeting to nucleus. In Cos-1 cells transfected with these constructs, human fumarase with either the native or COX MTSs was detected exclusively in mitochondria in >98% of the cells, while the remainder 1-2% of the cells showed varying amounts of nuclear labeling. In contrast, when human fumarase was fused to the yeast MTS, >50% of the cells showed nuclear labeling. Similar studies with yeast fumarase showed that with its native MTS, nuclear labeling was seen in 80-85% of the cells, but upon fusion to either human or COX MTS, nuclear labeling was observed in only 10-15% of the cells. These results provide evidence that extramitochondrial presence of yeast fumarase is mainly caused by the poor mitochondrial targeting characteristics of its MTS (but also affected by its primary sequence), and that the retrograde translocation mechanism does not play a significant role in the extramitochondrial presence of mammalian fumarase.

Inhibition of adenosine kinase by phosphonate and bisphosphonate derivatives.

The enzyme adenosine kinase (AK) plays a central role in regulating the intracellular and interstitial concentration of the purine nucleoside adenosine (Ado). In view of the beneficial effects of Ado in protecting tissues from ischemia and other stresses, there is much interest in developing AK inhibitors, which can regulate Ado concentration in a site- and event-specific manner. The catalytic activity of AK from different sources is dependent upon the presence of activators such as phosphate (Pi). In this work we describe several new phosphorylated compounds which either activate or inhibit AK. The compounds acetyl phosphate, carbamoyl phosphate, dihydroxyacetone phosphate and imidodiphosphate were found to stimulate AK activity in a dose-dependent manner comparable to that seen with Pi. In contrast, a number of phosphonate and bisphosphonate derivatives, which included clodronate and etidronate, were found to inhibit the activity of purified AK in the presence of Pi. These AK inhibitors (viz. clodronate, etidronate, phosphonoacetic acid, 2-carboxyethylphosphonic acid, N-(phosphonomethyl)-glycine and N-(phosphonomethyl)iminodiacetic acid), at concentrations at which they inhibited AK, were also shown to inhibit the uptake of (3)H-adenosine and its incorporation into macromolecules in cultured mammalian cells, indicating that they were also inhibiting AK in intact cells. The drug concentrations at which these effects were observed showed limited toxicity to the cultured cells, indicating that these effects are not caused by cellular toxicity. These results indicate that the enzyme AK provides an additional cellular target for the clinically widely used bisphosphonates and related compounds, which could possibly be exploited for a new therapeutic application. Our structure-activity studies on different AK activators and inhibitors also indicate that all of the AK activating compounds have a higher partial positive charge (delta(+)) on the central phosphorous atom in comparison to the inhibitors. This information should prove helpful in the design and synthesis of more potent inhibitors of AK.

Localization of mitochondrial DNA encoded cytochrome c oxidase subunits I and II in rat pancreatic zymogen granules and pituitary growth hormone granules.

Cytochrome c oxidase (COX) complex is an integral part of the electron transport chain. Three subunits of this complex (COX I, COX II and COX III) are encoded by mitochondrial (mit-) DNA. High-resolution immunogold electron microscopy has been used to study the subcellular localization of COX I and COX II in rat tissue sections, embedded in LR Gold resin, using monoclonal antibodies for these proteins. Immunofluorescence labeling of BS-C-1 monkey kidney cells with these antibodies showed characteristic mitochondrial labeling. In immunogold labeling studies, the COX I and COX II antibodies showed strong and specific mitochondrial labeling in the liver, kidney, heart and pancreas. However, in rat pancreatic acinar tissue, in addition to mitochondrial labeling, strong and specific labeling was also observed in the zymogen granules (ZGs). In the anterior pituitary, strong labeling with these antibodies was seen in the growth hormone secretory granules. In contrast to these compartments, the COX I or COX II antibodies showed only minimal labeling (five- to tenfold lower) of the cytoplasm, endoplasmic reticulum and the nucleus. Strong labeling with the COX I or COX II antibodies was also observed in highly purified ZGs from bovine pancreas. The observed labeling, in all cases, was completely abolished upon omission of the primary antibodies. These results provide evidence that, similar to a number of other recently studied mit-proteins, COX I and COX II are also present outside the mitochondria. The presence of mit-DNA encoded COX I and COX II in extramitochondrial compartments, provides strong evidence that proteins can exit, or are exported, from the mitochondria. Although the mechanisms responsible for protein exit/export remain to be elucidated, these results raise fundamental questions concerning the roles of mitochondria and mitochondrial proteins in diverse cellular processes in different compartments.

Phosphorylated derivatives that activate or inhibit mammalian adenosine kinase provide insights into the role of pentavalent ions in AK catalysis.

