Expression and characterization of recombinant human angiotensin I-converting enzyme. Evidence for a C-terminal transmembrane anchor and for a proteolytic processing of the secreted recombinant and plasma enzymes.

Chinese hamster ovary (CHO) cells have been transfected with either a full-length cDNA encoding human angiotensin I-converting enzyme (kininase II; EC 3.4.15.1) (ACE) or a mutated cDNA, in which the last C-terminal 47 amino acids, including the putative transmembrane domain, are not translated. Cell lines expressing high levels of the wild-type ACE or the mutant were established. The cells transfected with the wild-type cDNA (CHO-ACE) express a membrane-bound ectoenzyme with an intracellular C terminus, as shown by indirect immunofluorescence using an antiserum (28A7) raised against a synthetic peptide corresponding to the deduced C terminus of ACE. This enzyme is structurally, immunologically, and enzymatically identical to human kidney ACE. In addition, CHO-ACE cells also produce a secreted form of the enzyme. Neither this secreted form nor the enzyme purified from human plasma is recognized by the antiserum 28A7, indicating that they undergo a truncation in the C-terminal region. On the other hand, the transfected cells expressing the C-terminally truncated mutant (CHO-ACE delta COOH) do not retain ACE in the plasma membrane, but secrete it into the medium. These results indicate that ACE is anchored to the plasma membrane by the predicted C-terminal transmembrane domain, and the secreted form is derived from the membrane-bound form by a post-translational proteolytic cleavage of the C-terminal region.     

Generation of a biologically active, secreted form of human thyroid peroxidase by site-directed mutagenesis

Using site-directed mutagenesis, we introduced two stop codons immediately upstream of the putative transmembrane domain in human thyroid peroxidase (hTPO) cDNA, truncating the carboxyl terminus of hTPO (933 amino acids) by 85 residues. Mutated hTPO cDNA, inserted into a eukaryotic expression vector, was stably transfected into Chinese hamster ovary (CHO) cells. Immunoprecipitation of cellular 35S-methionine-labeled proteins with Hashimoto's serum revealed a 105-101 kilodalton doublet. In contrast, cells transfected with wild-type hTPO yielded a 112-105 kilodalton doublet. In pulse-chase experiments, CHO cells expressing the truncated hTPO protein secreted immunoprecipitable TPO into the culture medium after 4 h of chase, with levels accumulating progressively over a 24-h period. In contrast, CHO cells expressing wild-type hTPO released no immunoprecipitable TPO into the culture medium. The secreted, truncated form of hTPO appeared as a single band of lesser electrophoretic mobility, as opposed to the doublet expressed within cells. TPO enzymatic activity was present in conditioned media from CHO cells transfected with the mutated hTPO, but was absent in media from cells expressing wild-type hTPO. The stability of the mutated protein appeared similar to that of wild-type hTPO. In summary, we have generated a mutated, secreted form of hTPO that is enzymatically active and immunologically intact. Our data confirm the existence of a transmembrane domain in hTPO, and that hTPO is predominantly an enzyme with an extracellular orientation. The secreted form of hTPO has the potential for generating large amounts of soluble TPO protein for use in future structural and immunological studies.     

Cloning and characterization of a secreted form of angiotensin-converting enzyme 2

Angiotensin-converting enzyme 2 (ACE2) is a newly discovered, membrane-bound aminopeptidase responsible for the production of vasodilatory peptides such as angiotensin 1-7 (Ang 1-7). Thus, ACE2 is important in counteracting the adverse, vasoconstrictor effects of angiotensin II (Ang II). The objective of the present study was to clone and characterize a constitutively secreted form of ACE2 as a prelude to an investigation into its therapeutic potential in hypertension. A truncated form of ACE2 was cloned into a lentiviral vector behind the human elongation factor 1 alpha promoter (lenti-shACE2). Transfection experiments demonstrated that secreted human ACE2 (shACE2) was secreted constitutively into the medium. The kinetic properties of shACE2 were comparable to the human recombinant enzyme (rACE2). Transduction of human coronary artery endothelial cells and rat cardiomyocytes with lenti-shACE2 showed a significant secretion of the enzyme into the medium compared to its native, membrane-bound homolog (human ACE2 [hACE2]). In addition, systemic administration of lenti-shACE2 into neonatal rats resulted in a eightfold increase in ACE2 activity in the serum above control values. These observations establish that lenti-shACE2 can be used to transduce cardiovascularly relevant cells for the secretion of functional ACE2 enzyme both in vitro and in vivo. Collectively, these results set the stage for the use of these vectors to investigate the consequences of ACE2 over-expression in the pathogenesis of hypertension.     

