Endoprotease PACE4 is Ca2+-dependent and temperature-sensitive and can partly rescue the phenotype of a furin-deficient cell strain.

PACE4 is a member of the eukaryotic subtilisin-like endoprotease family. The expression of human PACE4 in RPE.40 cells (furin-null mutants derived from Chinese hamster ovary K1 cells) resulted in the rescue of a number of wild-type characteristics, including sensitivity to Sindbis virus and the ability to process the low-density-lipoprotein receptor-related protein. Expression of PACE4 in these cells failed to restore wild-type sensitivity to Pseudomonas exotoxin A. Co-expression of human PACE4 in these cells with either a secreted form of the human insulin pro-receptor or the precursor form of von Willebrand factor resulted in both proproteins being processed; RPE.40 cells were unable to process either precursor protein in the absence of co-expressed PACE4. Northern analysis demonstrated that untransfected RPE.40 cells express mRNA species for four PACE4 isoforms, suggesting that any endogenous PACE4 proteins produced by these cells are either non-functional or sequestered in a compartment outside of the secretory pathway. In experiments in vitro, PACE4 processed diphtheria toxin and anthrax toxin protective antigen, but not Pseudomonas exotoxin A. The activity of PACE4 in vitro was Ca2+-dependent and, unlike furin, was sensitive to temperature changes between 22 and 37 degrees C. RPE.40 cells stably expressing human PACE4 secreted an endoprotease with the same Ca2+ dependence and temperature sensitivity as that observed in membrane fractions of these cells assayed in vitro. These results, in conjunction with other published work, demonstrate that PACE4 is an endoprotease with more stringent substrate specificity and more limited operating parameters than furin.

Endoprotease activities other than furin and PACE4 with a role in processing of HIV-I gp160 glycoproteins in CHO-K1 cells.

We addressed the question of whether furin is the endoprotease primarily responsible for processing the human immunodeficiency virus type I (HIV-I) envelope protein gp160 in mammalian cells. The furin-deficient Chinese hamster ovary (CHO)-K1 strain RPE.40 processed gp160 as efficiently as wild-type CHO-K1 cells in vivo. Although furin can process gp160 in vitro, this processing is probably not physiologically relevent, because it occurs with very low efficiency. PACE4, a furin homologue, allowed processing of gp160 when both were expressed in RPE.40 cells. Further, PACE4 participated in the activation of a calcium-independent protease activity in RPE.40 cells, which efficiently processed the gp160 precursor in vitro. This calcium-independent protease activity was not found in another furin-deficient cell strain, 7.P15, selected from the monkey kidney cell line COS-7.

The low-density-lipoprotein receptor-related protein (LRP) is processed by furin in vivo and in vitro.

The low-density-lipoprotein receptor-related protein (LRP) is a multifunctional receptor involved in the clearance of a large number of diverse ligands, including proteases, protease-inhibitor complexes and lipoproteins. The mature receptor is composed of a 515 kDa and a 85 kDa subunit generated by proteolytic cleavage from a 600 kDa precursor polypeptide in a trans-Golgi compartment. Proteolytic processing occurs C-terminal to the tetrabasic amino acid sequence RHRR, a consensus recognition site for precursor processing endoproteases or convertases. In this study we have identified furin, a subtilisin-type protease, to be necessary for efficient processing of LRP in cells. Furin-deficient RPE.40 cells exhibited an impaired processing of endogenous LRP and of a recombinant soluble form of the receptor containing the processing site. The processing defect in RPE.40 cells could be complemented by expression of furin from a transfected cDNA in cultured cells and by purified furin in vitro. The impaired maturation of LRP in RPE.40 cells did not affect its intracellular transport, and correlated with a slight but consistent reduction in the endocytosis of LRP-specific ligands. These data suggest that proteolytic processing of LRP by furin is not necessary for intracellular trafficking but might be required for normal receptor activity.

Mutations in the elongation factor 2 gene which confer resistance to diphtheria toxin and Pseudomonas exotoxin A. Genetic and biochemical analyses.

Both diphtheria toxin and Pseudomonas exotoxin A inhibit eukaryotic protein synthesis by ADP-ribosylating diphthamide, a posttranslationally modified histidine residue present in the elongation factor 2 (EF-2) protein. Elongation factor 2 cannot be ADP-ribosylated by the toxins unless this histidine is modified. In this report we identify three new point mutations in toxin-resistant alleles of the Chinese hamster ovary cell elongation factor 2 gene. The mutations resulted in amino acid substitutions at positions 584 (serine to glycine), 714 (isoleucine to asparagine), and 719 (glycine to aspartic acid). All three amino acid substitutions prevented the biosynthesis of diphthamide. The amount by which the toxins reduced protein synthesis in each of these mutant cell strains suggested that all three mutations also either impaired the function of EF-2 or reduced its steady state level in the cytoplasm. Western blot analysis showed that equal amounts of EF-2 were present in each of the cell strains, indicating that the mutations impaired the catalytic function of EF-2.

Furin activates Pseudomonas exotoxin A by specific cleavage in vivo and in vitro.

We have demonstrated that the native proenzymatic form of Pseudomonas exotoxin A can be cleaved at its specific activation site by furin in intact Chinese hamster ovary cells or in vitro by furin in isolated membrane fractions from these cells. We have compared the activity of furin in cell membrane fractions with that of purified, recombinant human furin. We have verified that RPE.40, a Pseudomonas toxin-resistant mutant cell strain, is mutant in the fur gene, and we have demonstrated that these cells are deficient in cleavage of the toxin. We have also determined that this cleavage of Pseudomonas toxin by furin takes place at the authentic activation site to release the 37-kDa active fragment.

