An evaluation of the role of a pyroglutamyl peptidase, a post-proline cleaving enzyme and a post-proline dipeptidyl amino peptidase, each purified from the soluble fraction of guinea-pig brain, in the degradation of thyroliberin in vitro

The degradation of thyroliberin (less than Glu-His-Pro-NH2) to its component amino acids by the soluble fraction of guinea pig brain is catalysed by four enzymes namely a pyroglutamate aminopeptidase, a post-proline cleaving enzyme, a post-proline dipeptidyl aminopeptidase and a proline dipeptidase. 1. The pyroglutamate aminopeptidase was purified to over 90% homogeneity with a purification factor of 2868-fold and a yield of 5.7%. In addition to catalysing the hydrolysis of thyroliberin, acid thyroliberin and pyroglutamate-7-amido-4-methylcoumarin the pyroglutamate aminopeptidase catalysed the hydrolysis of the peptide bond adjacent to the pyroglutamic acid residue in luliberin, neurotensin bombesin, bradykinin-potentiating peptide B, the anorexogenic peptide and the dipeptides pyroglutamyl alanine and pyroglutamyl valine. Pyroglutamyl proline and eledoisin were not hydrolysed. 2. The post-proline cleaving enzyme was purified to apparent electrophoretic homogeneity with a purification factor of 2298-fold and a yield of 10.6%. The post-proline cleaving enzyme catalysed the hydrolysis of thyroliberin and N-benzyloxycarbonyl-glycylproline-7-amido-4-methylcoumarin. It did not catalyse the hydrolysis of glycylproline-7-amido-4-methylcoumarin or His-Pro-NH2. 3. The post-proline dipeptidyl aminopeptidase was partially purified with a purification factor of 301-fold and a yield of 8.9%. The post-proline dipeptidyl aminopeptidase catalysed the hydrolysis of His-Pro-NH2 and glycylproline-7-amido-4-methylcoumarin but did not exhibit any post-proline cleaving endopeptidase activity against thyroliberin or N-benzyloxycarbonyl-glycylproline-7-amido-4-methylcoumarin. 4. Studies with various functional reagents indicated that the pyroglutamate aminopeptidase could be specifically inhibited by 2-iodoacetamide (100% inhibition at an inhibitor concentration of 5 microM), the post-proline cleaving enzyme by bacitracin (IC50 = 42 microM) and the post-proline dipeptidyl aminopeptidase by puromycin (IC50 = 46 microM). Because of their specific inhibitory effects these three reagents were key elements in the elucidation of the overall pathway for the metabolism of thyroliberin by guinea pig brain tissue enzymes.     

Specificity of a serum peptidase hydrolyzing thyroliberin at pyroglutamyl-histidine bone

The substrate specificity of a serum enzyme which degrades thyroliberin (less than Glu-His-Pro-NH2) into pyroglutamic acid and His-Pro-NH2 has been investigated and compared with that of the pyroglutamyl aminopeptidase from calf liver. The latter enzyme has a broad specificity, causing rapid degradation of thyroliberin, pyroglutamyl beta-naphthylamide and luliberin. In contrast, the serum enzyme causes rapid stereospecific cleavage only of the pyroglutamyl-histidine bond of thyroliberin and closely related peptides. Compounds such as less than Glu-Ala, less than Glu-His and pyroglutamyl beta-naphthylamide, which are known substrates of the pyroglutamyl aminopeptidases (such as the liver enzyme), are not substrates of the serum enzyme, and inhibit it only poorly. Pyroglutamyl-containing peptides such as luliberin and neurotensin and thyroliberin analogues such as LLD-thyroliberin, less than Glu-His-Pro-NHCH3, less than Glu-His-Pro-Gly-NH2 and less than Glu-Phe-Pro-NH2 inhibit effectively the degradation of thyroliberin by the serum enzyme, but are not hydrolyzed by this enzyme. The high specificity of the serum enzyme implies a physiological function.     