Purification and characterization of recombinant human testis angiotensin-converting enzyme expressed in Chinese hamster ovary cells.

Enzymatically active human testis angiotensin-converting enzyme (ACE) was expressed in Chinese hamster ovary (CHO) cells stably transfected with each of three vectors: p omega-ACE contains a full-length testis ACE cDNA under the control of a retroviral promoter; and pLEN-ACEVII and pLEN-ACE6/5, in which full-length and membrane anchor-minus testis ACE cDNAs, respectively, are under the control of the human metallothionein IIA promoter and SV40 enhancer. In every case, active recombinant human testis ACE (hTACE) was secreted in a soluble form into the culture media, up to 2.4 mg/liter in the media of metal-induced, high-producing clones transfected with one of the pLEN vectors. In addition, membrane-bound recombinant enzyme was recovered from detergent extracts of cell pellets of CHO cells transfected with either p omega-ACE or pLEN-ACE-VII. Recombinant converting enzyme was purified to homogeneity by single-step affinity chromatography of conditioned media and detergent-extracted cell pellets in 85 and 70% overall yield, respectively. Purified hTACE from all sources comigrated with the native testis isozyme on sodium dodecyl sulfate-polyacrylamide gel electrophoresis with M(r) approximately 100 kDa. The native and recombinant proteins cross-reacted equally with anti-human kidney ACE antiserum on Western blotting. The catalytic activity of recombinant angiotensin-converting enzyme, in terms of angiotensin I and 2-furanacryloyl-Phe-Gly-Gly hydrolysis, chloride activation, and lisinopril inhibition, was essentially identical to that of the native enzyme. The facile recovery in high yield of fully active hTACE from the media of stably transfected CHO cells provides a suitable system for investigating structure-function relationships in this enzyme.     

Specific cellular proteins associate with angiotensin-converting enzyme and regulate its intracellular transport and cleavage-secretion

Angiotensin-converting enzyme (ACE) is an extensively glycosylated type I ectoprotein anchored in the plasma membrane by a hydrophobic transmembrane domain. In tissue culture as well as in vivo, the extracellular domain of ACE is released into the culture medium by a regulated proteolytic cleavage. To identify the cellular proteins that regulate ACE processing and cleavage-secretion, ACE-bound proteins were purified by affinity chromatography and characterized by microsequencing and Western blotting. One protein was identified as ribophorin and another as immunoglobulin-binding protein (BiP), a chaperone. Metabolic labeling and immunoprecipitation of ACE confirmed its interaction with BiP. Overexpression of BiP inhibited ACE secretion, an effect accentuated by the expression of an enzymatically inactive mutant BiP. This inhibition was caused by the retention of ACE precursors by BiP in the endoplasmic reticulum, as revealed by immunoprecipitation and immunofluorescence experiments. However, treatment with a phorbol ester, phorbol 12-myristate 13-acetate, enhanced ACE secretion even from cells overexpressing BiP. Western blot analysis of ACE-associated proteins with antibodies to protein kinase C (PKC) revealed the presence of its specific isozymes. Treatment with phorbol 12-myristate 13-acetate caused marked reduction in ACE association of selective PKC species. Thus, our studies have identified PKC and BiP as two proteins that directly interact with ACE and modulate its cell-surface expression and cleavage-secretion.     

Amyloid beta-protein is degraded by cellular angiotensin-converting enzyme (ACE) and elevated by an ACE inhibitor

Human genetic data have associated angiotensin-converting enzyme (ACE) with Alzheimer disease (AD), and purified ACE has been reported to cleave synthetic amyloid beta-protein (Abeta) in vitro. Whether deficiency in ACE activity, arising from genetic alteration or pharmacological inhibition, can decrease Abeta degradation and allow Abeta accumulation in intact cells is unknown. We cloned ACE from human neuroblastoma cells and showed that it had posttranslational processing and enzymatic activity typical of the endogenous protease. Cellular expression of ACE promoted degradation of naturally secreted Abeta40 and Abeta42, leading to significant clearance of both species. Using site-directed mutagenesis, we determined that both active sites within ACE contribute to Abeta clearance, and an ACE construct bearing mutations in each catalytic domain had no effect on Abeta levels. Pharmacological inhibition of ACE with a widely prescribed drug, captopril, promoted the accumulation of cell-derived Abeta in the media of beta-amyloid precursor-protein expressing cells. Together, these results show that ACE can lower the levels of secreted Abeta in living cells and that this effect is blocked by inhibiting the protease's activity with an ACE inhibitor. This work, combined with the genetic studies, supports the hypothesis that ACE may modulate the susceptibility to and progression of AD via degradation of Abeta. Our data encourage further analyses of the ACE gene for disease association and raise the question of whether currently prescribed ACE inhibitors could elevate cerebral Abeta levels in humans.     

