Philip N. McFadden

Molecular Phylogenetics of a Protein Repair Methyltransferase

Protein-L-isoaspartyl (D-aspartyl) O-methyltransferase (E.C. 2.1.1.77) is a well-conserved and widely distributed protein repair enzyme that methylates isomerized or racemized aspartyl residues in age-damaged proteins. We exploited the availability of protein sequences from 10 diverse animal, plant and bacterial taxa to construct a phylogenetic tree and determine the rates of amino acid substitution for this enzyme. We used a likelihood ratio test to show that this enzyme fulfills the conditions for a molecular clock. We found that the rate of substitution is 0.39 amino acid substitutions per site per 10(9) years and remains relatively constant from bacteria to humans. We argue that this degree of sequence conservation may result from the functional constraints necessitated by the requirement to specifically recognize altered aspartyl but not normal aspartyl residues in proteins. Relative rate analysis of the Caenorhabditis elegans sequence suggests that the amino acid substitution rate in the nematode lineage may be higher than that in other lineages and that the divergence of nematodes may have been a more recent event than suggested by previous analysis.

Injury-Induced Enzymatic Methylation of Aging Collagen in the Extracellular Matrix of Blood Vessels

As a result of blood vessel injury, protein D-aspartyl/L-isoaspartyl carboxyl methyltransferase (PIMT), a normally intracellular enzyme, becomes trapped within the meshwork of the vascular extracellular matrix where it can methylate substrate proteins. In this investigation we examined the distribution of such altered aspartyl-containing substrate proteins in the vascular wall. Nearly 90% of all the altered aspartyl residues were inaccessible to intracellular PIMT. Proteins of the extracellular matrix were found to be the major repository of altered aspartyl-containing polypeptides in the blood vessel wall, accounting for ∼70% of the total amount. Proteolytic cleavage of extracellular matrix proteins with cyanogen bromide (CNBr) revealed that collagens account for most of the altered aspartyl-containing proteins of the ECM. As a consequence of blood vessel injury, both type I and type III collagen along with other proteins were found to become methylated by injury-released PIMT. It is estimated that 1 cm of vein contains on the order of 5×1014 altered aspartyl residues involving between 1% and 5% of the total extracellular protein.

Automethylation of protein (d-aspartyl/l-isoaspartyl) carboxyl methyltransferase, a response to enzyme aging

A question that is central to understanding the mechanisms of aging and cellular deterioration is whether enzymes involved in recognition and metabolism of spontaneously damaged proteins are themselves damaged, either becoming substrates for their own activity; or being unable to act upon themselves, initiating cascades of cellular damage. We show here byin vitro experiments that protein (d-aspartyl/l-isoaspartyl) carboxyl methyltransferase (PCM) from bovine erythrocytes does methylate age-dependent amino acid damage in its own sequence. The subpopulation that is methylated, termed theαPCM fraction, appears to be formed through age-dependent deamidation of an asparaginyl site to either anl-isoaspartyl ord-aspartyl site because (a) the stoichiometry of automethylation of purified PCM is less than 1%, a value typical of the substoichiometric methylation of many other aged protein substrates, (b)αPCM is slightly more acidic than the bulk of PCM, and (c) the methyl esterified site inαPCM has the characteristic base-lability of this type of methyl ester. Also, the methyl group is not incorporated into the enzyme as an active site intermediate because the incorporated methyl group is not chased onto substrate protein. The effect of enzyme dilution on the rate of the automethylation reaction is consistent with methylation occurring between protein molecules, showing that the pool of PCM is autocatalytic even though individual molecules may not be. The automethylation and possible self-repair of the PCM pool has implications for maintaining thein vivo efficiency of methylation-dependent protein repair.

