Carboxypeptidase N Is Not Protein in the Mouse, Whereas Pro

Pro-Carboxypeptidase R is an Acute Phase Protein in the Mouse, Whereas Carboxypeptidase N Is Not

This information is current as Tomoo Sato, Takashi Miwa, Hiroyasu Akatsu, Noriyuki of September 23, 2021. Matsukawa, Kyoko Obata, Noriko Okada, William Campbell and Hidechika Okada J Immunol 2000; 165:1053-1058; ; doi: 10.4049/jimmunol.165.2.1053

http://www.jimmunol.org/content/165/2/1053 Downloaded from

References This article cites 26 articles, 10 of which you can access for free at: http://www.jimmunol.org/content/165/2/1053.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 23, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2000 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Pro-Carboxypeptidase R is an Acute Phase Protein in the Mouse, Whereas Carboxypeptidase N Is Not1,2

Tomoo Sato,*† Takashi Miwa,*† Hiroyasu Akatsu,† Noriyuki Matsukawa,† Kyoko Obata,* Noriko Okada,* William Campbell,† and Hidechika Okada3*

Carboxypeptidase R (EC 3.4.17.20; CPR) and carboxypeptidase N (EC 3.4.17.3; CPN) cleave carboxyl-terminal arginine and lysine residues from biologically active peptides such as kinins and anaphylatoxins, resulting in regulation of their biological activity. Human proCPR, also known as thrombin-activatable fibrinolysis inhibitor, plasma pro-carboxypeptidase B, and procarboxypeptidase U, is a plasma zymogen activated during coagulation. CPN, however, previously termed kininase I and anaphylatoxin inactivator, is present in a stable active form in plasma. We report here the isolation of mouse proCPR and CPN cDNA clones that can induce their respective enzymatic activities in culture supernatants of transiently transfected cells. Potato carboxypeptidase inhibitor can inhibit carboxypeptidase activity in culture medium of mouse proCPR-transfected cells. The expres- Downloaded from sion of proCPR mRNA in murine liver is greatly enhanced following LPS injection, whereas CPN mRNA expression remains unaffected. Furthermore, the CPR activity in plasma increased 2-fold at 24 h after LPS treatment. Therefore, proCPR can be considered a type of acute phase protein, whereas CPN is not. An increase in CPR activity may facilitate rapid inactivation of inflammatory mediators generated at the site of Gram-negative bacterial infection and may consequently prevent septic shock. In view of the ability of proCPR to also inhibit fibrinolysis, an excess of proCPR induced by LPS may contribute to hypofibrinolysis in patients suffering from disseminated intravascular coagulation caused by sepsis. The Journal of Immunology, 2000, 165: http://www.jimmunol.org/ 1053–1058.

t has been well established that human carboxypeptidase N sin-like enzymes such as thrombin (11), thrombin/thrombomodu- (CPN),4 which consists of two small active subunits and two lin complex (14), and plasmin (11), thereby acquiring enzymatic I large glycosylated subunits, is an important inactivator of activity by which it cleaves carboxyl-terminal arginine and lysine several potent peptides, including anaphylatoxins, kinins, and fi- residues from various peptides. This activity can be completely brinopeptides (1–4). CPN is present in the active form in plasma inhibited by potato carboxypeptidase inhibitor (PCI) at concentra- (5, 6). Mathews et al. (7) has reported familial partial carboxypep- tions at which CPN is not affected (15). Because the activity of by guest on September 23, 2021 tidase N deficiency resulting in angioedema/chronic urticaria, hay CPR is similar to that of CPN, this enzyme has been implicated in fever/asthma, or both. Over the last decade, a new basic car- the inactivation of several natural potent peptides, including ana- boxypeptidase that is generated from its zymogen during coagu- phylatoxins and bradykinin, as well as in the removal of arginine lation was identified independently by a number of groups (8–12). and lysine residues from partially plasmin-degraded fibrin (8, 15, We designated this enzyme carboxypeptidase R (CPR) and its zy- 16). To elucidate the in vivo roles of proCPR and CPN, the es- mogen, proCPR. The activated form appeared to remove arginine tablishment of an animal model is essential. For this purpose we residues (R) more rapidly than lysine residues (K) at the carboxyl identified cDNAs of mouse proCPR and CPN and studied the ef- terminus of substrates compared with CPN (5, 11, 13). It has also fects of LPS on the expression of these enzymes during the in- been termed plasma carboxypeptidase B (11), carboxypeptidase U flammatory process. (CPU) (8), and thrombin-activatable fibrinolysis inhibitor (TAFI) (12). It is known that human proCPR can be hydrolyzed by tryp- Materials and Methods Probes for screening a cDNA library and Northern blot analysis

