Localization and some properties of lysosomal dipeptidases in rat liver.

1. The rates of hydrolysis of 26 synthetic dipeptides by extracts from highly purified lysosomal fractions from rat liver at pH 5.0 and by whole liver homogenates at pH 7.4 have been determined. Extracts from the lysosomal fractions hydrolysed most peptides at a lower rate per mg protein than the homogenates, and some peptides not at all. 2. Properties of two dipeptidases present in the extracts from the lysosomal fractions, splitting Ile-Glu and Leu-Gly, respectively, were studied in greater detail. The enzyme that hydrolysed Ile-Glu was strongly activated by dithiothreitol, showed optimal activity at pH 4.5 and had a molecular weight of about 120 000. Leu-Gly dipeptidase did apparently not contain an essential thiol group and had a molecular weight of approx. 90 000. It showed maximal activity at pH 6.5. 3. After differential centrifugation of liver homogenates, Ile-Glu and Leu-Gly-splitting activities were determined in the fractions, under the optimal conditions mentioned above. The Ile-Glu-hydrolysing enzyme activity showed about the same distribution as the lysosomal marker enzyme acid phosphatase. Leu-Gly-splitting activity, however, was largely present in the cytosol fraction, with only a small peak in the lysosomal fraction. We obtained evidence that the activities present in the lysosomal fraction and in the cytosol fraction were due to different enzymes, and that one of these enzymes was localized exclusively in lysosomes. 4. It is concluded that some dipeptides originating from intralysosomal proteolysis might be split by lysosomal dipeptidases, whereas others are probably hydrolysed only in the extra-lysosomal compartment of the cell.

Rapid uptake by liver sinusoidal cells of serum albumin modified with retention of its compact conformation.

The clearance from the blood and the conformation of serum albumin modified by nitroguanidination and labeled with 125-I have been studied. Like formaldehyde-denatured albumin, but in contrast to native albumin, the nitroguanidinated derivative is rapidly cleared from the blood and taken up in lysosomes of liver sinusoidal cells. Although 94% of the free amino groups were blocked by nitroguanidination, we could not detect significant conformational changes using gel filtration, determination of reducible disulfide groups, and titration of tyrosine residues. It is concluded that extensive denaturation is no prerequisite for the uptake of albumin derivatives in liver sinusoidal cells. It is suggested that the nitroguanidinated protein, in contrast to native albumin, is bound on membrane receptors of sinusoidal cells. The nitroguanidino groups themselves might be bound on these receptors, but it seems equally possible that the blocking of positive charges of the albumin molecule or minor, local conformational changes of the protein are sufficient for the binding on the receptors.

Effect of size and charge on endocytosis of lysozyme derivatives by sinusoidal rat liver cells in vivo.

Hen egg-white lysozyme has been modified by intermolecular cross-linking with dimethyl suberimidate or by acylation with acetic or succinic anhydride. Retention of the native conformation of the modified enzyme was checked by measuring enzyme activity, resistance of disulfide bridges to reduction by thiols, and susceptibility to proteases. Unmodified lysozyme and its derivatives (labelled with 125I) were intravenously injected into nephrectomized rats, and plasma clearance and uptake by liver cells were determined. Under these conditions, about 6% of the unmodified lysozyme was taken up by liver 15 min after injection. Cross-linking led to a greatly increased uptake (up to 89% of the dose in 15 min), whereas acylation reduced the uptake to 3-4%. Cell isolations showed that the unmodified enzyme and the cross-linked derivatives were taken up by sinusoidal cells. Differential fractionation of liver homogenates indicated tht the unmodified enzyme was taken up in lysosomes. The cross-linked derivatives were concentrated in the nuclear and microsomal fractions as well as in the lysosomal fraction, suggesting adsorption on plasma membranes besides uptake in lysosomes. The experiments described in this paper, together with previous results on ribonuclease and lactate dehydrogenase, indicate that endocytosis of some proteins by sinusoidal liver cells is positively correlated with size and positive charge of the molecules.

Endocytosis and breakdown of ribonuclease oligomers by sinusoidal rat liver cells in vivo. I. Effect of size.