Post-proline dipeptidyl-aminopeptidase from synaptosomal membranes of guinea-pig brain. A possible role for this activity in the hydrolysis of His-ProNH2, arising from the action of synaptosomal membrane pyroglutamate aminopeptidase on thyroliberin

In this paper we report that while 55% of the total post-proline dipeptidyl-aminopeptidase activity in guinea-pig brain is associated with the soluble fraction of the cells, the remaining activity is widely distributed throughout the particulate fractions. A significant portion of this particulate activity is, however, associated with a synaptosomal membrane fraction. The specific activity of this enzyme rose as the synaptosomal membrane fraction was prepared from a synaptosomal fraction and had previously risen at the synaptosomal fraction was prepared from a postmitochondrial pellet. The synaptosomal membrane post-proline dipeptidyl-aminopeptidase was released from the membrane by treatment with Triton X-100 and partially purified by chromatography on Sephadex G-200. By contrast with the soluble enzyme the partially purified solubilised synaptosomal membrane post-proline dipeptidyl-aminopeptidase was not inhibited by 1.0 mM p-chloromercuribenzoate, 1.0 mM N-ethylmaleimide or 0.5 mM puromycin but was inhibited by 0.5 mM bacitracin. The partially purified solubilised enzyme was capable of releasing His-Pro from His-Pro-Val, His-Pro-Leu, His-Pro-Phe and His-Pro-Tyr and of releasing Gly-Pro from Gly-Pro-Ala but could not release Arg-Pro from Arg-Pro-Pro or from Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg (bradykinin). It was also unable to release Pro-Pro from Pro-Pro-Gly or Glp-Pro from Glp-Pro-Ser-Lys-Asp-Ala-Phe-Ile-Gly-Leu-MetNH2 (eledoisin). Using [Pro-3H]thyroliberin we show that the membrane-bound enzyme converts His-ProNH2, produced by the action of the synaptosomal membrane pyroglutamate aminopeptidase, to His-Pro thus competing with the spontaneous cyclisation of His-ProNH2 to His-Pro diketopiperazine. Purified preparations of synaptosomal membrane pyroglutamate aminopeptidase were used to generate His-ProNH2, which could then be converted to His-Pro by the presence of the partially purified synaptosomal membrane post-proline dipeptidyl-aminopeptidase. This preparation was free of contaminating post-proline cleaving endopeptidase, carboxypeptidase P, aminopeptidase P, prolyl carboxypeptidase or proline dipeptidase.     

Pyroglutamate aminopeptidase in rat submaxillary gland.

A significant amount of pyroglutamate aminopeptidase (PGAP) activity was found to be present in 27,000 x g supernatant of rat submaxillary gland, maximum activity being at pH 6.5. EDTA stimulated the enzyme activity by 95% at pH 8.0 while at pH 6.5 it did not have any significant effect. On comparison of its properties submaxillary PGAP appears to be different from brain, pituitary and other reported PGAPs. Submaxillary PGAP could also catalyze efficiently the formation of cyclo (His-Pro) from TRH. Cyclo (His-Pro) formation by submaxillary enzyme was more pronounced than that by liver PGAP.     

Purification of and kinetic studies on a narrow specificity synaptosomal membrane pyroglutamate aminopeptidase from guinea-pig brain

A pyroglutamate aminopeptidase activity, distinct from that of cytoplasm, was released from a synaptosomal membrane preparation of guinea-pig brain by papain treatment. This activity was further purified 3560-fold relative to the homogenate with a yield of 17% by a procedure involving gel filtration chromatography, calcium phosphate cellulose chromatography and hydrophobic interaction chromatography on phenyl-Sepharose CL-4B. The purified synaptosomal pyroglutamate aminopeptidase hydrolysed only thyroliberin, acid-thyroliberin, the luliberin N-terminal tripeptide (Glp-His-Trp) and, only slightly, Glp-His-Gly. No hydrolysis was observed with dipeptides containing N-terminal pyroglutamic acid (Glp) or with pyroglutamyl peptides containing more than three amino acids. A Km value of 40 microM was recorded when thyroliberin was used as substrate; however, luliberin was found to inhibit the hydrolysis of thyroliberin competitively with a Ki value of 20 microM.     

Developmental changes in the distribution of rat brain pyroglutamate aminopeptidase,a possible determinant of endogenous cyclo(His-Pro) concentrations

The distribution of cyclo(His-Pro), thyrotropin-releasing hormone (TRH) and Pyroglutamate aminopeptidase activity in adult and developing rat brains were studied. A comparison of the subcellular distribution of Pyroglutamate aminopeptidase activity in hypothalamic and cerebral cortical extracts from adult rats exhibited remarkable differences. In hypothalamus, the enzyme activity was mainly associated with the soluble fraction whereas in cortex it was predominantly associated with the particulate fractions. During postnatal development, the brain concentrations of cyclo(His-Pro) and Pyroglutamate aminopeptidase activities declined with age. These data suggest that Pyroglutamate aminopeptidase activity, but not TRH, plays an active role in determining the levels of endogenous cyclo(His-Pro) concentrations in brain.     