*Department of Molecular Biology, Nagoya City University Medical School, We obtained two partial mouse sequences similar to that of human plasma Nagoya, Japan; and †Choju Medical Institute, Fukushimura Hospital, Toyohashi, pro-carboxypeptidase B (proCPR) using Entrez (http://www.ncbi.nlm.nih- Japan .gov/Entrez/) from the National Center of Biotechnology Information. The GenBank accession numbers are AA244760 and AA254734. In the same Received for publication December 6, 1999. Accepted for publication May 1, 2000. manner we obtained data on three partial mouse sequences from clones The costs of publication of this article were defrayed in part by the payment of page similar to the human small subunit of CPN. The numbers for these charges. This article must therefore be hereby marked advertisement in accordance sequences are AA238838, AA285540, and AA396985. The sets of primers with 18 U.S.C. Section 1734 solely to indicate this fact. generated based on the above were: mCPR86ϩ,5Ј-CTGCTCTTCCAA 1 This work was supported by a research grant from the Organization for Pharma- GAACCTCC-3Ј; mCPR292Ϫ,5Ј-CCACGTTGTTCATCAGAACG-3Ј; ceutical Safety and Research. mCPR533ϩ,5Ј-GAATCCATGCCAGAGAATGG-3Ј; mCPR740Ϫ,5Ј- ϩ 2 The nucleotide sequence data reported in this paper will appear in the DDBJ/EMBL/ TTGTTCTTGTGAGCAGAGCG-3Ј; mCPN108 ,5Ј-CAAGGTGCA Ϫ GenBank nucleotide sequence databases under accession numbers AB021968 and CAACCAATGC-3Ј; mCPN510 ,5Ј-GAAGTAGGTGTTGAGATCCGG-3Ј; AB021969. mCPN609ϩ,5Ј-GATTCGCTCCTTGAACTTCG-3Ј; and mCPN1118Ϫ,5Ј- Ј 3 Address correspondence and reprint requests to Dr. H. Okada, Department of Mo- GTTCACCTGAAGTGACGTCG-3 . RT-PCR was performed with these sets lecular Biology, Nagoya City University Medical School, Mizuho-cho, Nagoya 467- of primers using mouse liver total RNA. PCR products were subcloned with 8601, Japan. E-mail address: [email protected] a TA cloning kit (Invitrogen, San Diego, CA). The presence of these inserts 4 Abbreviations used in this paper: CPN, carboxypeptidase N; mCPN, mouse CPN; was confirmed by sequencing. Clones were digested with several restric- CPR, carboxypeptidase R; TAFI, thrombin-activatable fibrinolysis inhibitor; CP, car- tion enzymes followed by electrophoresis on a 1.0% agarose gel. The in- boxypeptidase; EST, expressed sequence tags; TM, thrombomodulin; DIC, dissemi- serts were purified using a Prep-A-Gene kit (Bio-Rad, Hercules, CA) and nated intravascular coagulation; PCI, potato carboxypeptidase inhibitor. the probes were labeled CPR1, CPR2, CPN1, and CPN2, respectively.

Molecular cloning and base sequence analysis (Bio-Rad). The mCPN BamHI fragment was also subcloned into BamHI sites of pDR2EF1, and this form was designated m-CPN/pDR2EF1. The Dr. Mayumi Nonaka of our laboratory provided the cDNA library, which ϩ presence and fidelity of m-proCPR/pDR2EF1 and m-CPN/pDR2EF1 were was constructed using the ZAP II vector with poly(A) RNA isolated from confirmed by restriction enzyme digestion and sequencing, respectively. adult BALB/c mouse liver. Recombinants (1 ϫ 106) were plated at 50,000 plaques/137-mm plate. Screening was conducted under stringent condi- Transient expression of the cloned cDNAs in COS cells tions by the plaque hybridization method using each of the [␣-32P]dCTP- labeled probes (CPR1, CPR2, CPN1, and CPN2) described above. Hybrid- COS-7 cells derived from the kidney of an African green monkey were 2 ization was performed at 65°C for 16 h in a buffer containing 50 mM Tris cultured in 25-cm flasks in DMEM containing 10% FBS, 2 mM L-glu- (pH 7.4), 10ϫ Denhardt’s solution, 1 M NaCl, 10 mM EDTA, 0.1% SDS, tamine, 1 mM sodium pyruvate, 50 U/ml penicillin, 50 ␮g/ml streptomy- and 0.1 mg/ml denatured salmon sperm DNA (Wako, Osaka, Japan). Fil- cin, and 1 ␮g/ml fungizone in a 37°C humidified incubator (95% air/5% ϫ ters were washed twice with 2 SSC (0.3 M NaCl and 30 mM sodium CO2 atmosphere). The m-proCPR/pDR2EF1 and m-CPN/pDR2EF1 ex- citrate, pH 7.0) containing 0.1% SDS at 65°C for 30 min and once with pression vectors were transfected into COS-7 cells using Lipofectamine 0.2ϫ SSC containing 0.1% SDS at room temperature for 60 min. The dried reagent (Life Technologies) in a 24-well culture plate. When the cells were filters were then exposed for 16 h at Ϫ80°C to x-ray films (X-OMAT AR, 60–80% confluent, the conditioned medium in each well was replaced Eastman Kodak, Rochester, NY). The pBluescript phagemids were excised with 300 ␮l of Opti-MEM I Reduced Serum Medium (Life Technologies) from positive clones using R408 helper phage (Stratagene, La Jolla, CA). containing 5 ␮l of Lipofectamine and 1.6 ␮g of an appropriate plasmid The nucleotide sequence was determined by means of an A.L.F. sequence (pDR2EF1 alone as a control, m-proCPR/pDR2EF1, or m-CPN/ system (Pharmacia Biotech, Uppsala, Sweden). Sequence alignments and pDR2EF1). Five hours after transfection, the DNA/lipid complex medium analyses were performed using GENETYX-MAC software. was replaced with 1 ml of normal growth medium containing 10% FBS. To prevent the FBS from affecting the assay for carboxypeptidase (CP) activ- Mice ity, the medium was exchanged for serum-free DMEM after 12 h. Cells