Phosphatidylinositol linkage of a truncated form of the platelet-derived growth factor receptor

The platelet-derived growth factor (PDGF) receptor is usually anchored to the plasma membrane through a membrane-spanning hydrophobic amino acid sequence that splits the molecule into two approximately equal pieces, an amino-terminal external domain that contains the binding site for PDGF and a carboxyl-terminal cytoplasmic domain that includes the tyrosine kinase coding sequences. Here we report the expression of a truncated PDGF receptor that consists of the extracellular domain without the transmembrane and cytoplasmic domains. Unexpectedly, this form of the receptor that lacks a hydrophobic membrane-anchoring sequence was bound to the membrane and was not secreted into the culture media. Conventional methods to dissociate noncovalent protein-protein interactions failed to release the protein from the membrane. When the transmembrane and cytoplasmic sequences were artificially deleted from the PDGF receptor, the truncated extracellular domain was anchored to the membrane through phospholipids and could be released by phospholipase C treatment. This truncated form of the receptor bound PDGF with an affinity 5-20-fold lower than the full-length receptor.     

Glucose-induced degradation of the MDH2 isozyme of malate dehydrogenase in yeast.

MDH2, the nonmitochondrial isozyme of malate dehydrogenase in Saccharomyces cerevisiae, was determined to be a target of glucose-induced proteolytic degradation. Shifting a yeast culture growing with acetate to medium containing glucose as a carbon source resulted in a 25-fold increase in turnover of MDH2. A truncated form of MDH2 lacking amino acid residues 1-12 was constructed by mutagenesis of the MDH2 gene and expressed in a haploid yeast strain containing a deletion disruption of the corresponding chromosomal gene. Measurements of malate dehydrogenase specific activity and determination of growth rates with diagnostic carbon sources indicated that the truncated form of MDH2 was expressed at authentic MDH2 levels and was fully active. However, the truncated enzyme proved to be less susceptible to glucose-induced proteolysis, exhibiting a 3.75-fold reduction in turnover rate following a shift to glucose medium. Rates of loss of activity for other cellular enzymes known to be subject to glucose inactivation were similarly reduced. An extended lag in attaining wild type rates of growth on glucose measured for strains expressing the truncated MDH2 enzyme represents the first evidence of a selective advantage for the phenomenon of glucose-induced proteolysis in yeast.     

Angiotensin-converting enzyme 2 (ACE2), but not ACE, is preferentially localized to the apical surface of polarized kidney cells

Angiotensin-converting enzyme-2 (ACE2) is a homologue of angiotensin-I converting enzyme (ACE), the central enzyme of the renin-angiotensin system (RAS). ACE2 is abundant in human kidney and heart and has been implicated in renal and cardiac function through its ability to hydrolyze Angiotensin II. Although ACE2 and ACE are both type I integral membrane proteins and share 61% protein sequence similarity, they display distinct modes of enzyme action and tissue distribution. This study characterized ACE2 at the plasma membrane of non-polarized Chinese hamster ovary (CHO) cells and polarized Madin-Darby canine kidney (MDCKII) epithelial cells and compared its cellular localization to its related enzyme, ACE, using indirect immunofluorescence, cell-surface biotinylation, Western analysis, and enzyme activity assays. This study shows ACE2 and ACE are both cell-surface proteins distributed evenly to detergent-soluble regions of the plasma membrane in CHO cells. However, in polarized MDCKII cells under steady-state conditions the two enzymes are differentially expressed. ACE2 is localized predominantly to the apical surface ( approximately 92%) where it is proteolytically cleaved within its ectodomain to release a soluble form. Comparatively, ACE is present on both the apical ( approximately 55%) and basolateral membranes ( approximately 45%) where it is also secreted but differentially; the ectodomain cleavage of ACE is 2.5-fold greater from the apical surface than the basolateral surface. These studies suggest that both ACE2 and ACE are ectoenzymes that have distinct localization and secretion patterns that determine their role on the cell surface in kidney epithelium and in urine.     