Aspects of protein structure determining endocytosis of proteins by sinusoidal rat liver cells in vivo have been studied, using cross-linked or aggregated derivatives of bovine pancreatic ribonuclease A (labelled with 125I) as probes. Ribonuclease was cross-linked by reaction with dimethylsuberimidate, a way of modification that does not change the charge of the protein. Monomer, dimer and polymer fractions were isolated by gel filtration and characterized in respect of size and number of amino groups modified. Maintenance of enzyme activity, stability of disulfide bonds, and lack of susceptibility to endoproteases showed that the cross-linking procedure did not result in gross conformational changes of the ribonuclease molecules. Monomer, dimer and polymer fractions were injected into nephrectomized rats and plasma clearance and uptake in liver and spleen were determined. About 30% of the injected polymer fraction was found in liver 15 min after injection; for dimer and monomer fractions values of 6% and 2% of the dose were found. Similar differences were found in spleen. Autoradiography, cell isolation, and subcellular fractionation showed that in liver the radioactive proteins were taken up in lysosomes of sinusoidal cells. Similar results were obtained with fractions of aggregated ribonuclease prepared by freeze-drying the protein from 50% acetic acid. Our results demonstrate that the rate of uptake of the ribonuclease derivatives is positively correlated with the size of the molecules. Similarity of the results obtained with cross-linked and aggregated fractions suggests that the number of ribonuclease 'subunits'/molecule, rather than the procedures used to prepare the polymers, determine the rate of uptake by liver and spleen.

Endocytosis and breakdown of ribonuclease oligomers by sinusoidal rat liver cells in vivo. II. Effect of charge.

Experiments presented in this paper suggest that sinusoidal rat liver cells recognize basic groups on proteins and that this recognition results in endocytosis of the proteins. Evidence for involvement of basic groups was obtained in two ways. Firstly, we changed the positively charged amino groups of the cross-linked ribonuclease molecules to neutral or negative by acetylation or succinylation, respectively. The modified proteins did not contain easily reducible disulfide bonds and they were not very sensitive to endoproteases, suggesting that they were not denatured by the acetylation procedures. Acetylation and succinylation reduced uptake of the injected cross-linked ribonuclease derivatives by liver and spleen and abolished their rapid clearance from plasma. In nephrectomized rats about 75% of the polymer, 36% of the acetylated polymer and 32% of the succinylated polymer were endocytosed by liver after 6 h. For the dimer fractions these values were 59%, 23% and 27%, respectively. Autoradiography and subcellular fractionation of liver 30 min post-injection localized the acetylated polymer in the lysosomal/microsomal fraction of sinusoidal liver cells, probably endothelial cells. Secondly, a positive correlation was found between binding of a number of ribonuclease derivatives to the cation exchanger SP-Sephadex G-25 and the rate of endocytosis by sinusoidal liver cells.

Induction of lysosomal storage by suramin.

Isolated livers of rats injected with saline or with suramin (250 mg per kg body weight) 24h previously were perfused with a medium containing radioactively labeled formaldehyde-treated albumin. Suramin-loaded livers released breakdown products at a much lower rate than controls and contained about the double amount of undigested radioactive protein up to about 3 h after the start of the perfusion. These results show that inhibition of proteolysis by suramin as reported previously (Davies et al., 1971; Buys et al., 1973) is not caused by binding of the drug to the substrate in the bloodstream. Electron micrographs of liver sections of suramin-treated rats showed that lysosomes of sinusoidal cells resembled those seen in certain lysosomal storage diseases. The effect of suramin on lysosomal enzymes was studied in vitro. When used at a concentration corresponding to the putative concentration in lysosomes in vivo, the drug inhibited the lysosomal endopeptidases cathepsin Bl and D as well as acid phosphatase. Inhibition of acid phosphatase by suramin in vivo could also be demonstrated by histochemical methods. These results suggest that the observed storage phenomena may be mainly caused by inhibition of lysosomal enzymes.

Some aspects of plasma protein metabolism as compared with intracellular protein breakdown.

Many plasma proteins are cleared according to first-order kinetics. Half-lives of individual plasma proteins are widely different, just like those of intracellular proteins. Plasma proteins with molecular weights up to about 50000 are often mainly cleared by glomerular filtration, especially if the protein has a high isoelectric point. An other important clearance mechanism is endocytosis. Fluid endocytosis might contribute significantly to the catabolism of plasma proteins with long-lives like serum albumin. Adsorptive endocytosis is responsible for the rapid clearance of complexes of some plasma proteins with other molecules, like antigen-antibody complexes. Other plasma proteins like certain glycoproteins and lipoproteins are bound to be specific receptors on the surface of endocytosing cells without previous complex formation. Experiments with sucrose-containing labels suggest that certain tissues are specifically involved in endocytosis of some plasma proteins. If endocytosis and renal filtration constitute the main mechanisms for the clearance of plasma proteins, one cannot expect extensive homologies between this process and catabolism of intracellular proteins.

