Regulation of transcription by a protein methyltransferase.

The p160 family of coactivators, SRC-1, GRIP1/TIF2, and p/CIP, mediate transcriptional activation by nuclear hormone receptors. Coactivator-associated arginine methyltransferase 1 (CARM1), a previously unidentified protein that binds to the carboxyl-terminal region of p160 coactivators, enhanced transcriptional activation by nuclear receptors, but only when GRIP1 or SRC-1a was coexpressed. Thus, CARM1 functions as a secondary coactivator through its association with p160 coactivators. CARM1 can methylate histone H3 in vitro, and a mutation in the putative S-adenosylmethionine binding domain of CARM1 substantially reduced both methyltransferase and coactivator activities. Thus, coactivator-mediated methylation of proteins in the transcription machinery may contribute to transcriptional regulation.

Hormone-dependent, CARM1-directed, arginine-specific methylation of histone H3 on a steroid-regulated promoter.

Activation of gene transcription involves chromatin remodeling by coactivator proteins that are recruited by DNA-bound transcription factors. Local modification of chromatin structure at specific gene promoters by ATP-dependent processes and by posttranslational modifications of histone N-terminal tails provides access to RNA polymerase II and its accompanying transcription initiation complex. While the roles of lysine acetylation, serine phosphorylation, and lysine methylation of histones in chromatin remodeling are beginning to emerge, low levels of arginine methylation of histones have only recently been documented, and its physiological role is unknown. The coactivator CARM1 methylates histone H3 at Arg17 and Arg26 in vitro and cooperates synergistically with p160-type coactivators (e.g., GRIP1, SRC-1, ACTR) and coactivators with histone acetyltransferase activity (e.g., p300, CBP) to enhance gene activation by steroid and nuclear hormone receptors (NR) in transient transfection assays. In the current study, CARM1 cooperated with GRIP1 to enhance steroid hormone-dependent activation of stably integrated mouse mammary tumor virus (MMTV) promoters, and this coactivator function required the methyltransferase activity of CARM1. Chromatin immunoprecipitation assays and immunofluorescence studies indicated that CARM1 and the CARM1-methylated form of histone H3 specifically associated with a large tandem array of MMTV promoters in a hormone-dependent manner. Thus, arginine-specific histone methylation by CARM1 is an important part of the transcriptional activation process.

In vitro aging of calmodulin generates isoaspartate at multiple Asn-Gly and Asp-Gly sites in calcium-binding domains II, III, and IV.

We have determined the major sites responsible for isoaspartate formation during in vitro aging of bovine brain calmodulin under mild conditions. Protein L-isoaspartyl methyltransferase (EC 2.1.1.77) was used to quantify isoaspartate by the transfer of methyl-3H from S-adenosyl-L-[methyl-3H]methionine to the isoaspartyl (alpha-carboxyl) side chain. More than 1.2 mol of methyl-acceptor sites per mol of calmodulin accumulated during a 2-week incubation without calcium at pH 7.4, 37 degrees C. Analysis of proteolytic peptides of aged calmodulin revealed that > 95% of the methylation capacity is restricted to residues in the four calcium-binding domains, which are predicted to be highly flexible in the absence of calcium. We estimate that domains III, IV, and II accumulated 0.72, 0.60, and 0.13 mol of isoaspartate per mol of calmodulin, respectively. The Asn-97-Gly-98 sequence (domain III) is the greatest contributor to isoaspartate formation. Other major sites of isoaspartate formation are Asp-131-Gly-132 and Asp-133-Gly-134 in domain IV, and Asn-60-Gly-61 in domain II. Significant isoaspartate formation was also localized to Asp-20, Asp-22, and/or Asp-24 in domain I, to Asp-56 and/or Asp-58 in domain II, and to Asp-93 and/or Asp-95 in domain III. All of these residues are calcium ligands in the highly conserved EF-hand calcium-binding motif. Thus, other EF-hand proteins may also be subject to isoaspartate formation at these ligands. The results support the idea that isoaspartate formation in structured proteins is strongly influenced by both the C-flanking residue and by local flexibility.