A soluble form of the glycolipid-anchored receptor for urokinase-type plasminogen activator is secreted from peripheral blood leukocytes from patients with paroxysmal nocturnal hemoglobinuria.

The cellular urokinase-type plasminogen-activator (uPA) receptor (uPAR) is a glycolipid-anchored membrane protein thought to be involved in pericellular proteolysis during cell migration and tumor invasion. In the present study, we have identified and characterized two soluble forms of uPAR which have retained their ligand-binding capability. One variant was generated in vitro by treatment of intact normal cells with either a phosphatidylinositol-specific phospholipase C (PLC) or endoproteinase Asp-N. The other soluble uPAR variant was secreted in vivo from peripheral blood leukocytes affected by the stem-cell disorder paroxysmal nocturnal hemoglobinuria (PNH), and was found in the plasma from these PNH patients as well as in the conditioned medium from cultured PNH leukocytes. Under normal conditions, we find no evidence for any shedding or secretion of a soluble uPA-binding counterpart to human uPAR in plasma. Unlike normal leukocytes, the PNH-affected cells do not express uPAR on the cell surface, although they do contain apparently normal levels of uPAR-specific mRNA. The secreted uPAR derived from PNH cells has a mobility in SDS/PAGE that is slightly higher than that of uPAR solubilized by PtdIns-specific PLC or detergent, but resembles that of a truncated, recombinant uPAR variant, which has its C-terminus close to the proposed glycolipid-attachment site, suggesting that the secreted protein has been proteolytically processed for glycolipid attachment. The presence in plasma from PNH patients of such a secreted, hydrophilic form of uPAR lends support to the hypothesis that the lesion underlying the PNH disorder resides either in glycolipid biosynthesis or in the function of an as-yet-unidentified transamidating enzyme assumed to cleave and assemble the truncated uPAR with the preformed glycolipid moiety.     

Mammalian plasma membrane ecto-nucleoside triphosphate diphosphohydrolase 1, CD39, is not active intracellularly. The N-glycosylation state of CD39 correlates with surface activity and localization

CD39 is a member of the membrane-bound ecto-nucleoside triphosphate diphosphohydrolase family. The active site for native CD39 is located on the outer surface of the cellular plasma membrane; however, it is not yet known at what stage this enzyme becomes active along the secretory pathway to the plasma membrane. In this study, sucrose density fractionations performed on CD39-transfected COS-7 cell membranes suggest that CD39 activity resides primarily in the plasma membrane. Furthermore, we have created recombinant, soluble versions of CD39, one that is secreted and others that are retained in the endoplasmic reticulum, to demonstrate that CD39 is not active until it reaches the plasma membrane both in yeast and COS-7 cells. Moreover, the secreted active soluble CD39 in COS-7 cells is found to receive a higher degree of N-glycan addition than the inactive form retained intracellularly. When COS-7 cells were treated with tunicamycin to prevent N-glycosylation, soluble CD39 was not detected in the extracellular medium and remained inactive intracellularly. Surface biotinylation analysis also revealed that surface-expressed wild type CD39 receives a higher degree of N-glycosylation than intracellular forms and that inhibition of N-glycosylation prevents its plasma membrane localization. In addition, both intact and digitonin-permeablized COS-7 cells transfected with CD39 possess similar ecto-ATPase activities, further supporting the conclusion that only surface-expressed CD39 is enzymatically active. All of these data suggest that intracellular CD39 is inactive and that only a fully glycosylated CD39 has apyrase activity and is localized at the cell surface.     

Release of angiotensin I-converting enzyme by endothelial cells in vitro

Bovine fetal aortic endothelial cells cultured in serum-containing medium accumulate angiotensin I-converting enzyme (ACE) activity and also release it into the culture medium. Following subcultivation of a confluent culture using trypsin-EDTA, cellular ACE activity falls 50% within 8 h, but no ACE activity is detected in the medium, suggesting intracellular loss of the enzyme activity. ACE activity reappears in both the cell lysate and culture medium after the culture becomes confluent. The rate of accumulation of ACE activity released into the medium is always greater than that for cellular activity. For example, 21 days following subcultivation 80-85% of the total culture activity is detected in the medium. Both cellular and medium-associated ACE decrease proportionately as the culture progresses through its in vitro lifespan.     