Detection and Characterization of a Protein Isoaspartyl Methyltransferase Which Becomes Trapped in the Extracellular Space During Blood Vessel Injury

Injury to rat blood vessels in vivo was found to release intracellular pools of protein D-aspartyl/L-isoaspartyl carboxyl methyltransferase (PIMT) into the extracellular milieu, where it becomes trapped. This trapped cohort of PIMT is able to utilize radiolabeled S-adenosyl-L-methionine (AdoMet) introduced into the circulation to methylate blood vessel proteins containing altered aspartyl residues. As further shown in this study, methylated substrates are detected only at the specific site of injury. In vitro studies more fully characterized this endogenous PIMT activity in thoracic aorta and inferior vena cava. Methylation kinetics, immunoblotting, and the lability of methylated substrates at mild alkaline pH were used to demonstrate that both types of blood vessel contain an endogeneous protein D-aspartyl/L-isoaspartyl carboxyl methyltransferase (PIMT). At least 50% of the PIMT activity is resistant to nonionic detergent extraction, suggesting that the enzyme activity becomes trapped within or behind the extracellular matrix (ECM). Quantities of lactate dehydrogenase (LDH), another soluble enzyme of presumed intracellular origin, were found to be similarly trapped in the extracellular space of blood vessels.

Biosensors based on the chromatic activities of living, naturally pigmented cells : digital image processing of the dynamics of fish melanophores

The activities of fish chromatophores, the colorful living cells in the skin and scales of fish, were tested for their suitability as optically interfaced, digitally processed biosensors of the chemical environment. Norepinephrine and one of its pharmacologically active antagonists, yohimbine, were among the agents tested on melanophores, the black-pigmented variety of chromatophores. The melanophore response consisted of movements of cellular pigment toward and/or away from the center of the cell. This chromatic activity yielded high-contrast video images that were well-suited to digital image processing. Two variations of digital image processing were successful at quantifying pigment movements over the non-changing background: subtraction of the final image in a time-lapsed sequence of images; and subtraction of adjacent images in an image sequence. In both of these variations, a further refinement (noise reduction) of the digital measurements of pigment movement was achieved by taking averages of adjacent pixels in the differential images. Finally, pixel-counting algorithms were used to quantify changes in the area, perimeter and circularity of melanophore pigment. Measurements of pigment movements obtained in this automated, objective manner compared well to measurements obtained by manually tracing melanophore images and calculating areas and perimeters. In addition, the continuous measurement variables achieved by these digital image processing procedures were appropriate for statistical analyses of melanophore dynamics (e.g. population mean and standard deviation), unlike traditional ranked variable measurements such as the five-point ranking system widely known as the melanophore index.

A two-cell biosensor that couples neuronal cells to optically monitored fish chromatophores

A two-cell biosensor was developed that uses optically detected changes in naturally colored fish chromatophores to measure the neurosecretory output of mammalian neuronal cells. The specific version of the biosensor described here is a continuous flow device that places red-pigmented, dendritic erythrophore cells directly downstream of an immobilized population of PC12 neuronal cells, a well-established model cell-line having neuroendocrine function. Agents known to stimulate catecholamine neurosecretion (secretagogues) were presented to the PC12 cells. It was found that the varying level of neurosecretion from the PC12 cells was measurable by judging the degree of pigment aggregation in the erythrophores. Increases in catecholamine secretion and consequent pigment aggregation were observed for several known secretagogues, including receptor agonists (ATP, acetylcholine), membrane depolarizing agents (high K(+) concentration), and specific neurotoxins (black widow spider venom, alpha-latrotoxin). This particular two-cell biosensor, which is applicable to the detection of any agents that affect the levels of catecholamine secretion from PC12 cells, demonstrates the general principle that the breadth of sensitivity of a biosensor is increased by employing coupled cell types.

Incorporation of two 18O atoms into a peptide during isoaspartyl repair reveals repeated passage through a succinimide intermediate.

To study the mechanism of protein carboxyl methyltransferase-driven repair of age-damaged sites in polypeptides, a modell-isoaspartyl peptide,l-isotetragastrin, was enzymatically repaired to normall-tetragastrin in the presence of18O-enriched water. By this design, the enrichment of18O atoms in the peptide would reflect the number of passages through a hydrolyzable succinimide intermediate during formation of the repaired product. Mass determinations by FAB mass spectrometry revealed repaired peptide with two18O atoms incorporated, demonstrating that more than a single cycle of methylation and demethylation is necessary to ensure stoichiometric repair.