were cultured for 4 days post-transfection. The harvested culture medium Downloaded from Male BALB/c mice, 8–10 wk of age, were purchased from Japan SLC was clarified by passing through a 0.22-␮m pore size membrane filter and (Shizuoka, Japan). They were allowed free access to food and water. An- was concentrated from 500 to 20 ␮l using a microconcentrator (Microcon- imal experiments were conducted according to the Guidelines for the Care 10, Amicon, Beverly, MA). The concentrated medium was used to measure and Use of Laboratory Animals of Nagoya City University Medical CP activity. School. Experimental protocols were approved by the institutional animal care and use committee of Nagoya City University Medical School. Preparation of fresh serum and plasma samples to measure CP RT-PCR activity http://www.jimmunol.org/ Total RNA extracts were prepared from normal BALB/c mouse brain, BALB/c mice, 9–12 wk of age, were injected i.p. with 0.5 or 4 mg/kg of lung, liver, stomach, intestine, spleen, and kidney using Trizol reagent LPS (E. coli O111:B4) in sterile saline. They were anesthetized with pen- (Life Technologies, Gaithersburg, MD). Each cDNA was synthesized from tobarbital before cardiac puncture at 0, 1, 12, 24, 36, and 48 h after LPS 1 ␮g of total RNA treated with DNase I using Superscript II reverse tran- injection. Each blood sample was immediately separated into a 1.5-ml scriptase (Life Technologies) and random hexamer; 2.5% of this reaction plastic tube containing 10 ␮l of heparin and a silicone-coated glass tube product was used as a template for PCR. To confirm the integrity of each kept on ice. The sample in the plastic tube was immediately mixed and then total RNA obtained, GAPDH mRNA was amplified from the same cDNA centrifuged at 8000 rpm for 10 min at 4°C. This was used as the source of preparation. The following primers were used: mCPN, mCPN1ϩ (5Ј-AT mouse plasma. The glass tube sample was incubated for 12 h at 4°C and GCCAGACCTGCCCTCAG-3Ј) and mCPN510Ϫ; mouse proCPR, then centrifuged at 3000 rpm for 15 min at 4°C. This was used as the source ϩ Ϫ Ј of mouse serum. Results were obtained from three mice at each time point. mCPR86 and mCPR740 ; and mouse GAPDH (5 -CATCACCATCT by guest on September 23, 2021 TCCAGGA-3Ј and 5Ј-TTGTCATGGATGACCTTGGC-3Ј). DNA was amplified by PCR with AmpliTaq Gold DNA polymerase (Perkin-Elmer Measurement of CP activity Applied Biosystems, Foster City, CA) for 30 cycles of 95°C for 30 s, 55°C for 1 min, and 72°C for 1 min, followed by 72°C for 5 min. PCR products CP activity was determined using hippuryl-L-arginine (Sigma) as a syn- were separated by electrophoresis on a 1.2% agarose gel and visualized by thetic substrate. The amount of hippuric acid generated by the enzyme was ethidium bromide staining. determined by means of a modification of a liquid chromatography method described previously (18). For concentrated culture medium, 20 ␮l of each Northern blot analysis sample and 40 ␮l of 30 mM hippuryl-L-arginine in 50 mM HEPES (pH 8.2) as the substrate solution were mixed with or without 20 ␮l of 1 mg/ml Total RNA was extracted from normal mouse brain, lung, liver, stomach, trypsin solution (Sigma) and then incubated at 37°C for 45 min. With intestine, spleen, and kidney using Trizol reagent (Life Technologies). A mouse fresh serum and plasma, the sample volume was 10 ␮l. After in- 15-␮g amount of each total RNA preparation was electrophoresed in 1% cubation, 20 ␮l of 2.5 M HCl was added to stop the enzyme reaction. After formaldehyde-agarose gel and transferred to a nylon membrane (Hy- extraction with 300 ␮l of ethyl acetate, the top 200 ␮l was removed and bond-N, Amersham, Arlington Heights, IL). The membrane was hybrid- evaporated, and the residue was dissolved in 200 ␮l of double-distilled 32 ized with P-labeled mouse proCPR and CPN cDNA probes (CPR2 and water. The OD228 of the reaction product (hippuric acid) was compared CPN2) in QuikHyb solution (Stratagene). It was then washed twice with with that of dilutions of a standard solution of hippuric acid. When the 2ϫ SSC containing 0.1% SDS at room temperature for 15 min and once potato carboxypeptidase inhibitor (PCI; Calbiochem, La Jolla, CA) was with 0.1ϫ SSC containing 0.1% SDS at 60°C for 10 min and exposed to used, each sample was incubated with 1 ␮l of PCI solution (adjusted to the X-OMAT AR film (Kodak) at Ϫ80°C overnight. In another experiment, appropriate concentration with 50 mM Tris buffer, pH 7.5) for Ͼ5 min BALB/c mice, 9–12 wk of age, were injected i.p. with 0.5 or 4 mg/kg of before measurement of CP activity. LPS (from Escherichia coli 0111:B4, Sigma, St. Louis, MO) in 0.5 ml of saline or with 0.5 ml of saline alone and killed under anesthesia at 4, 8, 12, and 24 h after LPS injection. Total RNA of the livers was prepared using Results Trizol reagent. The rest of the procedure was conducted as described above Cloning of mouse proCPR and mouse CPN cDNAs except for the use of another mouse proCPR probe (CPR1). Based on several partial mouse sequences similar to those of hu- Construction of eukaryotic expression vectors for transfection of man plasma procarboxypeptidase B (ϭproCPR) and the human mouse proCPR and mCPN CPN active subunit acquired from the mouse EST database, two sets of primers for each enzyme were synthesized. Using these The eukaryotic expression vector pDR2EF1 synthesized by Dr. I. Anegon (Institut National de la Sante´ et de la Recherche Me´dicale, Unite´ 437, primers, we performed PCR amplification on cDNA that had been Nantes, France) was a gift from Dr. B. P. Morgan (University of Wales reverse transcribed from mouse liver total RNA. Probes were then College of Medicine, Cardiff, U.K.) (17). pDR2EF1 contains the powerful made of the PCR products. These probes were labeled with polypeptide chain elongation factor-1 ␣ promoter. Mouse proCPR cDNA [␣-32P]dCTP and used to screen ϳ1,000,000 plaques from the in pBluescript plasmid vector was subcloned into BamHI and EcoRV sites of pDR2EF1, which is termed m-proCPR/pDR2EF1. Mouse CPN cDNA in mouse liver cDNA library. The screening identified two clones the pBluescript vector was digested with BamHI, and an ϳ1.6-kb fragment containing a full-length cDNA of either the mouse homologue of containing an open reading frame was purified using a Prep-A-Gene kit human proCPR or the mouse homologue of the active subunit of The Journal of Immunology 1055 Downloaded from http://www.jimmunol.org/ by guest on September 23, 2021

