JBC Connect with Cosmo for Collagen Detection

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M203473200 on September 3, 2002

J. Biol. Chem., Vol. 277, Issue 45, 42654-42662, November 8, 2002
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
277/45/42654    most recent
M203473200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mahnke-Zizelman, D. K.
Right arrow Articles by Sabina, R. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mahnke-Zizelman, D. K.
Right arrow Articles by Sabina, R. L.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

N-terminal Sequence and Distal Histidine Residues Are Responsible for pH-regulated Cytoplasmic Membrane Binding of Human AMP Deaminase Isoform E*

Donna K. Mahnke-Zizelman and Richard L. SabinaDagger

From the Department of Biochemistry, Medical College of Wisconsin, Milwaukee, Wisconsin 53226

Received for publication, April 10, 2002, and in revised form, August 18, 2002

    ABSTRACT
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES

Mammalian AMP deaminase 3 (AMPD3) enzymes reportedly bind to intracellular membranes, plasma lipid vesicles, and artificial lipid bilayers with associated alterations in enzyme conformation and function. However, proteolytic sensitivity of AMPD polypeptides makes it likely that prior studies were performed with N-truncated enzymes. This study uses erythrocyte ghosts to characterize the reversible cytoplasmic membrane association of human full-sized recombinant isoform E (AMPD3). Membrane-bound isoform E exhibits diminished catalytic activity whereas low micromolar concentrations of the cationic antibiotic, neomycin, disrupt this protein-lipid interaction and relieve catalytic inhibition. The cytoplasmic membrane association of isoform E also displays an inverse correlation with pH in the physiological range. Diethyl pyrocarbonate (DEPC) modification of isoform E nearly abolishes its cytoplasmic membrane binding capacity, and this effect can be reversed by hydroxylamine. Difference spectra reveal that 18 of 29 histidine residues in each isoform E subunit are N-carbethoxylated by DEPC. These combined data demonstrate that protonated imidazole rings of histidine residues mediate a pH-responsive association of isoform E with anionic charges on the surface of the cytoplasmic membrane, possibly phosphatidylinositol 4,5-bisphosphate, a pure noncompetitive inhibitor of the enzyme. Finally, AMPD1 and a series of N-truncated AMPD3 enzymes are used to show that these behaviors are specific to isoform E and require up to 48 N-terminal amino acids, even though this stretch of sequence contains no histidine residues. The pH-responsive cytosol-membrane partitioning of isoform E may be an important mechanism for branch point regulation of adenylate catabolism.

    INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES

AMP deaminase (AMPD1; EC 3.5.4.6) is a highly regulated enzyme catalyzing a branch point reaction in the adenylate catabolic pathway, and its expression in mammalian tissues and cells is characterized by a multigene family (1-5). In humans, the AMPD1 gene encodes isoform M (6), AMPD2 encodes isoform L (7), and AMPD3 encodes isoform E (8, 9). Primary amino acid sequence alignments identify divergent N-terminal and conserved C-terminal domains across human AMPD isoforms (7, 8). In addition, all three human AMPD genes produce multiple transcripts that encode additional variation at or near, the N terminus of each isoform (8, 10, 11).

Human AMPD isoforms have been purified and characterized from endogenous sources (12-15). However, extreme N-terminal regions in mammalian AMPD polypeptides are highly sensitive to proteolysis during purification and subsequent storage of the enzyme at 4 °C (16-19). Recombinant technology provides a means to overexpress AMPD enzymes that can be purified with subunits predominantly intact, although these are also subject to proteolysis during storage at 4 °C (20). The availability of purified AMPD enzymes with subunits predominantly intact has stimulated interest in the structural and functional significance of extreme N-terminal sequences that were likely missing from previously characterized enzyme preparations.

Recent studies have shown that extreme N-terminal sequences only subtly influence the catalytic and regulatory properties of human AMPD isoforms (20, 21). Conversely, divergent N-terminal residues may play an important role in the intracellular distribution of AMPD as evidenced by their effects on contractile protein binding behavior (20, 21). Although a more C-terminal conserved contractile protein binding domain has been identified in all AMPD isoforms, a stretch of sequence in the unique N-terminal region of the AMPD1 polypeptide is required for the high actomyosin binding capacity of isoform M (21). This observation is functionally significant because contractile protein binding is an important physiological regulator of catalytic activity in stimulated skeletal muscle (22, 23), the primary site of AMPD1 expression. Although more widely distributed across human tissues and cells (15), the highest level of AMPD3 expression is also observed in skeletal muscle (3), where isoform E appears confined predominantly to type I fibers (24). However, up to 48 amino acids in the unique N terminus of isoform E dramatically suppress contractile protein binding capacity of this enzyme (21), a behavior that could facilitate other intracellular interactions.

Consistent with this hypothesis, available information indicates that mammalian AMPD3 enzymes can reversibly associate with the cytoplasmic membrane. This was initially revealed by the observation that human erythrocyte membrane preparations contained AMPD activity (25). Subsequent work using purified human erythrocyte AMPD (isoform E) demonstrated an association with the cytoplasmic face of erythrocyte ghost (EG) membranes that was also accompanied by reduced catalytic activity (26). Moreover, membrane association and the related inhibition of catalytic activity could be reversed in the presence of small molecule effectors, including substrate (26). More recently, a series of studies has shown that interactions between purified pig heart AMPD, the porcine ortholog of human isoform E (27), and isolated cytoplasmic membrane vesicles and artificial lipid bilayers (28-30) alters secondary structure and regulatory behavior of the enzyme.

Knowledge of the physical basis for these protein-lipid interactions and the effects they have on catalytic behavior of AMPD3 enzymes have been advanced by two recent observations (31). 1) Phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P2), an integral membrane phospholipid, is a potent noncompetitive inhibitor of human recombinant isoform E and its primary proteolytic product, Delta M90. 2) Incubation of rat PC12 cells with taurine causes an increase in total phosphoinositide concentrations that is accompanied by a higher percentage of membrane-associated AMPD activity.

However, as detailed above, it is likely that much of the available information on protein-lipid interactions involving AMPD has been obtained with proteolyzed AMPD3 enzymes. This study was designed to characterize the functional and structural basis for the cytoplasmic membrane association of human full-sized recombinant AMPD isoform E.

    EXPERIMENTAL PROCEDURES
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES

Materials-- Grace's insect cell medium and fetal calf serum were purchased from Invitrogen. Phosphocellulose (P11) was obtained from Whatman, Ltd. Disposable glass columns and a Protein Assay kit were available through Bio-Rad. Outdated human blood was provided by The Blood Center of Southeastern Wisconsin. Neomycin sulfate was purchased from CalBiochem NovaBiochem Corp. Goat anti-rabbit IgG was obtained from Santa Cruz Biotechnology, Inc. All other chemicals and reagents were of the highest qualities commercially available.