Defective processing of the insulin receptor in an endoprotease-deficient Chinese hamster cell strain is corrected by expression of mouse furin.

Characterization of an endoprotease-deficient mutant Chinese hamster ovary (CHO) cell, designated RPE.40, revealed that it bound less than 10% as much insulin as did its parent, CHO-K1. We examined processing of the endogenous insulin receptor in CHO-K1 and RPE.40 cells, and processing of the human insulin receptor expressed in these cells. RPE.40 cells did not process the endogenous insulin proreceptor to its subunit forms, and processed the human insulin proreceptor inefficiently. Accumulation of the proreceptor form of the insulin receptor was seen in both cases. Furin is a mammalian endoprotease that cleaves proproteins at a consensus sequence of basic amino acids found in the insulin proreceptor. Expression of mouse furin in RPE.40 cells restored normal processing of the endogenous and the human insulin receptor in these cells. In addition, expression of mouse furin corrected the reduced binding of insulin in RPE.40 cells, indicating that receptor function as well as processing was restored.

Expression of mouse furin in a Chinese hamster cell resistant to Pseudomonas exotoxin A and viruses complements the genetic lesion.

RPE.40 is a strain of mutated CHO-K1 cells with elevated resistance to Pseudomonas exotoxin A, Sindbis virus, and Newcastle disease virus. Virus resistance is due to an inability to cleave precursor viral membrane glycoproteins and produce infectious virions. Transfection of RPE.40 cells with cDNA for mouse furin causes them to lose all resistance and become as sensitive as wild-type cells to the toxin and viruses. Transfection of RPE.40 cells with cDNA for the related yeast protease Kex2 reduces their resistance to the toxin and viruses, but does not completely eliminate it.

A mutant CHO-K1 strain with resistance to Pseudomonas exotoxin A is unable to process the precursor fusion glycoprotein of Newcastle disease virus.

RPE.40, a mutant strain of CHO-K1 cells isolated for resistance to Pseudomonas exotoxin A and cross-resistant to alphaviruses, is also highly resistant to virulent strains of Newcastle disease virus. The resistance of RPE.40 cells to Newcastle disease virus results from the failure to cleave the viral envelope precursor glycoprotein Fo to fusion glycoprotein F1 at the consensus sequence (Lys/Arg)-Arg-Gln-(Lys/Arg)-Arg.

A mutation in codon 717 of the CHO-K1 elongation factor 2 gene prevents the first step in the biosynthesis of diphthamide.

The histidine residue at position 715 of elongation factor 2 (EF-2) is posttranslationally modified in a series of enzymatic reactions to 2-[3-carboxyamido-3-(trimethylammonio)-propyl]histidine, which has been given the trivial name diphthamide. The diphthamide residue of EF-2 is the target site for ADP ribosylation by diphtheria toxin and Pseudomonas exotoxin A. ADP-ribosylated EF-2 does not function in protein synthesis. EF-2 that has not been posttranslationally modified at histidine 715 is resistant to ADP ribosylation by these toxins. In this report we show that a G-to-A transition in the first position of codon 717 of the EF-2 gene results in substitution of arginine for glycine and prevents addition of the side chain of diphthamide to histidine 715 of EF-2. EF-2 produced by the mutant gene is fully functional in protein synthesis.

Characterization of the endogenous ADP-ribosylation of wild-type and mutant elongation factor 2 in eukaryotic cells.

Anti-[ADP-ribosylated elongation factor 2 (EF-2)] antiserum has been used to immunoprecipitate the modified form of EF-2 from polyoma-virus-transformed baby hamster kidney (pyBHK) cells [Fendrick, J. L. & Iglewski, W. J. (1989) Proc. Natl Acad. Sci. USA 86, 554-557]. This antiserum also immunoprecipitates a 32P-labelled protein of similar size to EF-2 from a variety of primary and continuous cell lines derived from many species of animals. One of these cell lines, chinese hamster ovary CHO-K1 cells was further characterized. The time course of labelling of ADP-ribosylated EF-2 with [32P]orthophosphate was similar in pyBHK cells and in CHO-K1 cells. The kinetics of labelling were more rapid for cells cultured in 2% serum than 10% serum, with incorporation of 32P reaching a maximum at 6 h and 10 h, respectively. EF-2 mutants of pyBHK and CHO-K1 cells resistant to diphtheria-toxin-catalyzed ADP-ribosylation of EF-2 remain sensitive to cellular ADP-ribosylation of EF-2. The 32P-labelled moiety of ADP-ribosylated EF-2 was digested by snake venom phosphodiesterase and the product was identified as AMP. The same 32P-labelled tryptic peptide was modified by toxin in wild-type EF-2 and by the cellular transferase in mutant EF-2. When purified EF-2 from pyBHK cells was incubated with [carbonyl-14C]nicotinamide and diphtheria toxin fragment A, under conditions for reversal of the ADP-ribosylation reaction, [14C]NAD was generated. The results suggest that cellular ADP-ribosylated EF-2 exists in a variety of cell types, and the ribosylated product is identical to that produced by toxin ADP-ribosylation of EF-2, except in diphthamide mutant cells. Studies with the mutant cell lines indicate that the toxin and the cellular transferase, however, recognize different determinants at the ADP-ribose acceptor site in EF-2. The cellular transferase does not require the diphthamide modification of the histidine ring in the amino acid sequence of EF-2 for the transfer of ADP-ribose to the ring. Therefore, we would expect the cellular transferase active site to be similar to, but not identical to, the critical amino acids demonstrated in the active site of diphtheria toxin and Pseudomonas exotoxin A.