Trapping succinimides in aged polypeptides by chemical reduction

Cyclization of aspartic acid and asparagine to succinimides is thought to be a common spontaneous aging reaction in proteins, but the instability of the succinimide ring has made it difficult to directly measure this structure. Chemical reduction has now been tested as a means of trapping succinimides as stable derivatives, homoserine and isohomoserine. Two succinimide-containing compounds were tested in this manner. First, polysuccinimide was reduced by sodium borohydride to a derivative that contained homoserine and isohomoserine in amounts that were consistent with the content of succinimide determined independently by quantitative hydrolysis. The identity of isohomoserine was confirmed by its resistance to degradation by L-amino acid oxidase, and through its synthesis by an alternate route involving borane reduction of asparagine. Second, in a test of this approach on a peptide mixture with only a trace-content of succinimide, isohomoserine and homoserine were formed as reduction products in amounts equivalent to the trace content of succinimide in the mixture. Detection of the products of the chemical reduction of polypeptides is therefore diagnostic of succinimides, and can be successfully applied at the trace sensitivity necessary for studies of naturally aging proteins. A related study of the reduction of aspartyl andΒ-aspartyl residues to, respectively, homoserine and isohomoserine, is described in the accompanying manuscript (Carter and McFadden, 1994).

[31] Isolation of d-aspartic acid β-methyl ester from erythrocyte carboxyl methylated proteins

Publisher Summary This chapter describes isolation and identification of D-aspartic acid β -methyl ester from human erythrocytes. It also discusses possible artifacts that can arise during such preparations from erythrocytes and other eukaryotic tissues. L-[ methyl - 3 H]methionine or S -adenosyl-L-[ methyl - 3 H] methionine are utilized to radiolabel methyl groups on erythrocyte proteins for preparation of 3 H-methylated proteins. The former isotope has been used to label methyl esters in intact cells, since it is taken up and converted into the active methyl group donor, S -adenosyl[ methyl - 3 H]methionine, by a cellular synthetase activity. The use of isotopes containing up to 3 tritium atoms per methyl group (60-90 Ci/nmol) gives the highest yield of aspartic acid β -methyl ester in subsequent digestions. It is possible that radioactive materials other than aspartic acid β -methyl ester, including undigested peptides, may comigrate with the non-isotopically labeled standard ester, therefore it is necessary to further characterize the radioactive product isolated from the ion-exchange chromatography step. The standard procedure is to perform a gel filtration step which functions not only to desalt the sample but also to separate the aspartic acid β -methyl ester from other radioactive components. The most convenient method to determine the optical configuration of isolated aspartic acid β -[ 3 H]methyl ester involves the synthesis of dipeptide leucyl diastereomers which can be separated on amino acid analysis resin after the procedure of Manning and Moore.

Determination of two sites of automethylation in bovine erythrocyte protein (D-aspartyl/L-isoaspartyl) carboxyl methyltransferase.

Protein (d-aspartyl/l-isoaspartyl) carboxyl methyltransferase (PCM, E.C. 2.1.1.77) was previously shown to be enzymatically methyl esterified in an autocatalytic manner at altered aspartyl residues; methyl esters are observed in a subpopulation of the enzyme termed theαPCM fraction [Lindquist and McFadden (1994),J. Protein Chem.13, 23–30]. The altered aspartyl sites serving as methyl acceptors inαPCM have now been localized by using proteolytic enzymes and chemical cleavage techniques in combination with matrix-assisted laser desorption/ionization (MALDI) mass spectrometry to identify fragments of the [3H]automethylated enzyme that contain a [3H]methyl ester. Methylation was positively identified at positions Asn188 and Asp217 in the enzyme sequence, a consequence of the spontaneous alteration of these sites tol-isoaspartyl ord-aspartyl sites and their methylation by active PCM molecules. The identification of more than one site of automethylation shows thatαPCM is not a homogeneous population of damaged PCM molecules, but rather a complex population of molecules with a variety of age-altered damage sites.