FIGURE 1. Nucleotide sequences of mouse proCPR (A) and mouse CPN (B) cDNAs and their deduced amino acid sequences. A, Nucleotides are numbered from the Met initiation codon. The ﬁrst arrowhead indicates the signal peptidase cleavage site, and others indicate trypsin cleavage sites deduced from human plasma proCPB (ϭproCPR) sequence analysis (11). The polyadenylation site is underlined. Residues implicated in zinc (circles) and substrate (squares) binding previously determined for human proCPR are indicated. Predicted N-glycosylation sites are shown by asterisks. B, The arrowhead indicates the signal peptidase cleavage site. The other symbols are the same in A. human CPN. We tentatively designated the isolated cDNAs as are probably present in the activation peptide as in human proCPR. mouse proCPR and mouse CPN, respectively. These were com- From the amino acid sequence, two glycosylation sites may exist pletely sequenced using both universal primers and speciﬁc inter- in the catalytic domain. Sequence analysis of the mouse CPN clone nal primers of our own design. The DNA sequences of these clones revealed an insert of 1774 bp with an open reading frame of 1371 are shown in Fig. 1. Sequence analysis of the mouse proCPR clone bp coding for a protein of 457 aa in length (Fig. 1B). There is no revealed an insert of 1432 bp with an open reading frame of 1266 canonical polyadenylation signal preceding the poly(A) tail at the bp coding for a protein of 422 aa (Fig. 1A). The 3Ј-untranslated appropriate distance as is present in the cDNA of human CPN. sequence includes the canonical polyadenylation signal, However, the sequence ATTAAA, 22 bp upstream from the

AATAAA, 33 bp upstream from the poly(A) tail. The NH2-termi- poly(A) tail, probably serves this function. The 20 aa of the NH2- nal of 21 aa probably represents a portion of the signal peptide as terminal were suggested to be a portion of the signal peptide, as determined by the weight-matrix method (19). The deduced pro- determined using the weight-matrix method (19). The deduced tein sequence for mouse proCPR has 84% identity to human amino acid sequence for mouse CPN has 83% identity to the hu- proCPR. Based on analysis of human proCPR, mouse proCPR man CPN small subunit. Although it had been shown that the small appears to consist of a 92-aa activation peptide and a 309-aa ma- subunits of human CPN are not glycosylated, two N-linked gly- ture enzyme. Four N-linked glycosylation sites (Asn-X-Ser/Thr) cosylation sites were predicted for mouse CPN. 1056 MOUSE proCPR IS AN ACUTE PHASE PROTEIN