Determination of angiotensin I-converting enzyme activity in cell culture using fluorescence resonance energy transfer peptides

An assay using fluorescence resonance energy transfer peptides was developed to assess angiotensin I-converting enzyme (ACE) activity directly on the membrane of transfected Chinese hamster ovary cells (CHO) stably expressing the full-length somatic form of the enzyme. The advantage of the new method is the possibility of using selective substrates for the two active sites of the enzyme, namely Abz-FRK(Dnp)P-OH for somatic ACE, Abz-SDK(Dnp)P-OH for the N domain, and Abz-LFK(Dnp)-OH for the C domain. Hydrolysis of a peptide bond between the donor/acceptor pair (Abz/Dnp) generates detectable fluorescence, allowing quantitative measurement of the enzymatic activity. The kinetic parameter K(m) for the hydrolysis of the three substrates by ACE in this system was also determined and the values are comparable to those obtained using the purified enzyme in solution. The specificity of the activity was demonstrated by the complete inhibition of the hydrolysis by the ACE inhibitor lisinopril. Therefore, the results presented in this work show for the first time that determination of ACE activity directly on the surface of intact CHO cells is feasible and that the method is reliable and sensitive. In conclusion, we describe a methodology that may represent a new tool for the assessment of ACE activity which will open the possibility to study protein interactions in cells in culture.     

Molecular and enzymatic properties of furin, a Kex2-like endoprotease involved in precursor cleavage at Arg-X-Lys/Arg-Arg sites.

We have recently shown that furin, a mammalian homologue of the yeast precursor-processing endoprotease Kex2, is involved in precursor cleavage at sites marked by the Arg-X-Lys/Arg-Arg motif within the constitutive secretory pathway. In this study, we analyzed molecular and enzymatic properties of furin expressed in Chinese hamster ovary cells using gene transfer techniques. COOH-terminal truncation analyses indicate that the polypeptide region significantly conserved among the Kex2 family members is required for the endoprotease activity of furin, while the COOH-terminal unconserved region containing the Cys-rich domain and the transmembrane domain is dispensable. A mutant of furin truncated up to the transmembrane domain from the COOH-terminus was secreted into the culture medium as an active form. The sequence requirements for precursor cleavage of this truncated furin determined in vitro were similar to those of wild-type furin determined by expression studies in cultured cells. It had a strong resemblance to the Kex2 protease in the inhibitor profile and pH dependency. These observations support the notion that furin is the endogenous endoprotease involved in precursor cleavage at Arg-X-Lys/Arg-Arg sites.     

Secretion of the extracellular domain of the human insulin receptor from insect cells by use of a baculovirus vector.

To explore the utility of the baculovirus/insect-cell system for the expression of a soluble secreted human insulin-receptor (hIR) extracellular ligand-binding domain, we have engineered a recombinant virus encoding an hIR deletion mutant which is truncated eight residues from the beginning of the predicted transmembrane domain (i.e. 921 residues). Within 24 h after infection of Sf9 cells with virus, insulin-binding activity begins to accumulate in the culture medium, and reaches a maximum between 48 and 72 h. The intracellular transit and processing of this secreted receptor, designated 'AchIR01', is quite slow. After 24 h in pulse-chase experiments approximately 50% of the metabolically labelled protein is still inside the cell. This protein accumulates as a non-cleaved hIR precursor which is glycosylated, but the carbohydrate is entirely endoglycosidase H (endoH)-sensitive (i.e. high mannose). Approximately one-half of the receptor in the culture medium (i.e. approximately 25% of the total) is in the form of non-cleaved precursor, and about one half of its carbohydrate chains are now endoH-resistant. The remainder of the protein is proteolytically processed hIR (alpha-plus truncated beta-subunits). None of these hIR species exhibit O-linked carbohydrate. Only the processed form of the receptor in the medium binds insulin. This insulin-binding protein is secreted as a dimer (alpha beta)2, and binds insulin with an affinity which is comparable with that of both the wild-type hIR as well as the secreted form of the hIR expressed in mammalian cells. Despite the rather inefficient processing and altered glycosylation of the AchIR01 protein in insect cells, this high-affinity insulin-binding protein accumulates in the medium at levels (mg/litre) of about 100 times that achieved in a mammalian-cell system.     