Expression and Purification of Recombinant AMPD3 Enzymes-- Human AMPD3 recombinant (wild type 1b and all N-truncated) enzymes and an N-truncated AMPD1 (Delta M54) enzyme were expressed in Sf9 (Spodoptera frugiperda) cells using baculoviral technology as previously described (20, 21). All N-truncated constructs were produced by oligonucleotide-directed mutagenesis using wild type cDNAs (AMPD1[exon 2+] and AMPD3[1b]) as templates. In some cases, the ATG triplet contained within an NcoI restriction site (CCATGG; naturally occurring or created) corresponded to a methionine start codon in the wild type mRNA (Delta M54AMPD1 and N-truncated AMPD3 enzymes Delta M90 and Delta M127). In other cases, introduction of the ATG triplet in the mutant cDNA resulted in a methionine substitution at the N terminus of the encoded truncated polypeptide (AMPD3 enzymes Delta L20M, Delta I49M, and Delta E65M). Recombinant proteins were purified from clarified sonicates of infected cells by phosphocellulose chromatography with sequential linear salt (0.1-2.0 M potassium chloride) and phosphate (0.02-0.45 M potassium phosphate) gradient elution followed by ammonium sulfate (30% w/v) precipitation as previously described (21). The specific activities of the purified AMPD3 enzymes ranged from 550 to 2000 units/mg protein, whereas that of the Delta M54AMPD1 enzyme was 5000 units/mg (21). Purified enzymes were suspended in 50 mM imidazole, pH 6.5 (AMPD1) or 7.0 (AMPD3) containing 500 mM KCl and stored at -80 °C in 50% (v/v) glycerol.

Preparation of Unsealed Erythrocyte Ghosts-- Unsealed erythrocyte ghosts (EG) were prepared after the method of Steck and Kant (32). Briefly, 1 ml of outdated human blood was diluted 1:6 with ice-cold phosphate-buffered saline (PBS; 5 mM sodium phosphate, pH 8.0, 150 mM sodium chloride) and pelleted at 1877 × g for 10 min at 4 °C in a refrigerated tabletop centrifuge. Cell pellets were washed three times with ice-cold PBS to remove hemolyzed cells. Intact erythrocytes were lysed by rapidly and thoroughly resuspending the cells in 40 ml of ice-cold 5 mM sodium phosphate, pH 8.0 (5P8), then centrifuging at 22,000 × g for 10 min at 4 °C. Ghosts were washed four times with ice-cold 5P8 to remove soluble cellular contents, then resuspended in 1 ml of ice-cold 5P8, and stored at 4 °C until use. Residual AMPD activities in unsealed EG suspensions were below the level of detection. Total protein per unit volume of resuspended ghosts was determined by Bio-Rad assay.

Preparation of Sealed, Right-Side-Out Erythrocyte Ghosts-- Sealed, right-side-out EG were prepared from unsealed EG after the method of Steck and Kant (32). Briefly, 500 µl of unsealed EG were resuspended in 20 ml of PBS and incubated 40 min at 37 °C to induce resealing. Resealed ghosts were then pelleted at 22,000 × g for 10 min at 4 °C, washed twice more in PBS and stored at 4 °C until use. A parallel aliquot of unsealed ghosts was maintained in 5P8 and treated similarly, then used as a control for subsequent comparative binding studies with increasing amounts of enzyme (see below). Residual AMPD activities in these sealed and unsealed EG suspensions were also below the level of detection. Total protein per unit volume of resuspended ghosts was determined by Bio-Rad assay.

Association of AMPD Enzymes with Erythrocyte Ghosts-- 100-µl mixtures of AMPD enzyme and EG, the latter resuspended in binding buffer (5 mM potassium phosphate, pH 7.0, 45 mM potassium chloride, 0.1 mg/ml bovine serum albumin), were incubated on ice (see text for amounts and times of incubation). Partitioning of AMPD enzymes between the supernatant and pellet was then evaluated as follows: mixtures were centrifuged at 14,000 × g for 10 min at 4 °C. The supernatant was recovered and residual AMPD quantitated by enzyme assay and Western blot analysis (see below). EG pellets were resuspended in 100 µl of binding buffer and AMPD was quantitated as described above. Other experimental conditions are described in the text.

Chemical Modification with Diethyl Pyrocarbonate and Hydroxylamine (HA)-- AMPD enzymes were dialyzed overnight against 4000 volumes of 100 mM potassium phosphate, pH 7.0 containing 100 mM potassium chloride and 1 mM dithiothreitol in order to reduce the imidazole concentration in stored preparations. Stock solutions of diethyl pyrocarbonate (DEPC) were freshly prepared in absolute ethanol immediately before use. DEPC modification reactions were performed in an 82-µl mixture comprised of ~4.5 units of enzyme diluted 1:1 in dialysis buffer. Reactions were initiated by the addition of DEPC to a final concentration of 250 µM, incubated for 120 min on ice, then quenched by the addition of imidazole to a final concentration of 2.4 mM. As a control, an equivalent volume of absolute ethanol (minus DEPC) was added to a parallel mixture and treated identically. Ethanol concentrations were less than 2.5% (v/v). Aliquots of treated and untreated enzyme were then incubated with EG and association was evaluated as described above. Hydroxylamine hydrochloride (adjusted to pH 7.0) was then added to the remainder of each quenched mixture (DEPC-treated and untreated) to a final concentration of 150 mM. These mixtures were then incubated for an additional 60 min at room temperature. Aliquots were then removed, and EG association was evaluated as described above. Preliminary experiments were performed in which the time of incubation and concentrations of chemical modifiers were varied in order to determine optimal conditions.

Difference Spectroscopy-- DEPC modification of histidine residues was analyzed by recording difference spectra using a Shimadzu dual beam spectrophotometer. The formation of N-carbethoxyimidazole derivatives of histidine can then be quantitated by measuring an increase in absorbance at 240 nm using an extinction coefficient of 3200 M-1 cm-1 (33). The modification reaction included E1b enzyme (7.6 µM subunit concentration) suspended in 100 mM potassium phosphate, pH 7.0 containing 100 mM potassium chloride and was initiated by the addition of DEPC to a final concentration of 1 mM. This was referenced against a parallel suspension of enzyme containing an equivalent volume of added ethanol (minus DEPC). Aliquots were removed at appropriate time intervals and quenched in 10 mM imidazole so that EG association could also be evaluated as described above.