The narrow specificity pyroglutamate amino peptidase degrading TRH in rat brain is an ectoenzyme

In order to determine the pathway of extracellular metabolism of the thyrotropin releasing hormone (pyroglu-his-proNH₂) in brain, the topographical organization of pyroglutamate aminopeptidase II on the plasma membrane was investigated. Its activity was only slightly increased when intact brain synaptosomes were lysed by osmotic shock or detergent treatment. Trypsin treatment of intact synaptosomes destroyed 70–80% of enzyme activity without affecting lactate dehydrogenase. Pyroglutamate aminopeptidase II activity was present in primary cultures of foetal mice cortical cells. It was detected in intact cells, was not released by the cells and its activity was not increased by saponin pretreatment. Trypsin treatment of the cells reduced pyroglutamate aminopeptidase II by 70% but did not affect pyroglutamate aminopeptidase I and lactate dehydrogenase. These data support that brain pyroglutamate aminopeptidase II is an ectoenzyme. They suggest that this enzyme could be responsible for thyrotropin releasing hormone extracellular catabolism in brain.     

Unfolding of the immunoglobulin light and heavy chains is required for the enzymatic removal of N-terminal pyroglutamyl residues

To enable Edman sequencing of pyroglutamylated immunoglobulins, enzymatic deblocking by pyroglutamate aminopeptidase is performed, often with variable yield and compromised solubility. Recently, enzymatic deblocking of immunoglobulins without denaturation was described. Although the conditions ensured efficient removal of pyroglutamyl residues, we conclude that deblocking is preceded by denaturation, which results in aggregation of the immunoglobulins. To study the effect of folding status on deblocking we developed a methanol based deblocking solution, which preserved the enzymatic activity of pyroglutamate aminopeptidase, provided conditions compatible with sequencing and enhanced deblocking of electroblotted samples, as well. At 50 degrees C and 35% (v/v) methanol the immunoglobulin chains were completely aggregated, but the degree of deblocking was comparable to that obtained with the previously described method. At 37 degrees C, the immunoglobulins were partly aggregated, but the deblocked chains were completely in the insoluble fractions, whereas the soluble fractions had retained pyroglutamylation in both chains, suggesting that unfolding of the immunoglobulins is required for the excision of the pyroglutamates. Inspection of the structures of pyroglutamylated immunoglobulin and pyroglutamate aminopeptidase P. furiosus indicates that the enzyme requires the substrate in an extended conformation, a criterium, which we conclude not to be fulfilled in the native form of immunoglobulins. Unfolding of the N-terminus would disrupt the immunoglobulin fold by breaking interactions between secondary structure elements and expose surfaces prone to aggregation.     

Biosynthesis of Tetrapyrrole Pigment Precursors : Formation and Utilization of Glutamyl-tRNA for delta-Aminolevulinic Acid Synthesis by Isolated Enzyme Fractions from Chlorella Vulgaris

The universal tetrapyrrole precursor δ-aminolevulinic acid (ALA) is formed from glutamate (Glu) in algae and higher plants. In the postulated reaction sequence, Glu-tRNA is produced by a Glu-tRNA synthetase, and the product serves as a substrate for a reduction step catalyzed by a pyridine nucleotide-requiring Glu-tRNA dehydrogenase. The reduced intermediate is then converted into ALA by a transaminase. An RNA and three enzyme fractions required for ALA formation from Glu have been isolated from soluble Chlorella extracts. The recombined fractions catalyzed ALA production from Glu or Glu-tRNA. The fraction containing the synthetase produced Glu-tRNA from Glu and tRNA in the presence of ATP and Mg²⁺. The isolated product of this reaction served as substrate for ALA production by the partially reconstituted enzyme system lacking the synthetase fraction and incapable of producing ALA from Glu. The production of ALA from Glu-tRNA by this partially reconstituted system did not require free Glu or ATP, and was not affected by added ATP. These results show that (a) free Glu-tRNA is an intermediate in the formation of ALA from Glu, (b) ATP is required only in the first step of the reaction sequence, and NADPH only in a later step, (c) Glu-tRNA production is the essential reaction catalyzed by one of the enzyme fractions, (d) this enzyme fraction is active in the absence of the other enzymes and is not required for activity of the others. The specific Glu-tRNA synthetase required for ALA formation has an approximate molecular weight of 73,000 ± 5,000 as determined by Sephadex G-100 gel filtration and native polyacrylamide gel electrophoresis. Other Glu-tRNA synthetases were present in the cell extracts but were ineffective in the the ALA-forming process.     