Protein L-isoaspartyl methyltransferase in postmortem brains of aged humans.

The specific activity of protein L-isoaspartyl methyltransferase, an enzyme implicated in the metabolism of damaged, isoaspartate-containing proteins, has been measured in postmortem samples of parietal cortex from 30 individuals (19 with Alzheimer's disease and 11 controls). Methyltransferase specific activity was positively correlated with age at death, increasing by 2.9 pmol/min/mg of protein for every ten years of age (r = .51, p less than 0.005). This correlation was significant in the control and Alzheimer's disease groups alike. Specific activity also appeared to be about 15% higher in females than in age- and diagnosis-matched males (p less than 0.05). No significant differences were observed between age- and sex-matched Alzheimer patients and controls, suggesting that a deficiency in this enzyme is not responsible for the accumulation of abnormal proteins in Alzheimer's disease.

The effect of urea exposure on isoaspartyl content and protein L-isoaspartate methyltransferase activity in Drosophila melanogaster.

Urea is a protein unfolding agent that can accumulate to locally high concentrations in tissues of many organisms. We used Drosophila melanogaster to test the hypothesis that urea loading would promote formation of isoaspartate (beta-carboxyl-linked aspartate), a common form of protein damage that occurs most readily in unstructured polypeptides and flexible regions of folded proteins. Ten populations of flies were tested; five control populations of urea-sensitive flies and five previously selected urea-tolerant populations. We measured the effects of urea consumption on levels of both isoaspartate and protein L-isoaspartate methyltransferase (PIMT), an enzyme believed to function in the repair or removal of isoaspartyl proteins. For both sets of populations, urea feeding for 6 days increased isoaspartyl levels by approximately 60%, supporting the idea that disruption of protein secondary and tertiary structures can accelerate the formation of isoaspartate in vivo. Urea feeding tended to increase PIMT activity in both control and urea-tolerant populations. There were no significant differences in PIMT activities or isoaspartyl levels between the control and urea-tolerant flies raised on normal or urea food. The latter findings indicate that urea tolerance evolved in the selected populations without any significant change in PIMT expression or activity.

Isoaspartate in peptides and proteins: formation, significance, and analysis.

Formation of isoaspartyl peptide bonds (isoAsp) is one of the most common forms of non-enzymatic degradation of peptides and proteins under mild conditions. IsoAsp arises when certain Asn-Xaa and Asp-Xaa sites undergo a spontaneous intramolecular rearrangement to form a succinimide which subsequently hydrolyzes to generate a mixture of isoAsp-Xaa and Asp-Xaa linkages in a ratio of approximately 2:1. This pathway is responsible for the much greater susceptibility of asparagine, compared with glutamine, to deamidation at neutral and alkaline pH. Rearrangement occurs most readily at Asn-Gly, Asn-Ser, and Asp-Gly sequences where the local polypeptide chain flexibility is high. Formation of isoAsp can decrease the biological activity of a protein pharmaceutical, alter its susceptibility to proteolytic degradation, and elicit autoimmunity. The enzyme protein L-isoaspartyl methyltransferase can be used to measure isoAsp sites in the low pmol range with or without the use of radioisotopes.

Biosynthesis and processing of presumed neurosecretory proteins in single identified neurons of Aplysia californica.

The biosynthesis and processing of low molecular weight protein (presumed neurosecretory protein) in cells R15, R14 and L11 of Aplysia californica was studied at high resolution by polyacrylamide slab gel electrophoresis in sodium dodecylsulfate. The number of low molecular weight proteins detected in each cell ranges from 3 in R14 and L11 to 5 to 6 in R15. In each of the cells studied, the low molecular weight protein consists of a primary precursor of ca. 12,000 daltons, and its proteolytic processing products. In each cell, the smallest protein, or in the case of R14, one of the two smallest proteins, accumulates to a significant extent, suggesting that it might correspond to a final processed neurohormone. In cell R15, the biosynthesis of the primary precursor and its subsequent processing to smaller peptides is largely unaffected by removal of extracellular calcium, by replacement of calcium with cobalt or by inhibition of spontaneous bursting via stimulation of the brachial nerve.