Sf9 Insect Cell Membrane Preparation-- Sf9 insect cell membranes were prepared according to the method of Sarkadi et al. (34) with some modification. Briefly, six confluent T-185 flasks were infected with human AMPD3 wild type recombinant baculovirus for 4 days at 27 °C. Infected cells were collected and pelleted at 365 × g for 6 min in a refrigerated tabletop centrifuge, then suspended in ice-cold TMEP buffer (50 mM Tris-HCl, pH 7.0, containing 50 mM mannitol, 2 mM EGTA, 10 µg/ml leupeptin, 20 µg/ml E-64, and 0.1% beta -mercaptoethanol) supplemented with 150 mM potassium chloride. Cell suspensions were lysed using 50 strokes of a hand-held glass-Teflon homogenizer on ice. Undisrupted cells and nuclear debris were removed by centrifugation at 500 × g for 10 min in a refrigerated tabletop centrifuge. The supernatant fluid was then divided into two aliquots and centrifuged for 60 min at 100,000 × g. Pellets containing the membranes were resuspended either in ice-cold TMEP buffer alone or in that supplemented with 150 mM potassium chloride at a protein concentration of 3-5 mg/ml. Membrane suspensions were stored at -80 °C until further use.

Enzyme Assays-- AMP deaminase activity was measured in 100-µl reactions (including 30 µl of supernatant or resuspended EG pellet) containing 25 mM imidazole, pH 7.0, 100 mM potassium chloride, 20 µg of bovine serum albumin and 20 mM AMP. This saturating substrate concentration is also sufficient to disrupt the EG association of AMPD3 enzymes (see below). Substrate and product were resolved and quantitated by anion-exchange HPLC as previously described (7, 35).

Glyceraldehyde-3-phosphate dehydrogenase was chosen as a marker presumed to reside on the cytoplasmic side of the erythrocyte membrane, and its activity was determined by the method of Steck and Kant (32). Briefly, 10 µg of sealed or unsealed EG protein was added at room temperature to 1-ml reaction mixtures comprised of freshly prepared 24.6 mM sodium pyrophosphate, pH 8.4 containing 3.28 mM cysteine HCl, 12 mM sodium arsenate, 1 mM beta -NAD+, and 1.5 mM glyceraldehyde-3-phosphate. Activity was monitored at 340 nm in a spectrophotometer that was blanked against identical mixtures without erythrocyte ghost protein added. The reactions were linear with time up to 3 min.

Western Blot Analysis-- Proteins were fractionated by 9% SDS-PAGE, electroblotted onto nitrocellulose membranes and probed with rabbit polyclonal antisera raised against human AMPD recombinant enzymes, as previously described (20).

Computer-assisted Statistical Analysis-- Instat program software was used to generate data means and S.D. and to perform two-tailed Student's t tests.

    RESULTS
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES

Association of AMPD3 Enzymes with Erythrocyte Ghosts-- Preliminary experiments were performed to characterize the association of isoform E (E1b) and Delta M90 enzymes with EG. E1b is one of three identified AMPD3 spliceoforms, and the Delta M90 variant is modeled after the primary N-terminal proteolytic product generated from this protein during purification and extended storage at 4 °C (20). Delta M90 was typically included in most experiments to simulate proteolyzed endogenous enzyme likely used in a previous study that examined the membrane association of human erythrocyte AMPD (26). Fig. 1 presents the results of enzyme assay and Western blot analysis in which increasing amounts of E1b and Delta M90 enzymes were incubated either with unsealed or sealed, right-side out ghosts. Both enzymes exhibit a substantially greater association with the unsealed preparation suggesting that interactions occur primarily with intracellular components of the ghost structure. Glyceraldehyde-3-phosphate dehydrogenase activity, a marker for the cytoplasmic side of the membrane, was also measured in both ghost preparations. The activity of this enzyme was measurable only in the unsealed ghost suspension (unsealed, 0.096 ± 0.012 Delta 340 nm/min/µg ghost protein (n = 3); sealed, <0.003 Delta 340 nm/min/µg ghost protein). All subsequent experiments were performed using unsealed erythrocyte ghosts.


View larger version (27K):
[in this window]
[in a new window]
  		Fig. 1.   Association of AMPD3 enzymes with unsealed and sealed, right-side-out EG. Mixtures of unsealed (U) and sealed (S) EG (10 µg of EG protein) were incubated with increasing amounts of human AMPD3 wild type (E1b) or N-truncated (Delta M90) enzymes on ice for 30 min, then partitioned by centrifugation at 14,000 × g for 10 min at 4 °C. AMPD was quantitated in resuspended EG pellets by enzyme assay (A) and Western blot analysis (B).

Enzyme assay and Western blot data also reveal that EG associations of the E1b and Delta M90 proteins are saturable with respect to the time of incubation and amounts of EG and enzyme added to the mixture (data not shown). This information was used to standardize subsequent experiments that unless otherwise stated employed 30-min incubations with 10 µg of EG protein and less than saturating amounts of AMPD3 enzymes (0.1-1.2 units; see text and figure legends for specific amounts).

Substrate-induced Dissociation of AMPD3 Enzymes from Erythrocyte Ghosts-- Previous work has shown that EG association of purified human erythrocyte AMPD results in catalytic inhibition of the enzyme, although substrate concentrations greater than 100 µM can disrupt this interaction (26). Therefore, it was necessary to determine a substrate concentration where the EG dissociation of the E1b and Delta M90 proteins was essentially complete in order that enzyme assay could be used as a reliable quantitative index following their partitioning into EG pellets. Parallel ghost resuspensions with associated enzymes were incubated with increasing amounts of AMP and partitioning was re-evaluated. Enzyme assay and Western blot analysis were used to show that the interaction between EG and either AMPD3 protein was sensitive to increasing substrate concentration and essentially eliminated in the presence of 5 mM AMP (data not shown). This information confirmed that a saturating level of substrate (20 mM) was adequate to enable enzyme assay to be used as a reliable quantitative indicator of AMPD3 protein association with EG pellets.

Neomycin-induced Dissociation of AMPD3 Enzymes from Erythrocyte Ghosts-- Neomycin was evaluated as an antagonist of the associations between AMPD3 enzymes and EG under the premise that these interactions may be related to their affinity for PtdIns(4,5)P2, an integral membrane phospholipid that is also a potent noncompetitive inhibitor of catalytic activity (31). Neomycin is an aminoglycoside antibiotic containing six primary amine groups that bind to PtdIns(4,5)P2 with micromolar affinity (36-39). Increasing amounts of neomycin were added to parallel resuspensions of ghosts containing associated AMPD3 enzyme, and partitioning was immediately re-evaluated at each concentration of the drug. As shown in Fig. 2 by both enzyme assay and Western blot analysis, low micromolar concentrations of neomycin produce dramatic EG dissociation of the E1b protein, with ~60 and 85% of the protein released at 5 and 50 µM, respectively. Conversely, EG association of the Delta M90 enzyme is relatively unaffected, and only about 20% of this protein is released by drug concentrations as high as 500 µM. The observed EG dissociation of this N-truncated activity is likely due to the reported ability of neomycin to act as an anion-exchanger at high (mM) concentrations (40).


