Lys and Fibrinogen Binding of Wild-Type (Trp72) and Mutant (Arg72) Human Apo(a) Kringle IV-10 Expressed in E coli and CHO Cells
Abstract In a previous study, we identified a lysine (Lys)-binding–defective form of human lipoprotein(a) and attributed this defect to the presence of a Trp72→Arg mutation in apolipoprotein(a) [apo(a)] kringle IV-10. To document this relationship, we expressed both wild-type (wt) and mutant (mut) forms of kringle IV-10 in Escherichia coli (nonglycosylated form) and Chinese hamster ovary (CHO) cells (glycosylated form). The Arg72 mut was prepared by introducing the T→A mutation in apo(a) kringle IV-10 amplified from human liver mRNA by the reverse-transcriptase polymerase chain reaction technique. All expressed kringles were tested for their ability to bind Lys and plasmin-modified fibrinogen (PM-fibrinogen). wt kringle IV-10 expressed in both E coli and CHO cells bound to Lys-Sepharose with comparable affinity. In contrast, the Arg72 mut expressed in both systems exhibited no Lys-binding capacity. Moreover, the wt kringle IV-10 expressed in both systems bound to PM-fibrinogen and exhibited two binding components, one Lys mediated (inhibitable by ε-amino-n-caproic acid) and one Lys insensitive, occurring in about the same proportions. Only the latter type of binding was present in the Arg72 mut expressed in E coli. We conclude that kringle IV-10 of human apo(a) has Lys-and PM-fibrinogen–binding capacities that are independent of glycosylation and require the presence of Trp72, one of the seven amino acids that constitute the Lys-binding site of kringle IV-10. Our results also show that the binding of kringle IV-10 to PM-fibrinogen is more complex than that to Lys, in that the former requires an additional binding site or sites outside the Lys-binding site.
- Received June 29, 1995.
- Accepted November 10, 1995.
One of the properties of human Lp(a) is its ability to bind Lys,1 a property that is not exhibited by the Lp(a) counterpart in rhesus monkeys.2 Although the molecular basis of this functional difference has not been clearly established, sequence analysis and modeling data3 have suggested a relationship between the Lys-binding defect and the presence of a Trp72→Arg mutation in rhesus apo(a) kringle IV-10.2 Furthermore, we have found the same mutation in humans with Lys binding–defective Lp(a) by examining the DNA region that codes for the LBS of kringle IV-10.4 The LBS of apo(a) kringle IV-10 comprises three aromatic (Trp72, Trp62, and Phe64), two anionic (Asp55 and Asp57), and two cationic (Arg35 and Arg71) amino acid residues (see Reference 3 for a review). This LBS is identical to that of plasminogen kringle 4, except that in the latter, Lys has replaced Arg in position 35. The importance of kringle IV-10 in Lys binding has also emerged from recent studies of human apo(a) kringle IV-10 expressed in Escherichia coli5 and human kidney embryonic cells6 and of human apo(a) recombinants expressed in HepG2 cells.7 On the other hand, Rouy et al8 observed that the binding of recombinant apo(a) to fibrin was not inhibited by the isolated kringle 4 of plasminogen and concluded that human apo(a) kringle IV-10 had no affinity for the Lys residues of fibrin. This same conclusion was reached by Sangrar et al6 from studies of recombinant kringle IV-10 secreted into the medium of transfected human kidney embryonic cells. Prompted by these divergent observations and our general interest in establishing a relationship between kringle structural mutability and function, we performed studies of both wt Trp72 and mut Trp72→Arg forms of this kringle that were expressed in E coli and CHO cells. The immediate goal of the study was to directly assess the role of Trp72 in Lys and fibrin binding of the individual kringle IV-10 sequences and the influence of the carbohydrate moiety on Lys/fibrin binding. The results of these studies are the subject of this report.