Endocytosis and breakdown of proteins by sinusoidal liver cells.

We studied the uptake of modified forms of bovine pancreatic ribonuclease A (labeled with 125-iodine) by rat liver in vivo. On one hand, these experiments were intended to investigate a possible role of sinusoidal cells in the uptake of plasma proteins; on the other hand, the effect of well-defined modifications of the enzyme on the role of uptake might give us a key to the factors determining life time of plasma proteins. We used nephrectomized rats in most experiments to avoid uptake by the kidneys. Preparations of ribonuclease oligomers prepared by cross-linking with dimethyl-suberimidate enabled us to study a possible relation between uptake and molecular size. This modification does not lead to changes in charge of the protein. Monomer, dimer and polymer fractions were isolated by gel filtration on Sephadex G-75. Of the 11 amino groups in ribonuclease A, 9, 8 and 6 remained unaltered in the monomer, dimer and polymer fraction, respectively. The maintenance of biological activity, the stability of disulfide bonds, and the unchanged susceptibility to endoproteases of the cross-linked products established that gross conformational changes had not occurred. At 1 h after injection, 1% of themonomer, 7% of the dimer and 19% of the polymer were recovered per g of liver protein. Combination of autoradiography, subcellular fractionation, and the determination of labeled ribonuclease derivatives in the spleens showed that the dimer and polymer fractions were mainly present in the lysosomes of sinusoidal cells.

O-(4-Diazo-3,5-di[125I]iodobenzoyl)sucrose, a novel radioactive label for determining organ sites of catabolism of plasma proteins.

A method is described for radiolabelling proteins with O-(4-diazo-3,5-di[125I]iodobenzoyl)sucrose (DD125IBS). When proteins so labelled were degraded within lysosomes, the radioactive fragments were largely retained within the organelle. High specific radioactivities were obtained without changing the properties of the protein. The validity of the method was demonstrated in vivo in rats using the short-lived protein lactate dehydrogenase, isoenzyme M4, and the long-lived protein bovine serum albumin. Derivatization with DD125IBS did not alter the clearance of either protein. Uptake of DD125IBS-labelled lactate dehydrogenase, isoenzyme M4, by liver and spleen of rats was determined. Radioactivity in these tissues increased up to about 2 h after injection (at this time the protein has been almost completely cleared from the blood) and subsequently declined with a half-life of approx. 20 h. After differential fractionation of liver, radioactivity was largely found in the mitochondrial and lysosomal fraction. The results of these studies establish that DD125IBS covalently coupled to plasma proteins should be a useful radioactive tracer for identifying the tissue and cellular sites of catabolism of relatively long-lived circulating proteins.

Endocytosis and breakdown of 125I-labelled lactate dehydrogenase isoenzyme M4 by rat liver and spleen in vivo.

1. Porcine lactate dehydrogenase isoenzyme M4 was labelled with 125I and injected intravenously into rats. Enzyme activity and radioactivity in plasma were cleared in an identical way with a half-life of about 30 min. This half-life was the same as that of unlabelled enzyme. 2. Uptake of label by liver and spleen was determined. Radioactivity in these tissues increased up to about 13 min after injection and subsequently declined. Radioautography indicated uptake of the enzyme by sinusoidal liver cells (probably Kupffer cells) and by spleen macrophages. After differential fractionation of liver, acid-precipitable radioactivity was largely found in the light mitochondrial and microsomal fractions, suggesting localization in lysosomes and endosomes respectively. 3. The amount of acid-soluble radioactive breakdown products in plasma started to rise between 7 and 15 min after injection. Breakdown in liver and spleen was retarded by previous injection of suramin, an inhibitor of lysosomal proteolysis. 4. The contribution of liver and spleen towards the clearance of the enzyme could be calculated from its half-life in plasma and its uptake by the organs within the first 13 min period after injection. Our results indicate that about 65% and 12% of the injected dose was taken up, and subsequently broken down, by liver and spleen respectively. 5. Unlabelled porcine lactate dehydrogenase isoenzyme H4 showed a plasma half-life of about 8 h. This isoenzyme is therefore endocytosed by liver at a much slower rate than isoenzyme M4 (if it is taken up at all).

Plasma clearance and endocytosis of cytosolic malate dehydrogenase in the rat.