View larger version (36K):
[in this window]
[in a new window]
  		Fig. 2.   Neomycin-induced EG dissociation of E1b and a series of N-truncated (19-126 residues) enzymes. Mixtures of E1b and N-truncated enzymes (1.2 units of each) and unsealed EG (40 µg of EG protein) were incubated on ice for 30 min, then partitioned as described in the legend to Fig. 1. Resuspended EG pellets were aliquoted (10 µg of EG protein each), then immediately re-partitioned following the addition of increasing concentrations of neomycin (0-500 µM). Re-partitioned EG pellets were resuspended and AMPD was quantitated by enzyme assay (A) and Western blot analysis (B). Graphical data represent the mean ± S.D. from independent experiments (E1b, n = 5; Delta L20M, n = 4; all other enzymes, n = 3). *, p < 0.05 when either E1b or Delta L20M were compared with other AMPD3 enzymes in a two-tailed Student's t test. +, p < 0.05 when E1b was compared with Delta L20M in a two-tailed Student's t test. Dagger , p < 0.05 when Delta I49M was compared with Delta M127 in a two-tailed Student's t test. All other comparisons were not significant.

Parallel analyses were performed using several other recombinant enzymes with progressive N-terminal deletions. As also illustrated in Fig. 2, an activity lacking 19 N-terminal residues, Delta L20M, behaves in a fashion similar to the E1b protein, although it is not as profoundly affected. Conversely, three other AMPD3 enzymes with truncations of 49-126 amino acids mimic the behavior of the Delta M90 protein and are relatively resistant to EG dissociation by low micromolar concentrations of neomycin.

Effect of Membrane Association on AMPD3 Enzyme Activity-- Catalytic activities of membrane-associated AMPD3 enzymes were examined following the resuspension of parallel ghost pellets containing bound E1b or Delta M90 proteins in the presence and absence of 62.5 µM neomycin. Partitioning was re-evaluated in each of these mixtures following exposure to increasing amounts of substrate (90-454 µM AMP). Aliquots of each supernatant were frozen for subsequent direct injection onto the HPLC column as a means of evaluating relative catalytic activity in each mixture during the 180-min incubation period. Fig. 3 reveals that the catalytic activity of the E1b enzyme is diminished (upper graph) when it remains associated with EG in the absence of neomycin (lower graph). The addition of this drug results in EG dissociation (middle graph) and a simultaneous release of catalytic inhibition (upper graph). Similarly, the Delta M90 enzyme also displays lower catalytic activity (upper graph) while it is associated with EG in the absence of neomycin (lower graph). However, catalytic inhibition of the Delta M90 enzyme is also relieved in the presence of neomycin (upper graph) even though this enzyme maintains its EG association (middle graph). A parallel experiment performed in the absence of erythrocyte ghosts shows that neomycin alone has no effect on the catalytic activity of either enzyme (data shown in Fig. 3, upper graphs).


View larger version (43K):
[in this window]
[in a new window]
  		Fig. 3.   Effect of membrane association on the catalytic behavior of E1b and Delta M90 enzymes. Parallel mixtures of enzymes (100 milliunits) and unsealed EG (10 µg of EG protein) were incubated on ice for 180 min with and without neomycin (62.5 µM) in the presence of increasing concentrations of AMP (90-454 µM). Each mixture was partitioned as described in the legend to Fig. 1. Upper graphs, an aliquot of each supernatant was frozen for subsequent injection onto the HPLC column to determine relative catalytic activity during the 180-min incubation. Middle and lower graphs, AMPD partitioning into supernatants (unbound) and resuspended pellets (bound) was quantitated by enzyme assay. As a control, parallel mixtures of each enzyme were incubated and partitioned as described above, but in the absence of EG. An aliquot of each supernatant was then frozen for subsequent analysis of relative catalytic activity as described above (data included in upper graphs).

Effect of pH on the Membrane Association of AMPD3 Enzymes-- Based on a prediction for ionic interactions between positively charged residues of isoform E and negatively charged phosphate moieties of PtdIns(4,5)P2, pH was also evaluated as an effector of the association between AMPD3 enzymes and EG membranes. Parallel mixtures of EG and E1b or the Delta M90 enzyme were incubated in binding buffers adjusted to pH 6.5, 7.0, and 7.5. As shown in Fig. 4 by both enzyme assay and Western blot analysis, pH has a dramatic effect on partitioning of the E1b enzyme. Relative to neutral pH, there is an increase in membrane association under acidotic conditions (pH 6.5) and diminished binding at alkaline pH (pH 7.5). Conversely, changes in pH have little effect on the EG association of the Delta M90 enzyme.


View larger version (51K):
[in this window]
[in a new window]
  		Fig. 4.   Effect of pH on membrane association of E1b and a series of N-truncated (19-126 residues) enzymes. Parallel mixtures of enzymes (0.5 units of each) and unsealed EG (10 µg of EG protein) were incubated in binding buffers of different pH on ice for 30 min, then partitioned as described in the legend to Fig. 1. AMPD was quantitated in supernatants (UNBOUND) and resuspended pellets (BOUND) by enzyme assay (graphs) and Western blot analysis (panels). Graphical data represent the mean ± S.D. for three independent experiments. *, p < 0.05 when compared with pH 6.5 in a two-tailed Student's t test. +, p < 0.05 when pH 7.0 and pH 7.5 were compared in a two-tailed Student's t test. All other comparisons were not significant.

Again using the series of deletion constructs, the region of N-terminal sequence responsible for the observed pH effect on membrane association was refined. Similar to that observed with neomycin-induced release, the Delta L20M enzyme behaves more like isoform E. Conversely, deletions of 48-126 N-terminal amino acids result in enzymes exhibiting membrane associations that are relatively unaffected by changes in pH.

In order to demonstrate that pH-induced effects on membrane association of isoform E are reversible, E1b enzyme partitioning was evaluated prior to, during, and following acidotic conditions in the same EG suspensions. Data presented in Fig. 5 illustrate that membrane association of the E1b enzyme is enhanced during acidosis, then the protein re-distributes toward steady-state conditions following a return to neutral pH.


View larger version (65K):
[in this window]
[in a new window]
  		Fig. 5.   Reversible membrane association of isoform E in response to changes in pH. A 300-µl mixture of E1b enzyme (1.0 unit) and unsealed EG (30 µg of EG protein) was incubated at pH 7.0 on ice for 30 min, then a 100-µl aliquot was removed and partitioned as described in the legend to Fig. 1. The remainder of the mixture was adjusted to pH 6.5 with hydrochloric acid, incubated on ice for 5 min, and a second 100-µl aliquot was removed and partitioned. Finally, the remainder of the mixture was adjusted back to pH 7.0 with potassium hydroxide, incubated on ice for 5 min, and partitioned. AMPD was quantitated in all supernatants and resuspended EG pellets by enzyme assay (upper graph) and Western blot analysis (lower panel). Graphical data represent the mean ± S.D. from three independent experiments. *, p < 0.05 when pH 7.0(first) was compared with pH 6.5 in a two-tailed Student's t test. +, p < 0.05 when pH 6.5 was compared with pH 7.0(second) in a two-tailed Student's t test. All other comparisons were not significant. B, bound; UB, unbound.