The enzyme adenosine kinase (AK) exhibits a nearly complete dependency on the presence of pentavalent ions (PVI) such as phosphate, arsenate, and vanadate. To understand its basis, the effect of a large number of phosphorylated compounds on AK activity was examined. Several compounds, such as phosphoribosyl pyrophosphate, phosphoenol pyruvate, creatine phosphate, phosphorous acid, phosphonoformic acid, and inorganic pyrophosphate, were found to substitute for PVI in stimulating AK activity. Similar to PVI, these compounds lowered the Km of AK for adenosine. In contrast, many other structurally related compounds (i.e., phosphonoacetic acid, 2-carboxyethyl phosphonic acid, N-phosphonomethyl glycine, N-phosphonomethyl iminodiacetic acid) inhibited AK activity. These compounds seemed to compete with the activators for binding to AK. Structural comparisons of different compounds indicate that all activating compounds contain a net positive charge on the pentavalent atom (e.g., phosphorous), which should enable it to act as an acceptor for a nucleophilic group. We suggest that a phosphate (or other activator) bound near the active site participates in AK catalysis by forming a transient pentavalent intermediate with a nonbridging oxygen of the beta-phosphate in ATP. This interaction likely facilitates the transfer of gamma-phosphate from ATP to adenosine, thus accounting for the stimulating role of PVI in AK catalysis. The insight provided by these studies concerning the structural features of activators and inhibitors should also prove helpful in the design of more potent inhibitors of AK.

Genomic organization and linkage via a bidirectional promoter of the AP-3 (adaptor protein-3) mu3A and AK (adenosine kinase) genes: deletion mutants of AK in Chinese hamster cells extend into the AP-3 mu3A gene.

The cDNA and genomic DNA for the mu3A subunit of the AP-3 (adaptor protein-3) complex were cloned from Chinese hamster cells. The AP-3 mu3A genes in Chinese hamster, human and mouse each comprise nine exons and eight introns, with all introns located in identical positions in the species studied. The AP-3 mu3A genes in these species are linked in a head-to-head fashion with the gene for the purine salvage pathway enzyme AK (adenosine kinase). These genes share the first exon, and a 512 bp fragment covering the intervening untranslated sequence has the characteristic of a CpG island promoter, and it effectively carried out transcription in both directions. Deletion studies indicate that this region contains both positive and negative regulatory elements affecting transcription of these genes. In comparison with the AP-3 mu3A gene (27 kb), the AK gene in human is very large (558 kb), with average exon and intron lengths of approx. 100 bp and 55.7 kb respectively. The ratio of non-coding to coding sequence in the human AK gene is >550, which is the highest reported for any gene. We also present evidence that a number of AK- mutants of Chinese hamster ovary cells contain large deletions that affect both of these genes. In addition to lacking part of the AK gene, two of these mutants also lacked all of the exons and introns corresponding to the AP-3 mu3A gene. These mutants should prove useful in elucidating the role of AP-3 mu3A in vesicle-mediated protein sorting--a process that is altered in Hermansky-Pudlak syndrome. Detailed phylogenetic analysis of the micro family of proteins presented here also provides insight into how different AP complexes are related and may have evolved.

Pentavalent ions dependency is a conserved property of adenosine kinase from diverse sources: identification of a novel motif implicated in phosphate and magnesium ion binding and substrate inhibition.

The catalytic activity of adenosine kinase (AK) from mammalian sources has previously been shown to exhibit a marked dependency upon the presence of pentavalent ions (PVI), such as phosphate (PO4), arsenate, or vanadate. We now show that the activity of AK from diverse sources, including plant, yeast, and protist species, is also markedly enhanced in the presence of PVI. In all cases, PO4 or other PVI exerted their effects primarily by decreasing the Km for adenosine and alleviating the inhibition caused by high concentrations of substrates. These results provide evidence that PVI dependency is a conserved property of AK and perhaps of the PfkB family of carbohydrate kinases which includes AK. On the basis of sequence alignments, we have identified a conserved motif NXXE within the PfkB family. The N and E of this motif make close contacts with Mg2+ and PO4 ions in the crystal structures of AK and bacterial ribokinase (another PfkB member which shows PVI dependency), implicating these residues in their binding. Site-directed mutagenesis of these residues in Chinese hamster AK have resulted in active proteins with greatly altered phosphate stimulation and substrate inhibition characteristics. The N239Q mutation leads to the formation of an active protein whose activity was not stimulated by PO4 or inhibited by high concentrations of adenosine or ATP. The activity of the E242D mutant protein was also not significantly altered in the presence of phosphate. Although PO4 had no effect on the KmAdenosine for this mutant, the KmATP, K(i)Adenosine, and K(i)ATP were significantly decreased. In contrast to these mutations, N239L or E242L mutant proteins showed greatly decreased activity with an altered Mg2+ requirement. These observations support the view that N239 and E242 play an important role in the binding of PO4 and Mg2+ ions required for the catalytic activity of adenosine kinase.