FIGURE 2. CP activity in culture media of COS-7 cells transfected with m-proCPR/pDR2EF1 (A) or m-CPN/pDR2EF1 (B) and with pDR2EF1 vector alone as a control. CP activity was measured with or without 0.25 mg/ml trypsin to establish whether each of our isolated gene products is a proenzyme. Preincubation with 50 ␮g/ml of PCI can inhibit the CP activity of trypsin-treated culture medium of m-proCPR cDNA-transfected cells. Activ- ity was determined with hippuryl-L-arginine as substrate. One unit of enzyme activity was defined as the amount of enzyme required to hydrolyze 1 ␮mol of substrate/min at 37°C. Data are the mean Ϯ SD of triplicate samples. Downloaded from Transient expression and characterization of mouse proCPR and vealed two species of 1.6 and 1.9 kb that were expressed abun- mouse CPN dantly in the liver and more weakly in the stomach (Fig. 3B), and We confirmed that the products of our isolated genes have the on longer exposure were also seen in the lung and kidney; how- same properties as human proCPR and CPN. Since we detected ever, only the 1.6-kb species was seen in the lung. Expression of little CP activity in the culture medium of COS-7 cells alone (data the 1.9-kb species was much stronger than that of the 1.6-kb species in the liver and stomach. The 1.9-kb species appears to be not shown), we used these cells as host cells for cDNA transfec- http://www.jimmunol.org/ tion. COS-7 cells were transiently transfected with the following from the mouse CPN gene isolated in the present study. Northern expression vectors: pDR2EF1 alone as a control, m-proCPR/ blot analysis of mouse proCPR mRNA showed only a 1.5-kb spe- pDR2EF1, or m-CPN/pDR2EF1. Because mouse proCPR and cies that was expressed abundantly in the liver and was undetected CPN cDNAs both harbor a portion of the signal peptide, it was in other tissues. This 1.5-kb Northern blot species corresponded to predicted that the recombinant COS-7 cells would secrete the gene the length of our isolated gene. products. Therefore, serum-free medium was used to measure the Expression of proCPR and CPN mRNAs in liver of LPS- CP activity of recombinant proteins secreted from the transfected inoculated mice cells, since its use ensured that there would be little interference from FBS CPs. Enzyme activity was determined from the amount We examined how LPS influences mouse proCPR and CPN gene by guest on September 23, 2021 of hippuric acid generated by the cleavage of hippuryl-L-arginine expression in murine liver, which is the major site of synthesis. As as substrate. To demonstrate whether the gene product is a proen- shown in Fig. 4, i.p. administration of LPS (0.5 mg/kg) caused an zyme, enzyme activities were also measured after trypsin treatment up-regulation of mouse proCPR gene expression, which reached a (16). As shown in Fig. 2A, the medium harvested from COS-7 cells maximal level at 12 h after LPS treatment. The level at 24 h was transiently transfected with m-proCPR/pDR2EF1 plasmid showed still higher than that in the control. On the other hand, expression much higher activity than medium without trypsin. The medium of both 1.6- and 1.9-kb mouse CPN mRNAs remained unchanged used for the control transfectant (vector alone) showed little CP activity even with trypsin. In addition, we demonstrated that PCI (50 ␮g/ml) could inhibit 92% of the trypsin-generated CP activity in the culture medium of mouse proCPR-transfected cells (Fig. 2A). In the culture medium of COS-7 cells transfected with m-CPN/pDR2EF1 plasmid, significant CP activity was observed without trypsin treatment (Fig. 2B). However, treatment with trypsin reduced the CP activity in the supernatant of mouse CPN cDNA transfectant cells to about half that without trypsin treatment despite the fact that limited trypsin treatment of human CPN was shown to enhance activity (W. Campbell, unpublished obser- vations). This may be due to a difference in experimental condi- tions, and additional experiments will establish whether there is a difference in sensitivity to trypsin between mouse CPN and human CPN.

Expression of mouse proCPR and CPN mRNAs in various normal mouse tissues FIGURE 3. Expression of mouse proCPR and CPN transcripts in tissues. A, RT-PCR ampliﬁcation of mouse CPN, mouse proCPR, and By means of RT-PCR, we detected mRNA for mouse proCPR and GAPDH cDNAs from total RNA of mouse tissues. The primers for mouse CPN in all tissues studied except mouse brain CPN or mouse CPN, mouse proCPR, and GAPDH generated ampliﬁcation products of spleen proCPR (Fig. 3A). A comparison of mouse proCPR and 510, 655, and 282 bp, respectively. B, Northern blot analysis of mouse CPN expression showed that they shared almost the same pattern, tissues. Total RNA (15 ␮g/lane) isolated from the indicated mouse tissues in that liver and stomach showed stronger expression than did was electrophoresed and hybridized with probes for mouse proCPR other tissues. Northern blot analysis of mouse CPN mRNA re- and CPN. The Journal of Immunology 1057