Stoichiometric methylation of porcine adrenocorticotropin by protein carboxyl methyltransferase requires deamidation of asparagine 25. Evidence for methylation at the alpha-carboxyl group of atypical L-isoaspartyl residues.

The enzymatic methylation of porcine adrenocorticotropin (ACTH) in both its native form and a form which is deamidated at asparagine 25 has been compared using purified protein carboxyl methyltransferase from bovine brain. Incubation of deamidated ACTH with high concentrations of methyltransferase resulted in near stoichiometric levels of methyl incorporation (78 mol %), while the methylation of native ACTH was highly substoichiometric (3-12 mol %). The Km and Vmax for deamidated ACTH were 1.9 microM and 11,200 pmol/min/mg, respectively, making this peptide the most specific substrate known for the mammalian methyltransferase. Deamidation of asparagine 25 leads to the formation of an atypical isopeptide bond in which the resulting aspartyl residue is linked to the adjacent glycine 26 via its side-chain beta-carboxyl group rather than the usual alpha-carboxyl linkage (Gráf, L., Bajusz, S., Patthy A., Barát, E., and Cseh, G. (1971) Acta Biochim. Biophys. Acad. Sci. Hung. 6, 415-418; Bornstein, P., and Balian, G. (1977) Methods Enzymol. 47, 132-145). A synthetic isopeptide (beta-linked) analog of deamidated ACTH serves as a highly effective substrate for the methyltransferase, but the corresponding normal (alpha-linked) peptide does not, indicating that this enzyme selectively recognizes the alpha-carboxyl group of atypical beta-linked L-aspartyl residues (see also accompanying paper (Murray, E.D., Jr., and Clarke, S. (1984) J. Biol. Chem. 259, 10722-10732]. Methylation of atypical beta-linked L-aspartyl residues resulting from deamidation can account for previous observations that in vitro protein carboxyl methylation in mammalian systems almost always occurs with a low stoichiometry and that these protein methyl esters are considerably less stable than most chemically formed protein methyl esters.

Determination of D- and L-aspartate in amino acid mixtures by high-performance liquid chromatography after derivatization with a chiral adduct of o-phthaldialdehyde.

A sensitive and convenient method for the simultaneous determination of D- and L-aspartic acid in amino acid mixtures is described. The method involves derivatization of the mixture with a chiral fluorogen, followed by high-performance liquid chromatography on a reverse-phase column. The fluorogen used is an adduct of o-phthaldialdehyde with an optically active thiol, N-acetyl-L-cysteine. The sensitivity and accuracy of this method is similar to that using adducts of o-phthaldialdehyde with the achiral thiol, 2-mercaptoethanol. Five picomoles of D-aspartate can be accurately detected in the presence of a 100-fold excess of L-aspartate with a total analysis time (including derivatization) of 10 min.

Fragmentation of isoaspartyl peptides and proteins by carboxypeptidase Y: release of isoaspartyl dipeptides as a result of internal and external cleavage.

Isoaspartate-containing versions of sea urchin sperm-activating peptide, delta sleep-inducing peptide, and lactate dehydrogenase (231-242) were cleaved at internal sites by carboxypeptidase Y. Cleavage occurred between the isoaspartate and the preceding amino acid, and it was accompanied by sequential digestion of amino acids from the two resulting carboxyl termini. Because the isoaspartyl bonds were not cleaved, isoaspartyl dipeptides were among the final products. The rate of release of isoaspartyl dipeptides was different for the three peptides, a 24-h digestion yielding 0.32 mol of isoaspartylglycine/mol of isoaspartyl sperm-activating peptide, 0.50 mol of isoaspartylalanine/mol of isoaspartyl delta sleep-inducing peptide, and 1.15 mol of isoaspartylserine/mol of isoaspartyl lactate dehydrogenase (231-242). The different rates could be explained by the slow cleavage of amino acids preceded by glycine. Isoaspartyl dipeptides were not detected in digests of the corresponding aspartate- or asparagine-containing forms of the peptides. Release of isoaspartyl dipeptides by carboxypeptidase Y was used to demonstrate the presence of isoaspartylglycine sequences in deamidated adrenocorticotropin (0.54 mol/mol), in a mixture of trypic fragments of base-treated calmodulin (0.20 mol/mol), and in a mixture of tryptic fragments of base-treated triosephosphate isomerase (0.08 mol/mol). These results confirm earlier work suggesting that isoaspartylglycine formation is prevalent in proteins exposed to alkaline conditions. They also provide a methodology that should prove useful in the characterization of natural substrates for protein L-isoaspartyl methyltransferase.