Is all cyclo(His-Pro) derived from thyrotropin-releasing hormone?

Cyclo(His-Pro), or histidyl-proline diketopiperazine, is an endogenous cyclic dipeptide that is ubiquitously distributed in tissues and body fluids of both man and animals. This cyclic dipeptide is not only structurally related to thyrotropin-releasing hormone (TRH, pGlu-His-ProNH₂), but it can also arise from TRH by the action of the enzyme pyroglutamate amino-peptidase (pGlu-peptidase). The data on the distribution of TRH, cyclo(His-Pro), and pGlu-peptidase under normal and abnormal conditions are summarized and potential relationships analyzed. We conclude that all of the cyclo(His-Pro) cannot be derived from TRH. Two additional sources of cyclo(His-Pro) are suggested. It is proposed that 29,247 molecular weight TRH prohormone, prepro TRH, which contains 5 copies of TRH sequence, can be processed to yield cyclo(His-Pro). Thus, both TRH and cyclo(His-Pro) share a common precursor, prepro[TRH/Cyclo(His-Pro)].     

The removal of pyroglutamic acid from monoclonal antibodies without denaturation of the protein chains

Typically, the removal of pyroglutamate from the protein chains of immunoglobulins with the enzyme pyroglutamate aminopeptidase requires the use of chaotropic and reducing agents, quite often with limited success. This article describes a series of optimization experiments using elevated temperatures and detergents to denature and stabilize the heavy chains of immunoglobulins such that the pyroglutamate at the amino terminal was accessible to enzymatic removal using the thermostable protease isolated from Pyrococcus furiosus. The detergent polysorbate 20 (Tween 20) was used successfully to facilitate the removal of pyroglutamate residues. A one-step digestion was developed using elevated temperatures and polysorbate 20, rather than chaotropic and reducing agents, with sample cleanup and preparation for Edman sequencing performed using a commercial cartridge containing the PVDF membrane. All of the immunoglobulins digested with this method yielded heavy chain sequence, but the extent of deblocking was immunglobulin dependent (typically>50%).     

Characterization of two carnosine-degrading enzymes from rat brain. Partial purification and characterization of a carnosinase and a beta-alanyl-arginine hydrolase

From rat brain extracts, two carnosine-degrading enzymes have been identified and partially purified by ion-exchange chromatography, hydrophobic interaction chromatography on phenyl-Sepharose CL-4B and gel filtration. These enzymes exhibit distinct differences in their chemical characteristics and substrate specificities. One enzyme, designated carnosinase, preferentially hydrolyzes carnosine and exhibits a low Km value (0.02 mM) towards this substrate. Carnosinase also degrades anserine but not homocarnosine or homoanserine. The other carnosine-degrading enzyme hydrolyzes beta Ala-Arg considerably faster than carnosine and, therefore, has been tentatively designated beta Ala-Arg hydrolase. This enzyme exhibits high Km values with carnosine (Km = 25 mM) and beta Ala-Arg (Km = 2 mM). Homocarnosine and gamma-aminobutyryl-arginine are not degraded by beta Ala-Arg hydrolase. Neither enzyme is inhibited by agents reactive on activated hydroxyl groups, such as diisopropyl fluorophosphate, and also not by a variety of peptidase inhibitors of microbial origin or from other sources. Carnosinase is also not inhibited by bestatin but beta Ala-Arg hydrolase, although not an aminopeptidase, is strongly inhibited by this aminopeptidase inhibitor (IC50 = 50 nM). While carnosinase is strongly inhibited by thiol-reducing agents such as dithioerythritol and 2-mercaptoethanol, beta Ala-Arg hydrolase is stabilized and activated by these substances. Both enzymes are strongly inhibited by metal-chelating agents. Carnosinase, however, is not dependent on exogeneously added metal ions and is strongly inhibited by Mn2+ as well as by heavy metal ions. In contrast, beta Ala-Arg hydrolase requires Mn2+ ions for full enzymatic activity. Based on these differences, selective incubation conditions could be evaluated in order to determine specifically both enzyme activities in crude tissue extracts. In rat, both enzymes are present in all tissues tested, except skeletal muscles, but considerable differences in their relative distribution among different tissues are also observed.     