1. Pig heart cytosolic malate dehydrogenase was radiolabelled with O-(4-diazo-3,5-di-[125I]iodobenzoyl)sucrose and intravenously injected into rats. Enzyme activity and radioactivity were cleared from plasma identically, with first-order kinetics, with a half-life of about 30 min. 2. The tissue distribution of radioactivity was determined at 2 h after injection. All injected radioactivity was recovered from the tissues. A high percentage of the injected dose was found in liver (37%), spleen (6%) and bone including marrow (19%). 3. Radioactivity in liver and spleen increased up to 2 h after injection and subsequently declined, with a half-life of about 20 h. 4. After differential fractionation of liver, radioactivity was largely found in the mitochondrial and lysosomal fraction. 5. Liver cells were isolated 1 h after injection of labelled enzyme. We found that Kupffer cells, endothelial cells and parenchymal cells had endocytosed the enzyme at rates corresponding to 2725, 94 and 63 ml of plasma/day per g of cell protein respectively. 6. Radioautography indicated that in spleen and bone marrow the enzyme is mainly taken up by macrophages. 7. Internalization of the enzyme by liver, spleen and bone marrow was saturable. This indicates that the enzyme is taken up in these tissues by adsorptive endocytosis. 8. The present results closely resemble those obtained previously for the mitochondrial isoenzyme of malate dehydrogenase and for lactate dehydrogenase M4. Since those enzymes are positively charged at physiological pH, whereas cytosolic malate dehydrogenase is negative, net charge cannot be the major factor determining the rate of uptake of circulating enzymes by reticuloendothelial macrophages, as has been suggested in the literature [Wachsmuth & Klingmüller (1978) J. Reticuloendothel. Soc. 24, 227-241].

Plasma clearance of mitochondrial aspartate aminotransferase in the rat: competition with mitochondrial malate dehydrogenase.

In previous experiments we have shown that the rapid clearance in rats of alcohol dehydrogenase, lactate dehydrogenase M4, and the mitochondrial and cytosolic isoenzymes of malate dehydrogenase is largely due to endocytosis by macrophages in liver, spleen and bone marrow. Competition experiments indicated that the dehydrogenases as well as adenylate kinase and creatine kinase MM are endocytosed via the same receptor. We suggested that this receptor contains a group with affinity for the nucleotide-binding sites of the enzymes. We now demonstrate that competition also occurs between mitochondrial malate dehydrogenase and mitochondrial aspartate aminotransferase, which does not require a nucleotide for its activity. At low doses, mitochondrial aspartate aminotransferase was cleared following first-order kinetics (half-life: 19 min). Simultaneous injection of a high dose of mitochondrial malate dehydrogenase strongly retarded the clearance of the aminotransferase. These results make unlikely the hypothesis that a nucleotide-binding site is involved in recognition of enzymes by macrophages.

Infection of mice with lactate dehydrogenase-elevating virus destroys the subpopulation of Kupffer cells involved in receptor-mediated endocytosis of lactate dehydrogenase and other enzymes.

In previous experiments in rats, we have shown that the rapid plasma clearance of a number of clinically important enzymes is due to receptor-mediated endocytosis by Kupffer cells and other resident macrophages. Others have shown that infection of mice with lactate dehydrogenase-elevating virus, a virus that proliferates in macrophages, leads to reduced plasma elimination of these enzymes. This paper integrates these two sets of experiments. Plasma elimination of intravenously injected, radioactively labeled lactate dehydrogenase M4 and mitochondrial malate dehydrogenase in mice was shown to be caused in part by uptake in liver, spleen and bone. Uptake of lactate dehydrogenase M4 by these tissues was, to a large extent, saturable and the two dehydrogenases competitively inhibited each other's clearance. These results suggest that, also in mice, these enzymes are partly cleared from plasma by endocytosis by way of a common receptor on cells (probably macrophages) from liver, spleen and bone marrow. Morphometrical data showed that normal mouse liver contains 23 x 10(6) Kupffer cells/cm3. This number was reduced to about 30% of that of controls 24 hr after infection of mice with lactate dehydrogenase-elevating virus but returned to normal within the next 9 days. The saturable component of uptake of lactate dehydrogenase M4 by liver, spleen and bone had disappeared 24 hr after infection with the virus, and did not return after the Kupffer cell population had recovered. Our findings suggest that lactate dehydrogenase M4 is, to a large extent, removed from the circulation by way of a receptor on a subpopulation of macrophages that is permissive for replication of lactate dehydrogenase-elevating virus.(ABSTRACT TRUNCATED AT 250 WORDS)