Enzymatic protein carboxyl methylation at physiological pH: cyclic imide formation explains rapid methyl turnover.

At pH 7.4, 37 degrees C, bovine brain protein carboxyl methyltransferase transiently methylates deamidated adrenocorticotropin. The methylation occurs at the alpha-carboxyl group of an atypical beta-carboxyl-linked isoaspartyl residue (position 25). Several lines of evidence indicate that the immediate product of demethylation is an aspartyl cyclic imide involving positions 25 and 26. The evidence includes (1) the rapid rate of methyl ester hydrolysis, which is consistent with intramolecular catalysis, (2) the inability of the demethylated product to be remethylated, (3) the charge of this product, and (4) its rate of breakdown. The eventual hydrolysis of the cyclic imide produces a 30/70 mixture of peptides containing either alpha- or beta-carboxyl-linked aspartyl residues, respectively. Cyclic imide formation is nonenzymatic and can explain the unusual lability of mammalian protein methyl esters in general. These findings suggest that protein carboxyl methylation in mammalian tissues is not a simple on/off reversible modification as it apparently is in chemotactic bacteria. Carboxyl methylation may serve to activate selected protein carboxyl groups for subsequent longer lasting modifications, possibly subserving a role in protein repair, degradation, cross-linking, or some other as yet undiscovered alteration of protein structure.

Identification and topography of substrates for protein carboxyl methyltransferase in synaptic membrane and myelin-enriched fractions of bovine and rat brain.

The major components of crude brain synaptosomes (synaptic membranes, mitochondria, and myelin) have been separated and analyzed by polyacrylamide gel electrophoresis for the presence of proteins that serve as substrates for protein carboxyl methyltransferase. Of the three fractions, synaptic membranes contain the largest number of individual methyl acceptors (at least seven), while mitochondria contain no well-defined methyl acceptors. Undisrupted myelin contains a single major methyl acceptor with a very low apparent molecular weight. The patterns of protein methylation in synaptic membranes prepared from cerebral cortex, hippocampus, striatum, thalamus, and tectum showed marked differences; however, these differences could largely be explained by differential degrees of myelin contamination in synaptic membranes from the different regions. The effect of trypsin pretreatment on the carboxyl methylation of intact and lysed synaptosomes was studied to estimate the sidedness of the major methylation sites on synaptic membranes. One of the methyl acceptors (Mr 48K) appears to be facing the intracellular surface of the synaptosome, but most sites appear to be outward facing.

Identification of a 59-kDa Substrate for cAMP-Dependent Protein Kinase That Is Enriched in Rat Cerebellar Membranes.

Following incubation of rat brain membranes with the catalytic subunit of cAMP-dependent protein kinase and [(32)P]ATP, a previously unreported phosphoprotein, pp59, was found to be enriched in cerebellar synaptic plasma membrane preparations, but not in those prepared from cerebral cortex, hippocampus, olfactory bulb, or striatum. This protein, which has an M(r) of 59,000 by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, is not phosphorylated by the cGMP-dependent protein kinase. While pp59 was consistently detected in cerebellar membranes from adult Sprague-Dawley rats, it was not detected in bovine or rabbit cerebellar membranes. Moreover, pp59 did not comigrate with any of the autophosphorylated subunits of the Ca(2+)/calmodulindependent protein kinase in rat cerebellar membranes. Extraction of pp59 from these membranes could be accomplished with 6 M urea, but not with 0.4 M NaCl or 0.5% (v/v) Triton X-100. The urea solubility suggests that pp59 is not an integral membrane protein. Acid hydrolysis of the protein phosphorylated in vitro yielded phosphoserine but no significant amount of phosphothreonine or phosphotyrosine. Further analysis of pp59 may provide new insights into the role of cAMP in modulation of synaptic function in the cerebellum.