Occurrence of epsilon-poly-L-lysine-degrading enzyme in epsilon-poly-L-lysine-tolerant Sphingobacterium multivorum OJ10: purification and characterization

Epsilon-poly-L-lysine (epsilon-PL)-degrading enzyme was found in the epsilon-PL-tolerant strain Sphingobacterium multivorum OJ10 and purified to homogeneity. The purified enzyme has a molecular mass of approximately 80 kDa. The enzyme catalyzed exo-type degradation of epsilon-PL and released L-lysine. The enzyme was a Co2+ or Ca2+ ion-activated aminopeptidase.     

Neuronal Localization of Pyroglutamate Aminopeptidase II in Primary Cultures of Fetal Mouse Brain

Pyroglutamate aminopeptidase II is a highly specific membrane-bound ectopeptidase proposed to inactivate thyrotropin releasing hormone (TRH) in brain extracellular space. Its activity was measured in primary cell cultures of fetal brain in an attempt to define its cellular localization. Enzyme activity was detected in hypothalamic or cortical cell membrane fractions from 4- to 12-day-old cultures. When proliferation of nonneuronal cells was abolished by cytosine arabinoside treatment, pyroglutamate aminopeptidase II specific activity was increased as compared to untreated cultures, the opposite was observed for pyroglutamate amino-peptidase I activity. Treatment of cortical cells with the neurotoxic agent glutamate reduced simultaneously pyroglutamate aminopeptidase II and glutamate decarboxylase activities. Glial cell cultures expressed pyroglutamate aminopeptidase I or glutamate synthase activities but not pyroglutamate aminopeptidase II. The data suggest that pyroglutamate aminopeptidase II is predominantly localized in neuronal cells. This is consistent with a role for pyroglutamate aminopeptidase II in TRH-ergic synaptic transmission.     

Ontogeny of two topologically distinct TRH-degrading pyroglutamate aminopeptidase activities in the rat liver.

We recently separated and characterized two topologically distinct pyroglutamate aminopeptidase (PAP) activities in adult rat liver, which convert TRH to cyclo His-Pro (cHP). The liver possesses high-affinity binding sites to the biologically active dipeptide cHP and is thus a potential target tissue for pancreatic TRH and/or its conversion product cHP, and may be a site of TRH conversion and/or inactivation. This report describes the ontogenic development of two liver PAP activities and compares them with that of plasma thyroliberinase. The particulate high-molecular-weight PAP was absent at birth and during the neonatal period, while the soluble, low-molecular-weight PAP was present at all the developmental stages tested. The changes in particulate PAP activity are similar to those in the plasma of age-matched rats. The peculiar age-dependent changes in particulate PAP activity, plus its cellular location, suggest that it has a regulatory role.     

Subcellular localization and levels of aminopeptidases and dipeptidase in Saccharomyces cerevisiae.

Three aminopeptidases (L-aminoacyl L-peptide hydrolases, EC 3.4.11) and a single dipeptidase (L-aminoacyl L-amino acid hydrolase, EC 3.4.13) are present in homogenates of Saccharomyces cerevisiae. Bassed on differences in substrate specificity and the sensitivity to Zn2+ activation, methods were developed that allow the selective assay of these enzymes in crude cell extracts. Experiments with isolated vacuoles showed that aminopeptidase I is the only yeast peptidase located in the vacuolar compartment. Aminopeptidase II (the other major aminopeptidase of yeast) seems to be an external enzyme, located mainly outside the plasmalemma. The synthesis of aminopeptidase I is repressed in media containing more than 1% glucose. In the presence of ammonia as the sole nitrogen source its activity is enhanced 3--10-fold when compared to that in cells grown on peptone. In contrast, the levels of aminopeptidase II and dipeptidase are less markedly dependent on growth medium composition. It is concluded that aminopeptidase II facilitates amino acid uptake by degrading peptides extracellularly, whereas aminopeptidase I is involved in intracellular protein degradation.     