Effect of Diethyl Pyrocarbonate and Hydroxylamine Modification on the Membrane Association of AMPD Enzymes-- The pH sensitivity of membrane association exhibited by isoform E suggested that histidine residues could play a role in this behavior of the enzyme. DEPC N-carbethoxylates the imidazole ring of histidine with sufficient specificity under appropriate conditions to be useful in defining the role of these residues in biochemical interactions. This correlation is strengthened if the observed effect can be reversed by subsequent exposure to HA, which can remove the N-carbethoxyl group from a modified histidine residue. E1b and Delta M90 AMPD3 enzymes were treated with DEPC and analyzed for EG membrane association by enzyme assay and Western blot analysis. An AMPD1 enzyme, Delta M54, was also included in this experiment after determining that it can associate with EG membranes (data not shown). This N-truncated version of the two AMPD1 (isoform M) splice variants was used because it retains the high contractile protein binding capacity of the wild type enzymes but is not prone to aggregation (21), thus making it more suitable for evaluation of isoform specificity in this experiment. As shown in Fig. 6, DEPC modification significantly reduces the membrane binding capacity of both AMPD3 enzymes, although quantitative differences are apparent. The membrane association of the E1b enzyme is virtually eliminated, whereas the effect on Delta M90 is only modest. Conversely, DEPC modification does not diminish the EG association of the Delta M54AMPD1 enzyme. It is also evident from the controls that pH has little effect on the membrane association of this AMPD1 enzyme. Exposure to DEPC reduced the catalytic activities of all enzymes following a 120-min exposure to this modifying agent (percent of control when assayed at pH 6.5 and 7.0, respectively: E1b, 34.7 ± 12.2 and 42.9 ± 9.8; Delta M90, 31.3 ± 1.2 and 31.7 ± 5.2; Delta M54AMPD1, 61.0 ± 9.1 and 65.3 ± 8.0).


View larger version (60K):
[in this window]
[in a new window]
  		Fig. 6.   Effect of DEPC and hydroxylamine modification on membrane binding of AMPD enzymes. AMPD enzymes (7-20 units) were reacted with 250 µM DEPC on ice for 120 min, then quenched by the addition of imidazole (2.4 mM final concentration). Parallel mixtures incubated in the presence of an equivalent volume of absolute ethanol used to prepare fresh stock concentrations of DEPC served as controls. Aliquots of each enzyme (0.3 units) were then incubated with unsealed ghosts (10 µg of EG protein) at pH 6.5 or 7.0 on ice for 30 min and then partitioned as described in the legend to Fig. 1. The remainder of each AMPD3 enzyme mixture was then reacted with 150 mM hydroxylamine (HA) for 60 min at room temperature. Aliquots (0.3 units) were then incubated with unsealed EG (10 µg of EG protein) at pH 6.5 or 7.0 on ice for 30 min, and then partitioned as described in the legend to Fig. 1. AMPD was quantitated in all supernatants and resuspended EG pellets by enzyme assay (reported as Percent Bound, left panels) and Western blot analysis (right panels). Graphical data represent the mean ± S.D. from three independent experiments. *, p < 0.05 when control was compared with DEPC-treated in a two-tailed Student's t test. +, indicates p < 0.05 when control was compared with DEPC/HA-treated in a two-tailed Student's t test. B, bound; UB, unbound.

Subsequent exposure to HA restores the EG binding capacities of E1b and Delta M90 enzymes (Fig. 6), strongly suggesting that the observed effects of DEPC modification on the membrane associations of these two AMPD3 enzymes is due to N-carbethoxylation of histidine residues.

Quantitation of N-Carbethoxylated Histidine Residues in Isoform E during Modification with DEPC-- Difference spectra were recorded between 230 nm and 300 nm during reactions of the E1b enzyme with DEPC. As illustrated in Fig. 7A, these spectra show a time-dependent increase in absorbance between 230 and 250 nm with observed maxima between 234 and 236 nm. In addition, there is no apparent valley in the spectrum at 278 nm that would indicate O-carbethoxylation of tyrosine residues (41). The number of N-carbethoxylated histidine residues were calculated based on the difference at 240 nm using an extinction coefficient of 3200 M-1 cm-1 (33). As shown in Fig. 7B, the modification of up to 18 histidine residues in each isoform E subunit is accompanied by the time-dependent loss of its membrane binding capacity.


View larger version (26K):
[in this window]
[in a new window]
  		Fig. 7.   UV absorbance difference spectra during a DEPC modification reaction and the stoichiometry of histidine residue N-carbethoxylation. E1b enzyme (subunit concentration, 7.6 µM) was incubated at 4 °C with 1 mM DEPC in 100 mM potassium phosphate, pH 7.0 containing 100 mM potassium chloride. Difference spectra were recorded every 5 min from 1 to 116 min. A, difference spectra from (bottom to top) 1, 16, 31, 46, and 61 min reaction times. B, stoichiometry of DEPC modification and effect on membrane binding capacity of the E1b enzyme (reported as Percent Bound). N-carbethoxylated histidine residues were quantitated using an extinction coefficient of 3200 M-1 cm-1 at 240 nm (33). Subunit molecular mass of the E1b enzyme is 88.8 kDa.

Effect of Ionic Strength on the Membrane Association of Isoform E-- In order to maintain the integrity of the unsealed ghost preparations, all experiments designed to examine the structural basis and catalytic effect of the protein-lipid interaction between isoform E and the cytoplasmic membrane were performed under hypotonic conditions (45 mM potassium chloride). However, the ionic nature of the cytoplasmic membrane association of isoform E suggests that this interaction may be less robust at physiological salt concentrations. Therefore, additional binding experiments were performed to assess the cytoplasmic membrane association of isoform E in the presence of 150 mM potassium chloride. Data presented in Fig. 8A show that the relative EG membrane association of isoform E is lower in the presence of 150 mM potassium chloride. Fig. 8B illustrates that simulated physiological acidosis causes a 2-fold enhancement in this pH-sensitive, reversible interaction similar to that observed under hypotonic conditions (see data presented in Fig. 5).


View larger version (24K):
[in this window]
[in a new window]
  		Fig. 8.   Effect of ionic strength on the reversible membrane association of isoform E. A, parallel mixtures of E1b enzyme (0.3 units) and unsealed EG (10 µg of EG protein) were incubated in binding buffers containing either 45 or 150 mM potassium chloride at different pH on ice for 30 min and then partitioned as described in the legend to Fig. 1. B, 300-µl mixture of E1b enzyme (1.0 unit) and unsealed EG (30 µg of EG protein) was incubated in binding buffer containing 150 mM potassium chloride at pH 7.2 on ice for 30 min, and then a 100-µl aliquot was removed and partitioned as described in the legend to Fig. 1. The remainder of the mixture was adjusted to pH 6.2 with hydrochloric acid, incubated on ice for 5 min, and a second 100-µl aliquot was removed and partitioned. Finally, the remainder of the mixture was adjusted back to pH 7.2 with potassium hydroxide, incubated on ice for 5 min, and partitioned. AMPD was quantitated in resuspended EG pellets by enzyme assay (upper graph) and Western blot analysis (lower panel). Graphical data represent the mean ± S.D. from three independent experiments. *, p < 0.05 when pH 6.2 was compared with either pH 7.2 condition in a two-tailed Student's t test.