Deamidation and isoaspartate formation in proteins: unwanted alterations or surreptitious signals?

Formation of betalinked Asp-Xaa peptide bonds--isoaspartyl (isoAsp) sites--arise in proteins via succinimide-linked deamidation of asparagine or dehydration of aspartate, reactions which represent a major source of spontaneous protein damage under physiological conditions. Accumulation of atypical isoaspartyl sites is minimized in vivo by the activity of protein L-isoaspartyl O-methyltransferase (PIMT), which regenerates a normal peptide bond. Loss of PIMT has harmful consequences, especially in neurons; thus, formation of isoAsp sites and their subsequent correction by PIMT is widely believed to constitute an important pathway of protein damage and repair. Recent evidence is mounting, however, that deamidation and isoaspartate formation may, in some instances, constitute a novel mechanism for intentional modification of protein structure. Herein we describe the mechanism of Asx rearrangement, summarize the evidence that PIMT serves an important repair function, and then focus on emerging evidence that deamidation and isoAsp formation may sometimes have a useful function.

Diversity of methyl acceptor proteins in rat pheochromocytoma (PC12) cells revealed after treatment with adenosine dialdehyde.

Protein N-methylation is a widespread modification whose functions are poorly understood. To overcome the inherent technical difficulties in identification of N-methylated proteins, we cultured PC12 cells with a methylation inhibitor, in expectation that proteins would accumulate in a hypomethylated state. Cell extracts were then incubated with [methyl-3H]S-adenosyl-L-methionine to label methyl-accepting sites via endogenous methyltransferases. Using two-dimensional gel electrophoresis we detected over 50 methyl acceptors, ranging from 18 to 120 kDa. Most had isoelectric points greater than 7.0. NG,NG-Dimethylarginine and NG-monomethylarginine accounted for about 90% of the methyl-3H-amino acids recovered after acid hydrolysis and thin-layer chromatography. The production of hypomethylated proteins should prove useful, not only in the identification of new methyl acceptors, but also in the isolation and characterization of new methyltransferases.

Optimal conditions for the use of protein L-isoaspartyl methyltransferase in assessing the isoaspartate content of peptides and proteins.

Protein L-isoaspartyl methyltransferase provides a basis for enzymatic measurement of atypical, isoaspartyl linkages which make a major contribution to protein microheterogeneity. The low Vmax of the methyltransferase reaction and the instability of the methyl ester can hinder accurate determinations, and different laboratories using different conditions have achieved discrepant values for the isoaspartate content of the same proteins. To investigate the effects of these conditions, and to optimize the assay, isoaspartyl delta sleep-inducing peptide was methylated under a variety of conditions. We found that 1 microM methyltransferase was required to obtain stoichiometric modification of 2 microM peptide in 40-min reactions at pH 6.2 and 30 degrees C. A computer model utilizing kinetic constants obtained from studies on initial rates of methylation predicted the same requirement for enzyme concentration. Carrier protein was necessary for optimal methyltransferase activity at enzyme concentrations below 0.4 microM. Stoichiometric methylation required concentrations of S-adenosylmethionine to be in substantial excess over those of peptide; 50 microM S-adenosylmethionine is the minimum needed for complete modification of 10 microM peptide. Spontaneous demethylation was significant under all conditions tested, so that the methyl ester itself never reached a ratio of 1 mol/mol of total peptide. These results demonstrate that the most accurate measurements of isoaspartate are obtained when reactions are carried out at low peptide concentrations, high S-adenosylmethionine concentrations, and high enzyme concentrations.(ABSTRACT TRUNCATED AT 250 WORDS)