Expression of wt and mut Apo(a) Kringle IV-10 in E coli
PCR Cloning and Mutagenesis
All enzymes were obtained from New England Biolabs unless specified otherwise. The human liver was obtained from the Liver Transplant Program of the University of Chicago. Liver total RNA was isolated by use of the GlassMAX RNA microisolation kit (GIBCO BRL) and converted to cDNA by use of the SUPERSCRIPT preamplification kit (GIBCO BRL). The DNA region that codes for kringle IV-10 and spans nucleotides 12 383 through 12 725 of the apo(a) sequence9 was amplified by using primers A and B (Table⇓). PCR conditions in a reaction volume of 100 μL were as follows: 200 ng human liver cDNA, 20 pmol of each primer, 200 μmol/L dNTP’s, 1 U Vent Polymerase, denaturation at 94°C for 45 seconds, annealing at 57°C for 45 seconds, and extension at 72°C for 1 minute for a total of 35 cycles. For all PCR reactions described in “Methods,” the reaction volume and concentrations of primers, dNTP’s, and Vent Polymerase were the same as stated above. The resulting PCR product (342 bp) was cloned into the EcoRV site of the pBlueScript KS− plasmid (Stratagene) and used to transform DH5α cells (GIBCO BRL) according to the standard techniques of Sambrook et al.10 Sequence analysis of the isolated plasmid DNA revealed that all positive clones contained Met in position 66 of kringle IV-10 (as originally reported by McLean et al9 ) and not Thr, which is known to be a frequent mutation in humans.11 12 13 The point mutation T→A that results in substitution of Trp by Arg at position 72 of kringle IV-10 was introduced into the wt kringle sequence by using recombinant PCR techniques.14 As a first step, we generated two fragments that overlapped in sequence and contained the T→A mutation as part of amplification primers D and E (Table⇓). The first PCR fragment was amplified with primers C and D; for the second fragment we used primers E and F (Table⇓). Primer C contained an EcoRI site, whereas primer F had the stop codon directly followed by an Xho I site. The following PCR conditions were the same for both fragments: 500 ng pBlueScript/KIV-10, denaturation at 94°C for 30 seconds, annealing at 57°C for 30 seconds, and extension at 72°C for 45 seconds for a total of 20 cycles. The PCR products were purified by using a QIAEX Gel Extraction kit (Qiagen), and the purified fragments were joined by reamplification with primers C and F. The conditions for the reaction were as follows: 200 ng of each fragment, denaturation at 94°C for 30 seconds, annealing at 57°C for 30 seconds, and extension at 72°C for 45 seconds for a total of five cycles, followed by denaturation at 94°C for 45 seconds, annealing at 63°C for 45 seconds, and extension at 72°C for 1 minute for a total of 10 cycles. The DNA fragment that codes for wt (Trp72) kringle IV-10 was amplified directly from pBlueScript/KIV-10 by using primers C and F under the following PCR conditions: 500 ng of plasmid DNA, denaturation at 94°C for 30 seconds, annealing at 57°C for 30 seconds, and extension at 72°C for 45 seconds for a total of five cycles, followed by denaturation at 94°C for 45 seconds, annealing at 63°C for 45 seconds, and extension at 72°C for 1 minute for a total of 15 cycles. Finally, the DNA fragments that code for wt and mut kringle IV-10 were gel purified and cloned into EcoRI- and Xho I–digested vector pGEX-KG driven by the tac promoter and containing the coding sequence for GST.15 The expression plasmids were designated pGEX-KG/Trp72 KIV-10 and pGEX-KG/Arg72 KIV-10. DNA sequences of wt and mut kringles were determined by using the fmol Sequencing Kit according to the manufacturer’s instructions (Promega).
Expression, Purification, and Thrombin Cleavage of Fusion Apo(a) Kringle IV-10
wt and mut forms of apo(a) kringle IV-10 were expressed in E coli strain DH5α as fusion GST-proteins according to the instructions provided by Pharmacia. The GST–kringle IV-10 fusion proteins bound to GST-Sepharose (Pharmacia) were cleaved by thrombin for 4 hours at 22°C. The kringles were eluted in the flow-through fraction, whereas the GST fusion component remained bound to the column.
Analysis of free sulfhydryl groups in recombinant kringle IV-10 was performed according to the method of Ampulski et al.16 Kringle IV-10 (20 μmol/L) was exposed to a six-molar excess of 4,4′-dithioldipyridine, and the reaction was monitored at 324 nm for 20 minutes at room temperature.
Expression of wt and mut Kringle IV-10 in CHO Cells
Plasmids and Amplification Primers
Plasmid pBlueScript/KIV-10 was used as a template for generating wt (Trp72) and mut (Arg72) kringle IV-10 expression constructs as described in the previous section, except that primer C was replaced with primer G containing a BsoFI site (Table⇑). The PCR products were digested with BsoFI and Xho I and then gel purified. The human apo(a) leader (95 bp), containing part of the 5′ untranslated region and the signal peptide, was obtained by digestion of cloned apo(a) DNA (PUC119Da18.4 plasmid, generously provided by Dr J.W. McLean) with EcoRI and BsoFI. Apo(a) leader DNA was gel purified, ligated to DNA fragments that code for either wt or mut kringle IV-10, and inserted into cytomegalovirus promoter–driven mammalian expression vector pcDNA3 (Invitrogen) restricted with EcoRI and Xho I. The expression plasmids were designated pcDNA3/Trp72 KIV-10 and pcDNA3/Arg72 KIV-10. DNA sequences of both wt and mut constructs were determined as described above.