FIGURE 4. Time course of induction of mouse proCPR and CPN mR- NAs in liver after LPS injection. Mice received an i.p. injection of 0.5 ml of saline alone as the control or LPS (E. coli O111: B4, 0.5 mg/kg) in 0.5 FIGURE 5. Induction of arginine CP activity by LPS. The CP activity ml of saline. Total RNA was isolated from mouse liver at 4, 8, 12, and 24 h of trypsin-treated plasma (f) and trypsin-treated plasma plus 100 ␮g/ml of after LPS injection and at 12 h after saline injection. Each total RNA (15 PCI (Œ) was measured before and after i.p. injection of 0.5 mg/kg of LPS. ␮ g/lane) was loaded onto a formaldehyde agarose gel. Blots were probed The CP activity of mouse plasma and serum at 0, 1, 6, 12, 24, 36, and 48 h 32 with P-labeled cDNA for the mouse proCPR or mouse CPN fragment. following treatment with LPS is indicated (mouse plasma, 0.5 mg/kg (Ⅺ) Lower bands contain 18S ribosomal RNA in the total RNA loaded onto and 4 mg/kg (E) of LPS; mouse serum, 0.5 mg/kg (छ) and 4 mg/kg (‚) the gel. of LPS). Data represent the mean Ϯ SD of three mice at the indicated times. During these experiments, a total of seven mice injected with 4 mg/kg of LPS died from 12 to 36 h following injection. These mice were Downloaded from following LPS injection (0.5 mg/kg). A high dose (4 mg/kg) and a replaced with seven animals who survived the 4 mg/kg challenge to ana- low dose (0.5 mg/kg) of LPS elicited the same response (data not lyze results for three mice from each group. shown). Furthermore, to examine whether the levels of expression of mouse proCPR and CPN mRNAs in various other tissues are affected by LPS treatment, we compared levels at 12 h after LPS coagulation, the CP activity in plasma was also determined in the

or saline injection by Northern blot analysis. Northern blotting did http://www.jimmunol.org/ not detect mouse proCPR mRNA in brain, lung, stomach, intes- presence of trypsin. As expected, CP activity in the presence of tine, or kidney despite LPS treatment. However, LPS stimulation trypsin was much higher in plasma from mice administered LPS reduced transcription of mouse CPN mRNA in the stomach to half 24 h earlier than in plasma from untreated mice. Furthermore, the that in the untreated control, although it did not change the level in enhanced CP activity in trypsin-treated plasma returned to the the kidney or lung (data not shown). level of activity in untreated plasma in the presence of PCI, indi- cating that the increase in CP activity was due to CPR. Kinetics of CP activity in fresh mouse serum and plasma after LPS stimulation Discussion