Association of Recombinant Isoform E with Insect Cell Membranes-- Sf9 insect cells infected with human AMPD3 wild type recombinant baculovirus were analyzed as a means to evaluate the ability of isoform E to interact with membranes in situ. Light membrane fractions, normally prepared in a low strength ionic buffer (34), were isolated in the presence of 150 mM potassium chloride. Enzyme assay and Western blot data presented in Fig. 9 show that recombinant isoform E co-purifies with insect cell membranes isolated in the presence of this physiological concentration of salt. Similar to that observed with erythrocyte ghosts, the interaction between isoform E and insect cell membranes is sensitive to ionic strength. This is apparent when the resuspended preparation is subsequently partitioned under hypotonic and normal salt concentrations.


View larger version (29K):
[in this window]
[in a new window]
  		Fig. 9.   Recombinant isoform E co-purifies with Sf9 insect cell membranes. Membranes were prepared in the presence of 150 mM potassium chloride from insect cells infected with human AMPD3 wild type recombinant baculovirus and resuspended either in TMP buffer alone (which contains no salt) or that supplemented with 150 mM potassium chloride. AMPD was quantitated in suspensions and following partitioning by centrifugation at 14,000 × g for 5 min. Upper, enzyme activity in suspensions (Total) and in membrane-cleared supernates (Unbound). Lower, Western blot analysis: lane 1, purified isoform E; lane 2, total; lane 3, partitioned pellets; lane 4, partitioned supernatants. Graphical data represent the mean ± S.D. from three independent determinations. *, p < 0.05 when unbound was compared between the no salt and 150 mM KCl conditions in a two-tailed Student's t test.


    DISCUSSION
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES

Mammalian AMPD expression is characterized by a multigene family that encodes three parental isoforms with divergent N-terminal and conserved C-terminal domains. Although AMPD is located at a highly regulated branch point in the adenylate catabolic pathway, the functional significance of multiple isoforms is not well understood. Data presented in this study advance this issue by demonstrating that the human AMPD3 enzyme, isoform E, can reversibly bind to the cytoplasmic membrane with an accompanying inhibition of catalytic activity and that the interaction is responsive to changes of pH in the physiological range. The structural basis of this protein-lipid interaction is an electrostatic attraction between anionic charges on the cytoplasmic membrane and histidine residues in isoform E that also requires up to 48 N-terminal amino acids of the AMPD3 polypeptide. These combined observations reveal a previously unrecognized regulatory mechanism for isoform E that, together with the established protein-protein interaction between isoform M (AMPD1) and elements of the contractile apparatus, suggests a functional and spatial separation of multiple AMPD activities expressed within the same cell.

Several features exhibited by the cytoplasmic membrane association of isoform E reflect the electrostatic interaction involving histidine residues. For example, membrane binding of isoform E exhibits an inverse relationship with pH in the physiological range of 6.1 to 7.5. Furthermore, exposure to DEPC essentially eliminates the membrane binding capacity of isoform E. In addition, hydroxylamine reverses the effect that DEPC has on membrane binding capacity of isoform E. Finally, UV absorbance difference spectra collected during modification of isoform E with DEPC demonstrate an increase in absorbance between 230 and 250 nm characteristic of N-carbethoxylation of histidine and the lack of a valley at 278 nm that would indicate O-carbethoxylation of tyrosine residues. These combined data establish a role for histidine residues in the reversible cytoplasmic membrane association of isoform E. The imidazole rings of this amino acid are more likely to be protonated under mildly acidotic conditions and would promote membrane association (see Fig. 10). Conversely, the enzyme would be more soluble under steady-state conditions and mild alkalosis due to fewer protonated histidine residues.


View larger version (22K):
[in this window]
[in a new window]
  		Fig. 10.   Model for adenylate catabolic branchpoint regulation by the reversible cytoplasmic membrane association of isoform E. The rapid on-off cytoplasmic membrane association of isoform E serves as a switch that can modulate catabolic flow from AMP. Metabolic acidosis (top) favors histidine residue protonation that can promote the membrane recruitment of isoform E, where it is simultaneously anchored and inhibited by PtdIns(4,5)P2. This situation leads to more adenosine production by AMP-preferring cN-I. Conversely, isoform E is predominantly cytosolic under steady-state conditions (bottom) that favor histidine residue deprotonation. This situation leads to enhanced production of IMP. Although the depicted regulatory scheme involves changes in intracellular pH, other effectors of the membrane partitioning of isoform E may contribute, as well as factors that control the activity of cN-I. For simplicity, only one subunit of the AMPD3 tetramer is shown.

The identity of the anionic attractant on the cytoplasmic membrane for protonated histidine residues in isoform E is less certain. Negative charge potentials of intracellular membranes are due, in part, to the presence of a variety of anionic phospholipids. Among these, PtdIns(4,5)P2 is a component of the cytoplasmic membrane that can interact with isoform E. PtdIns(4,5)P2 is a pure noncompetitive inhibitor of this enzyme with a Ki of 110 nM (31), which may also explain the associated catalytic inhibition of membrane-bound isoform E. While other membrane-derived sources of phosphate groups could bind and inhibit this enzyme, low micromolar concentrations of neomycin are able to simultaneously relieve catalytic inhibition and disrupt the protein-lipid interaction involving isoform E. Neomycin is an aminoglycoside antibiotic with six primary amine groups that binds strongly to PtdIns(4,5)P2 in intracellular membranes (36, 38, 39). Although neomycin does not distinguish between PtdIns(4,5)P2 and PtdIns(3,4)P2 (39), the former is the major bisphosphate inositol in mammalian cells and therefore a more likely in vivo target for isoform E.

Ionic interactions are frequently involved in lipid binding of non-integral membrane proteins. For example, the hydrophilic faces of amphipathic alpha -helices can interact with anionic patches on the surface of membranes (42, 43). Acidic phospholipid and phosphoinositide-specific binding domains have also been identified and among these are pleckstrin homology (PH) domains found in a number of signal transduction proteins (44). These ~100 amino acid modules have diverse primary amino acid sequences but similar tertiary structures. A positively charged surface within the PH domain has been shown to represent the ligand binding site to phosphoinositides (45). In fact, the C-terminal region of isoform E has two putative PH domains (31), and these stretches of sequence contain 11 of the 29 histidine residues found in the AMPD3 polypeptide. Data presented in this study have shown that N-carbethoxylation of as many as 18 histidine residues in each isoform E subunit is accompanied by the near elimination of its membrane binding capacity. Regardless of whether those primarily responsible for the reversible membrane association of isoform E are located in these PH domains, to our knowledge a pH-responsive protein-lipid interaction involving histidine residues has not previously been described as a mechanism for the membrane association of a non-integral protein.