A small region in the angiotensin-converting enzyme distal ectodomain is required for cleavage-secretion of the protein at the plasma membrane

Both germinal and somatic isoforms of ACE are type I ectoproteins expressed on the cell surface from where the enzymatically active ectodomains are released to circulation by a regulated cleavage-secretion process. Our previous studies have shown that ACE-secretase activity is regulated by the ACE distal ectodomain and not by sequences at or around the cleavage site. In the current study we have identified that the ACE residues encompassing 343 to 655 of the germinal form are needed for its cleavage-secretion. To narrow down this region further, we have examined the cleavage-secretion of ACE-CD4 chimeric proteins in mammalian cells and Pichia pastoris. These experiments identified five residues (HGEKL) in the ACE region of the chimeric proteins that were essential for their cleavage-secretion. When the corresponding residues were substituted by alanine in native germinal and somatic ACE, the mutant proteins were not cleaved, although they were displayed on the cell surface and enzymatically active. These results demonstrated that a small region in the ectodomain of ACE is required for its cleavage at the juxtamembrane domain. This conclusion was further supported by our observation that secreted ACE inhibited cell-bound ACE cleavage-secretion, although the secreted form did not contain the cleavage site.     

Recombinant expression of a secreted form of the atrial natriuretic peptide clearance receptor

A general structure for the atrial natriuretic peptide clearance receptor (ANP C-receptor) has been proposed based on hydropathicity analysis of the deduced amino acid sequence of this membrane protein (Fuller, F., Porter, J.G., Arfsten, A., Miller, J., Schilling, J., Scarborough, R.M., Lewicki, J.A., and Schenk, D.B. (1988) J. Biol. Chem. 263, 9395-9401). The ANP C-receptor is believed to possess a large amino-terminal extracellular domain (436 amino acids), a single hydrophobic transmembrane anchor (23 amino acids), and a short cytoplasmic tail (37 amino acids). As a means of testing the structure and proposed cellular orientation of this protein, we have employed the technique of in vitro mutagenesis to prepare a receptor mutant (anc-) lacking the transmembrane and cytoplasmic domains. Expression of this mutant in mammalian cells using a vaccinia virus vector results in secretion of a truncated soluble form of the ANP C-receptor which binds native ANP and synthetic ANP analogs with a specificity similar to that of the native ANP C-receptor. In contrast to the native ANP C-receptor that exists predominantly as a homodimer on the cell surface, the secreted receptor exists as a monomeric species. The results are consistent with the proposed structure of this receptor with the amino-terminal domain containing the ANP-binding site oriented extracellular to the plasma membrane. In addition, these data demonstrate that the receptor does not require association with the plasma membrane or its native dimeric configuration in order to bind ANP ligands with high affinity and specificity.     

Liver cells secrete the plasma form of platelet-activating factor acetylhydrolase.

Platelet-activating factor (PAF) is a phospholipid (1-O-alkyl-2-acetyl-sn-glycero-3-phosphocholine) with diverse physiological effects. It has been implicated as a mediator of inflammation, allergy, shock, and thrombosis. Plasma contains an enzyme, PAF acetylhydrolase, that catalyzes the degradation of PAF, and the level of this enzyme may regulate the concentration of PAF in the blood and extracellular spaces under some conditions. Thus, the cellular source(s) of this enzyme and the factors that regulate its synthesis and secretion are issues that may have important physiological and pathological implications. We found that cultures of Hep G2, a human hepatocarcinoma line, secreted PAF acetylhydrolase activity. Optimal secretion occurred in medium that contained serum, and the newly secreted PAF acetylhydrolase was associated with high density and low density lipoproteins (LDL and HDL, respectively), just as the enzyme is in plasma. In the absence of serum. PAF acetylhydrolase was secreted with a particle that had a density similar to HDL. Apolipoproteins B and E were found in the same fractions. We tested the effects of a variety of hormones on the secretion of PAF acetylhydrolase and found that secretion was inhibited by 17 alpha-ethynylestradiol with a maximal effect at 30 microM. This may account for the observation of others that estrogens reduce the activity of PAF acetylhydrolase in the plasma. The PAF acetylhydrolase secreted by Hep G2 cells appeared to be identical to the enzyme in human plasma based on substrate specificity, association with LDL and HDL, response to inhibitors, and reactivity with antibodies against the plasma PAF acetylhydrolase. In conclusion, we have demonstrated that hepatocytes in culture secrete a PAF acetylhydrolase that is apparently identical to the plasma form. The secretion is constitutive but may also be regulated in response to hormonal stimulation.