Human CPR was initially identified following the observation that In the present study we isolated mouse homologues of human by guest on September 23, 2021 the CP activity of fresh serum is higher than that of plasma (8), and proCPR cDNA and human CPN small active subunit cDNA. The we have now observed the same phenomenon in the mouse. The deduced amino acid sequences revealed Ͼ80% homology between CP activity of fresh mouse serum was approximately twice that of mice and humans. Substrate and zinc binding sites were conserved. plasma. This augmented activity decreased in a time-dependent Culture media of COS-7 cells transfected with these homologues manner at 37°C and returned to the plasma level within 1 h. Fur- had enzyme activities similar to those of the human enzymes, as thermore, PCI could lower this enhanced activity in a dose-depen- mentioned below. The trypsin-treated culture medium of the dent fashion, and activity was completely suppressed to the plasma proCPR transfectant had high CP activity, and that of the CPN level using 50 ␮g/ml of this inhibitor (data not shown). These transfectant also had significant carboxypeptidase activity without results agree with the finding that PCI inhibits human CPR but not trypsin treatment. Furthermore, PCI inhibited the CP activity of human CPN. Therefore, we determined that the CP activity in mouse CPR as is seen with human CPR; therefore, we considered fresh serum and plasma of LPS-treated mice is representative of the isolated genes to be those of mouse proCPR and CPN, respec- the CP activity of CPR plus CPN and of CPN alone, respectively. tively. Northern blot analysis revealed that mouse CPN mRNA At 24 h following LPS administration, the CP activity of fresh consists of two species of 1.6 and 1.9 kb. The mouse CPN cDNA serum was significantly increased to 5 times the plasma level (Fig. that we isolated was regarded as the 1.9-kb species because it was 5). The CP activity of serum did not rise above the plasma level in more abundantly expressed than was the 1.6-kb species. It has the presence of 100 ␮g/ml of PCI (preliminary results). The CP been demonstrated that human proCPR and human CPN are syn- activity of fresh serum was reduced at 6 h following a transient thesized in the liver (11, 20, 21). This organ appears to be the increase at 1 h, but it increased again at 24 h to a level about 2.5 major site of synthesis in the mouse as well. Although mouse CPN times higher than that in sera of untreated mice. The CPR activity, mRNA expression was not affected by LPS, mouse proCPR therefore, was 4 times greater than that in untreated mouse serum, mRNA expression steadily increased for at least 12 h after LPS as calculated by subtracting the plasma CP activity, corresponding treatment. It is conceivable that the CP activity inhibited by PCI of to CPN activity, from the total activity in serum. ProCPR con- trypsin-treated plasma reflects the level of proCPR in plasma. Its sumption during inflammation induced by LPS may be responsible value increased 2-fold at 24 h after LPS injection, suggesting that for the decrease in serum CP activity observed in the early stage proCPR expression was up-regulated at the protein level in addi- after injection. Similarly, plasma CP activity corresponding to tion to the mRNA level. Furthermore the CPR activity (fresh se- CPN had decreased slightly at 6 and 12 h after LPS injection. This rum CP activity minus the plasma CP activity) at 24 h after LPS difference in activity between serum and plasma was the same stimulation increased 4-fold, whereas the plasma CP activity was whether mice were injected with 0.5 or 4 mg/kg of LPS. essentially unaffected. This increase in CPR activity demonstrates To show that the enhanced CPR activity was not due to an that LPS induces activation of proCPR as well as enhanced ex- increased efficiency in the conversion of proCPR to CPR during pression, and CPN is not affected. These results indicate that 1058 MOUSE proCPR IS AN ACUTE PHASE PROTEIN mouse proCPR is an acute phase protein, whereas mouse CPN is References not. CPN may be required for homeostasis to remove terminal 1. Erdo¨s, E. G. 1979. Kininases. In Handbook of Experimental Pharmacology, Vol. lysine and arginine from a variety of peptides. This was supported 25: Bradykinin, Kallidin, and Kallikrein. E. G. Erdsˇs, ed. Springer-Verlag, Berlin, by the finding that CPN is related to the pathophysiology of p. 428. chronic diseases such as chronic urticaria or asthma (7). An in- 2. Bokisch, V. A., and H. J. Muller-Eberhard. 1970. Anaphylatoxin inactivator of human plasma: its isolation and characterization as a carboxypeptidase. J. Clin. crease in CPR activity may facilitate rapid inactivation of inflam- Invest. 49:2427. matory mediators such as anaphylatoxins and kinins generated at 3. Gorski, J. P., T. E. Hugli, and H. J. Muller-Eberhard. 1979. C4a: the third ana- sites of Gram-negative bacterial infections and may consequently phylatoxin of the human complement system. Proc. Natl. Acad. Sci. USA 76: prevent septic shock. Based on these findings, it is conceivable that 5299. administration of proCPR could prevent death from septic shock. 4. Belew, M., B. Gerdin, G. Lindeberg, J. Porath, T. Saldeen, and R. Wallin. 1980. Structure-activity relationships of vasoactive peptides derived from fibrin or fi- However, several recent reports suggest that excess activation of brinogen degraded by plasmin. Biochim. Biophys. Acta 621:169. proCPR may exacerbate a state of disseminated intravascular co- 5. Plummer, T. H., Jr., and M. Y. Hurwitz. 1978. Human plasma carboxypeptidase agulation (DIC) resulting from a Gram-negative bacterial infection N: isolation and characterization. J. Biol. Chem. 253:3907. (14, 15, 22–27). Following activation by thrombin, proCPR, oth- 6. Levin, Y., R. A. Skidgel, and E. G. Erdos. 1982. Isolation and characterization of the subunits of human plasma carboxypeptidase N (kininase i). Proc. Natl. Acad. erwise known as TAFI, exerts an antifibrinolytic effect by remov- Sci. USA 79:4618. ing carboxyl-terminal lysines on fibrin, preventing plasminogen 7. Mathews, K. P., P. M. Pan, N. J. Gardner, and T. E. Hugli. 1980. Familial binding and activation (15). On the other hand, it has been estab- carboxypeptidase N deficiency. Ann. Intern. Med. 93:443. lished that CPN does not remove carboxyl-terminal lysine residues 8. Campbell, W., and H. Okada. 1989. An arginine specific carboxypeptidase generated in blood during coagulation or inflammation which is unrelated to car- from the clot surface despite the preference for lysine residues boxypeptidase N or its subunits. Biochem. Biophys. Res. Commun. 162:933. Downloaded from rather than arginine residues (15, 26). Furthermore, according to 9. Hendriks, D., S. Scharp, M. van Sande, and M. P. Lommaert. 1989. Character- the present study, LPS stimulation did not essentially alter CPN ization of a carboxypeptidase in human serum distinct from carboxypeptidase N. expression or consumption of CPN compared with proCPR fol- J. Clin. Chem. Clin. Biochem. 27:277. 10. Hendriks, D., W. Wang, S. Scharp, M. P. Lommaert, and M. van Sande. 1990. lowing LPS injection (Fig. 5), suggesting that CPN may not be Purification and characterization of a new arginine carboxypeptidase in human actively involved in the pathophysiology of DIC. Activated protein serum. Biochim. Biophys. Acta 1034:86.