Extreme N-terminal sequence is critical to functional and structural features of the interaction between histidine residues in isoform E and the cytoplasmic membrane. This was unexpected in light of the fact that all 29 histidine residues in the AMPD3 polypeptide are C-terminal to amino acid 170 (8). However, unlike isoform E, a series of N-truncated AMPD3 enzymes lacking 48-126 amino acids all exhibit an interaction with the cytoplasmic membrane that is mostly unresponsive to pH or low micromolar concentrations of neomycin. Furthermore, DEPC modification essentially eliminates the membrane binding capacity of isoform E, yet only modestly affects this behavior of the Delta M90 N-truncated enzyme. These combined data strongly suggest that N-truncated enzymes participate in additional protein-lipid interactions that are somehow suppressed by the presence of up to 48 N-terminal amino acids in the AMPD3 polypeptide. How N-terminal sequence achieves this apparent suppression of additional membrane interactions is unknown, but this same stretch of amino acids has a similar influence on the contractile protein binding capacity of isoform E (21). Consequently, the functional significance of AMPD3 N-terminal sequence is made apparent through the roles that this stretch of amino acids have in preventing interactions with elements of the contractile apparatus and promoting the pH-responsive cytoplasmic membrane association of isoform E.

The combined results of this study reveal an elegant yet simplistic regulatory mechanism for isoform E that may be important to branch point adenylate catabolism. Fig. 10 presents a model that relates documented and proposed structural information to functional features of this regulation. The modulation of AMPD catalytic activity that accompanies a rapid on-off cytoplasmic membrane binding of isoform E should impact competition between AMP deaminase and cytosolic 5'-nucleotidase I (cN-I). During periods of metabolic acidosis, for example, diminished AMPD catalytic activity would serve to augment adenosine production by cN-I. Such a mechanism may serve to facilitate the many physiological roles of adenosine as well as explain why ischemic heart predominantly dephosphorylates AMP even though reported AMP deaminase activities are higher than cN-I in soluble extracts prepared from normoxic tissue (46-48).

Relative AMPD1 and AMPD3 expression is highest in adult skeletal muscle (3, 7), although fiber-type differences have been noted, i.e. AMPD1 levels are high in all fibers but more so in glycolytic muscle, whereas AMPD3 is primarily restricted to oxidative fibers (24, 49, 50). The AMPD1 isoform exhibits a high contractile protein binding capacity that is, in part, attributable to N-terminal sequence (20, 21). The myofibrillar recruitment of this enzyme during intense skeletal muscle contractions precedes the detectable accumulation of IMP (22), suggesting that catalytic activation accompanies contractile protein binding. Intense skeletal muscle contractions also produce an acidotic environment in skeletal myocytes (51), but the predicted increase in cytoplasmic membrane binding of isoform E under these conditions would promote catalytic inhibition of this AMPD activity. Such spatial and functional divergence in AMPD isoform behavior under these physiological conditions could explain why AMPD3 expression is unable to compensate for the deficiency of myoadenylate deaminase (AMPD1), a common inherited disorder of skeletal muscle energy metabolism. These individuals accumulate little, if any IMP during strenuous exercise (52-54).

In conclusion, this study has presented data that suggests additional functional significance for divergent N-terminal sequences across human AMPD isoforms. N-terminal amino acids in the AMPD3 polypeptide are proposed to control the intracellular distribution of isoform E by simultaneously suppressing contractile protein binding (20, 21) and promoting its association with the cytoplasmic membrane. These dual effects may be particularly relevant to striated muscle where both interactions are possible. Moreover, multiple AMPD genes are expressed in these tissues (3, 7, 8), where mixed tetramers comprised of different ratios of isoform-specific subunits appear to represent a substantial portion of enzyme activity (15). Both gene-specific subunits can exert influence on the behaviors of these hybrid enzymes, as suggested by the intermediate contractile protein binding capacities exhibited by human AMPD1/AMPD3 recombinant tetramers compared with those of isoforms M and E (20). Consequently, the contractile apparatus and the cytoplasmic membrane may compete for the recruitment of these hybrid enzymes. A comprehensive understanding of the contractile protein and membrane binding behaviors of all mixed tetramers containing AMPD3 subunits seems warranted and should provide critical information for the interpretation of planned studies designed to examine in vivo protein-protein and protein-lipid interactions involving AMPD in striated muscle. In a broader sense, the spatial separation of AMPD catalytic activities in tissues and cells should contribute to localized adenylate catabolic regulation and may provide justification for multiple isoforms of this enzyme.

    Acknowledgment

Outdated blood was provided by the Blood Center of Southeast Wisconsin

    FOOTNOTES

* This work was supported by Public Health Service Grant DK-50902 from the National Institutes of Health.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Dagger To whom correspondence should be addressed: Dept. of Biochemistry, 8701 Watertown Plank Rd., Milwaukee, WI 53226. Tel.: 414-456-4697; Fax: 414-456-6510; E-mail: sabinar@mcw.edu.

Published, JBC Papers in Press, September 3, 2002, DOI 10.1074/jbc.M203473200

    ABBREVIATIONS

The abbreviations used are: AMPD, AMP deaminase; EG, erythrocyte ghost; PtdIns(4, 5)P2, phosphatidylinositol 4,5-bisphosphate; Sf9, Spodoptera frugiperda; PBS, phosphate-buffered saline; DEPC, diethyl pyrocarbonate; HA, hydroxylamine; PH, pleckstrin homology; cN-I, cytosolic 5'-nucleotidase I.