Post-translational modification of bovine pro-opiomelanocortin. Tyrosine sulfation and pyroglutamate formation, a mass spectrometric study

The amino-terminal fragment of beta-lipotropin (i.e. beta-lipotropin (1-40)) and joining peptide portions of pro-opiomelanocortin have been purified from extracts of bovine posterior pituitaries. Peptides were purified using a combination of reversed-phase and ion-exchange batch extraction procedures followed by reversed-phase high performance liquid chromatography. beta-Lipotropin (1-40) was found to consist of four major components while joining peptide was found to consist of two major components. Fast atom bombardment-mass spectrometric analysis of the tryptic fragments of both peptides revealed that the observed heterogeneity could be explained in terms of post-translational modifications. beta-Lipotropin (1-40) was found to be sulfated at tyrosine residue 28 to an extent of about 50%. The tyrosine residue in beta-lipotropin (1-40) is situated within an amino acid sequence with a preponderance of glutamate residues. Sulfation of this amino acid residue is entirely compatible with the known primary structure requirements of the sulfotransferase enzyme located in the trans-Golgi fraction. Both beta-lipotropin (1-40) and joining peptide were found to have pyroglutamate at their amino termini to an extent of about 50%. The cDNA sequence for bovine pro-opiomelanocortin predicts the presence of glutamic acid at position 1 of both peptides. Pyroglutamate is normally formed through the cyclization of glutamine. This reaction is thought to be catalyzed by a pyroglutamate forming enzyme located within the secretory granule fraction. Under certain circumstances peptides with glutamate at their amino termini may act as substrates for this enzyme.     

Glutaminyl cyclases unfold glutamyl cyclase activity under mild acid conditions

N-terminal pyroglutamate (pGlu) formation from glutaminyl precursors is a posttranslational event in the processing of bioactive neuropeptides such as thyrotropin-releasing hormone and neurotensin during their maturation in the secretory pathway. The reaction is facilitated by glutaminyl cyclase (QC), an enzyme highly abundant in mammalian brain. Here, we describe for the first time that human and papaya QC also catalyze N-terminal glutamate cyclization. Surprisingly, the enzymatic Glu(1) conversion is favored at pH 6.0 while Gln(1) conversion occurs with an optimum at pH 8.0. This unexpected finding might be of importance for deciphering the events leading to deposition of highly toxic pyroglutamyl peptides in amyloidotic diseases.     

Degradation of Melanotropin Inhibiting Factor by Brain

Degradation of melanotropin inhibiting factor (MIF) was measured by fluorometry, using pareptide as an internal standard, following the separation of the dansyl derivatives of MIF and its metabolites by HPLc. MIF was not split by carboxypeptidases A and B, prolidase, or pyroglutamate aminopeptidase. It was hydrolyzed by leucine aminopeptidase, aminopeptidase M, and carboxypeptidase Y. Rat brain hydrolyzed 159 nmol of MIF per mg of protein per h; the activity was linear with enzyme concentration. Hydrolysis start from the N-terminal end, as shown by the appearance of proline as the first metabolite of the MIF degradation, followed by leucine, glycinamide, leucylglycine, and glycine. Activity in the rat brain regions was in the order striatum, medulla oblongata > cortex, hippocampus, midbrain > hypothalamus, cerebellum, and pituitary. The enzyme was mostly in the supernatant, with significant amounts in the myelin and synaptosomal fractions. MIF aminopeptidase could be separated from carboxypeptidase by centrifugation at 30,000 x g for 20 min and precipitation with 45--75% (NH4)2SO4. It showed pH optima in the alkaline range (8.25 and 8.75) and was inhibited by EDTA, EGTA, SQ 14,225, puromycin, bacitracin, and bestatin.     

Freezing Injury and Phospholipid Degradation in Vivo in Woody Plant Cells: III. Effects of Freezing on Activity of Membrane-bound Phospholipase D in Microsome-enriched Membranes

Freeze-thawing of microsome-enriched membranes from living bark tissues of black locust trees, especially those from less hardy tissues, caused a drastic increase in sensitivity to Ca²⁺ and a complete loss of the regulatory action of Mg²⁺ in membrane-bound phospholipase D activity with endogenous (membrane-bound) substrates. Also, the freeze-thaw cycle made phospholipase D in these membranes more resistant to digestion by proteases. Thus, the regulatory properties of the membrane-bound phospholipase D seem to be dependent on the nature of the membranes and on the interaction between the enzyme and membranes as well. The alteration of regulatory properties by freezing was protected by sucrose, at lower concentrations, and more effectively for membranes from hardy tissues than for membranes from less hardy tissue. Addition of partially purified soluble phospholipase D to the reaction system containing membranes caused only a slight stimulation of the degradation of endogenous phospholipids. Phospholipid degradation in vivo during freezing of less hardy tissue may be catalyzed mainly by the bound enzyme. Disintegration of the tonoplast, however, besides releasing soluble phospholipase D into the cytosol, would release organic acids (lowering the pH) and free Ca²⁺. Both factors would stimulate drastically the membrane-bound phospholipase D, causing degradation of membrane phospholipids.