C specifically inactivates factors VIIIa and Va, two essential co- 11. Eaton, D. L., B. E. Malloy, S. P. Tsai, W. Henzel, and D. Drayna. 1991. Isolation, http://www.jimmunol.org/ factors in the intrinsic coagulation pathway, thereby attenuating molecular cloning, and partial characterization of a novel carboxypeptidase B from human plasma. J. Biol. Chem. 266:21833. further thrombin formation and consequent activation of TAFI 12. Bajzar, L., R. Manuel, and M. E. Nesheim. 1995. Purification and characteriza- (proCPR) (23). Expression of protein C mRNA in the mouse, tion of TAFI, a thrombin-activable fibrinolysis inhibitor. J. Biol. Chem. 270: mainly in liver, dropped to a minimal level at 8 h after LPS treat- 14477. ment (24). Analysis of serial plasma samples from patients with 13. Campbell, W., K. Yonezu, T. Shinohara, and H. Okada. 1990. An arginine carboxypeptidase generated during coagulation is diminished or absent in patients DIC revealed that the activity of protein C and its Ag decreases with rheumatoid arthritis. J. Lab. Clin. Med. 115:610. progressively during the initial stages of DIC and remains at a low 14. Bajzar, L., J. Morser, and M. Nesheim. 1996. TAFI, or plasma procarboxypep- level for 24–48 h (25). Our studies with a murine model demon- tidase B, couples the coagulation and fibrinolytic cascades through the thrombin- strated that proCPR increases and reaches a maximum level 24 h thrombomodulin complex. J. Biol. Chem. 271:16603. by guest on September 23, 2021 15. Redlitz, A., A. K. Tan, D. L. Eaton, and E. F. Plow. 1995. Plasma carboxypep- after LPS treatment. Therefore, endotoxin induces down-regula- tidases as regulators of the plasminogen system. J. Clin. Invest. 96:2534. tion of protein C and up-regulation of proCPR. Reduced protein C 16. Shinohara, T., C. Sakurada, T. Suzuki, O. Takeuchi, W. Campbell, S. Ikeda, activity favors the intrinsic pathway of coagulation, with a relative N. Okada, and H. Okada. 1994. Pro-carboxypeptidase R cleaves bradykinin fol- increase in thrombin formation and a consequent activation of lowing activation. Int. Arch. Allergy Immunol. 103:400. 17. Charreau, B., A. Cassard, L. Tesson, B. Le Mauff, J. M. Navenot, D. Blanchard, proCPR, which is up-regulated following LPS stimulation. Endo- D. Lublin, J. P. Soulillou, and I. Anegon. 1994. Protection of rat endothelial cells toxin thereby causes the hypercoagulability and hypofibrinolysis from primate complement-mediated lysis by expression of human CD59 and/or seen in DIC. Thrombomodulin (TM) is expressed on the surface of decay-accelerating factor. Transplantation 58:1222. endothelial cells, and soluble functional proteolytic fragments of 18. Hendriks, D., S. Scharp, and M. van Sande. 1985. Assay of carboxypeptidase N activity in serum by liquid-chromatographic determination of hippuric acid. Clin. TM are also present in circulating plasma. It has been reported that Chem. 31:1936. the thrombin/TM complex activates TAFI (proCPR) 1250 times 19. von Heijne, G. 1986. A new method for predicting signal sequence cleavage sites. more rapidly than does free thrombin (14) and that plasma TM, Nucleic Acids Res. 14:4683. including soluble TM fragments, inhibits fibrinolysis in a dose- 20. Oshima, G., J. Kato, and E. G. Erdos. 1975. Plasma carboxypeptidase N, subunits and characteristics. Arch. Biochem. Biophys. 170:132. dependent manner via activation of plasma proCPB (ϭproCPR) 21. Gebhard, W., M. Schube, and M. Eulitz. 1989. cDNA cloning and complete (26). Plasma levels of soluble TM fragments are elevated in pa- primary structure of the small, active subunit of human carboxypeptidase N (ki- tients with DIC (27). These findings suggest that high concentra- ninase 1). Eur. J. Biochem. 178:603. tions of these fragments in patients with DIC facilitate the activa- 22. Minnema, M. C., P. W. Friederich, M. Levi, P. A. von dem Borne, L. O. Mosnier, J. C. Meijers, B. J. Biemond, C. E. Hack, B. N. Bouma, and H. ten Cate. 1998. tion of proCPR. Enhancement of rabbit jugular vein thrombolysis by neutralization of factor XI: Our results suggest that CPR may actually play two important in vivo evidence for a role of factor XI as an antifibrinolytic factor [published roles in vivo: first, as an inactivator of inflammatory mediators to erratum appears in J. Clin. Invest. 1998 Feb 15;101(4):917]. J. Clin. Invest. 101: 10. prevent excessive inflammation, and secondly, as an inhibitor of 23. Bajzar, L., M. E. Nesheim, and P. B. Tracy. 1996. The profibrinolytic effect of fibrinolysis. Therefore, although a higher level of CPR activity activated protein C in clots formed from plasma is TAFI-dependent. Blood 88: may reduce susceptibility to shock, an increase in its activity could 2093. also facilitate DIC by preventing fibrinolysis. Further research on 24. Yamamoto, K., and D. J. Loskutoff. 1998. Extrahepatic expression and regulation CPR both in vitro and in vivo should provide insight into its im- of protein C in the mouse. Am. J. Pathol. 153:547. 25. Marlar, R. A., J. Endres-Brooks, and C. Miller. 1985. Serial studies of protein C portant dual function as a regulatory enzyme. and its plasma inhibitor in patients with disseminated intravascular coagulation. Blood 66:59. Acknowledgments 26. Hosaka, Y., Y. Takahashi, and H. Ishii. 1998. Thrombomodulin in human plasma contributes to inhibit fibrinolysis through acceleration of thrombin-dependent ac- We thank Yasushige Ishikawa, Chiharu Naito, Yoshiaki Tani, and Norihiro tivation of plasma procarboxypeptidase B. Thromb. Haemost. 79:371. Ogawa for their excellent technical assistance. We also thank Catherine 27. Takano, S., S. Kimura, S. Ohdama, and N. Aoki. 1990. Plasma thrombomodulin Campbell for her help in editing this manuscript. in health and diseases. Blood 76:2024.