    REFERENCES
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES

1. Sabina, R. L., Morisaki, T., Clarke, P., Eddy, R., Shows, T. B., Morton, C. C., and Holmes, E. W. (1990) J. Biol. Chem. 265, 9423-9433[Abstract/Free Full Text]
2. Morisaki, T., Sabina, R. L., and Holmes, E. W. (1990) J. Biol. Chem. 265, 11482-11486[Abstract/Free Full Text]
3. Mahnke-Zizelman, D. K., Eddy, R., Shows, T. B., and Sabina, R. L. (1996) Biochim. Biophys. Acta 1306, 75-92[Medline] [Order article via Infotrieve]
4. Mahnke-Zizelman, D. K., Van den Bergh, F., Bausch-Jurken, M. T., Eddy, R., Sait, S., Shows, T. B., and Sabina, R. L. (1996) Biochim. Biophys. Acta 1308, 122-132[Medline] [Order article via Infotrieve]
5. Sermsuvitayawong, K., Wang, X., Nagabukuro, A., Matsuda, Y., Morisaki, H., Toyama, K., Mukai, T., and Morisaki, T. (1997) Mamm. Genome 8, 767-769[CrossRef][Medline] [Order article via Infotrieve]
6. Sabina, R. L., Fishbein, W. N., Pezeshkpour, G., Clarke, P. R. H., and Holmes, E. W. (1992) Neurology 42, 170-177[Abstract/Free Full Text]
7. Bausch-Jurken, M. T., Mahnke-Zizelman, D. K., Morisaki, T., and Sabina, R. L. (1992) J. Biol. Chem. 267, 22407-22413[Abstract/Free Full Text]
8. Mahnke-Zizelman, D. K., and Sabina, R. L. (1992) J. Biol. Chem. 267, 20866-20877[Abstract/Free Full Text]
9. Yamada, Y., Goto, H., and Ogasawara, N. (1992) Biochim. Biophys. Acta 1171, 125-128[Medline] [Order article via Infotrieve]
10. Mineo, I., Clarke, P. R. H., Sabina, R. L., and Holmes, E. W. (1990) Mol. Cell. Biol. 10, 5271-5278[Abstract/Free Full Text]
11. Van den Bergh, F., and Sabina, R. L. (1995) Biochem. J. 312, 401-410[Medline] [Order article via Infotrieve]
12. Yun, S-L., and Suelter, C. H. (1978) J. Biol. Chem. 253, 404-408[Abstract/Free Full Text]
13. Stankiewicz, A. (1981) Int. J. Biochem. 13, 1177-1183[CrossRef][Medline] [Order article via Infotrieve]
14. Ashby, B., and Holmsen, H. (1981) J. Biol. Chem. 256, 10519-10523[Abstract/Free Full Text]
15. Ogasawara, N., Goto, H., Yamada, Y., Watanabe, T., and Asano, T. (1982) Biochim. Biophys. Acta 714, 298-306[Medline] [Order article via Infotrieve]
16. Ranieri-Raggi, M., and Raggi, A. (1980) Biochem. J. 189, 367-368[Medline] [Order article via Infotrieve]
17. Marquetant, R., Sabina, R. L., and Holmes, E. W. (1989) Biochemistry 28, 8744-8749[CrossRef][Medline] [Order article via Infotrieve]
18. Chilson, O. P., Kelly-Chilson, A. E., and Siegel, N. R. (1997) Comp. Biochem. Physiol. 116B, 371-377
19. Knecht, K., Wiesmuller, K.-H., Gnau, V., Jung, G., Meyermann, R., Todd, K. G., and Hamprecht, B. (2001) J. Neurosci. Res. 66, 941-950[CrossRef][Medline] [Order article via Infotrieve]
20. Mahnke-Zizelman, D. K., Tullson, P. C., and Sabina, R. L. (1998) J. Biol. Chem. 273, 35118-35125[Abstract/Free Full Text]
21. Mahnke-Zizelman, D. K., and Sabina, R. L. (2001) Biochem. Biophys. Res. Comm. 285, 489-495[CrossRef][Medline] [Order article via Infotrieve]
22. Rundell, K. W., Tullson, P. C., and Terjung, R. L. (1992) Am. J. Physiol. 263, C287-C293[Medline] [Order article via Infotrieve]
23. Rundell, K. W., Tullson, P. C., and Terjung, R. L. (1992) Am. J. Physiol. 263, C294-C299[Medline] [Order article via Infotrieve]
24. Kuppevelt, T. H. v., Veerkamp, J. H., Fishbein, W. N., Ogasawara, N., and Sabina, R. L. (1994) J. Histochem. Cytochem. 42, 861-868[Abstract]
25. Askari, A. (1963) Science 141, 44-45[Abstract/Free Full Text]
26. Pipoly, G. M., Nathans, G. R., Chang, D., and Deuel, T. F. (1979) J. Clin. Invest. 63, 1066-1076[Medline] [Order article via Infotrieve]
27. Hohl, A. M., and Hohl, C. M. (1999) Mol. Cell. Biochem. 201, 151-158[CrossRef][Medline] [Order article via Infotrieve]
28. Wozniak, M., Kossowska, E., Purzycka-Preis, J., and Zydowo, M. M. (1988) Biochem. J. 255, 977-981[Medline] [Order article via Infotrieve]
29. Tanfani, F., Kossowska, E., Purzycka-Preis, J., Zydowo, M. M., Wozniak, M., Tartaglini, E., and Bertoli, E. (1993) Biochem. J. 291, 921-926[Medline] [Order article via Infotrieve]
30. Tanfani, F., Kulawiak, D., Kossowska, E., Purzycka-Preis, J., Zydowo, M. M., Sarkissova, E., Bertoli, E., and Wozniak, M. (1998) Mol. Gen. Metab. (1998) 65, 51-58[CrossRef][Medline] [Order article via Infotrieve]
31. Sims, B., Mahnke-Zizelman, D. K., Profit, A. A., Prestwich, G. D., Sabina, R. L., and Theibert, A. B. (1999) J. Biol. Chem. 274, 25701-25707[Abstract/Free Full Text]
32. Steck, T. L., and Kant, J. A. (1974) Methods Enzymol. 31, 172-180[CrossRef][Medline] [Order article via Infotrieve]
33. Ovadi, J., Libor, S., and Elodi, P. (1967) Biochim. Biophys. Acta 2, 455-458
34. Sarkadi, B., Price, E. M., Boucher, R. C., Germann, U. A., and Scarborough, G. A. (1992) J. Biol. Chem. 267, 4854-4858[Abstract/Free Full Text]
35. Swain, J. L., Sabina, R. L., McHale, P. A., Greenfield, J. C., and Holmes, E. W. (1982) Am. J. Physiol. 242, H818-H826[Medline] [Order article via Infotrieve]
36. Schacht, J. (1976) J. Neurochem. 27, 1119-1124[CrossRef][Medline] [Order article via Infotrieve]
37. Schacht, J. (1978) J, Lipid Res. 19, 1063-1067[Abstract]
38. Gabev, E., Kasianowicz, J., Abbott, T., and McLaughlin, S. (1989) Biochim. Biophys. Acta 979, 105-112[Medline] [Order article via Infotrieve]
39. Arbuzova, A., Martushova, K., Hangyas-Mihalyne, G., Morris, A. J., Ozaki, S., Prestwich, G. D., and Mclaughlin, S. (2000) Biochim. Biophys. Acta 1464, 35-48[Medline] [Order article via Infotrieve]
40. Prentki, M., Deeney, J., Matschinsky, F. M., and Joseph, S. K. (1986) FEBS Lett. 197, 285-288[CrossRef][Medline] [Order article via Infotrieve]
41. Burstein, Y., Walsh, K. A., and Neurath, H. (1974) Biochemistry 13, 205-210[CrossRef][Medline] [Order article via Infotrieve]
42. Kalmar, G. B., Kay, R. J., Lachance, A., Aebersold, R., and Cornell, R. B. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 6029-6033[Abstract/Free Full T