Tissue Culture and Transfection Experiments
CHO cells, generously provided by Dr G. Dawson (University of Chicago, Chicago, Ill), were cultured at 37°C in the presence of 10% CO2 with DMEM-F12 medium (GIBCO BRL) supplemented with 10% fetal bovine serum (GIBCO BRL). The expression plasmids pcDNA3/Trp72 KIV-10 and pcDNA3/Arg72 KIV-10, purified by use of the Plasmid Midi Kit (Qiagen), were combined with Lipofectin reagent (GIBCO BRL); transfection was performed according to the GIBCO BRL protocol. Two days after transfection, the cells were plated in DMEM-F12 medium supplemented with 10% fetal bovine serum and 400 μg/mL geneticin (G418, GIBCO BRL). After 10 days stable, individual clones were identified. Several colonies were analyzed, and the clones that were producing the largest amount of recombinant kringles were selected for further examination.
Partial Purification of wt Kringle IV-10 From the Culture Media
Trp72 kringle IV-10 secreted by the stable transformants was isolated from conditioned media by affinity chromatography on Lys-Sepharose (for details, see the “Lys-Sepharose Binding Assay” section). The cells were washed twice with serum-free DMEM-F12 and incubated with the same medium overnight at 37°C. The medium (200 mL) was batch adsorbed with 5 mL Lys-Sepharose for 2 hours at 22°C. The resin was then packed in the column and washed with 10 bed volumes of PBS followed by 10 bed volumes of the same buffer containing 0.5 mol/L NaCl. Bound proteins were eluted with PBS in the presence of 0.2 mol/L EACA (Sigma). The kringle-containing fractions were combined and dialyzed against 50 mmol/L Tris HCl, pH 7.5/0.15 mol/L NaCl.
In Vitro Translation
In vitro translation was performed using the TNT/T7 coupled reticulocyte lysate system and canine pancreatic microsomes from Promega. Aliquots (1 μg) of pcDNA3/KIV-10 (either wt or mut) and l-[35S]Cys (DuPont-NEN) were added to the reaction mixture, and translation was performed according to the manufacturer’s protocol. The proteins were separated by 15% SDS-PAGE17 using a Mini-Gel electrophoresis system from Novex. The gels were treated with En3hancer Autoradiography Enhancer (DuPont-NEN), dried at 60°C under vacuum, and exposed to the film at −70°C.
Steady-State Labeling Experiments
CHO cells (80% confluent) were preincubated for 1 hour at 37°C in serum-free Met/Cys–depleted DMEM (GIBCO BRL) and then labeled for 5 hours in the same medium supplemented with Expre35S35S label containing both [35S]Met and [35S]Cys (125 μCi/mL, DuPont-NEN). The supernatants were harvested after the labeling period and either studied immediately or frozen at −20°C until use.
After preincubation for 1 hour at 37°C in serum-free Met/Cys–depleted DMEM, the CHO cells were labeled for 10 minutes with Expre35S35S label (250 μCi/mL) and chased from 0 to 1 hour at 37°C in serum-free DMEM. The supernatants were harvested and analyzed by immunoprecipitation. Cell monolayers were rinsed twice with cold PBS and scraped into a cold lysis buffer (50 mmol/L Tris HCl, pH 7.5; 0.15 mol/L NaCl; 5 mmol/L EDTA; 0.5% Triton X-100; 0.5% sodium deoxycholate; and 1 mmol/L PMSF). The lysates were clarified by centrifugation in an Eppendorf centrifuge for 10 minutes at 4°C and further analyzed by immunoprecipitation.
Preparation of Anti–Kringle IV-10 Antibody
A polyclonal antibody to human apo(a) kringle IV-10 was prepared at the Animal Facilities of the University of Chicago. Rabbits were injected with 0.5 mg wt human apo(a) kringle IV-10 expressed in E coli that had been mixed in complete Freund’s adjuvant (GIBCO BRL) and at monthly intervals were given two more injections of 0.2 mg kringle mixed with incomplete Freund’s adjuvant (Sigma). Antiserum was collected and frozen at −20°C.
Metabolically labeled supernatants or cell lysates were incubated for 1 hour at 22°C with anti–apo(a) kringle IV-10 serum (1 μL serum per 106 cpm). Thereafter, protein A–Sepharose (Pharmacia) was added and the samples were incubated for an additional 30 minutes at 22°C. The Sepharose was pelleted by brief centrifugation and washed twice with cold NETTAM buffer (50 mmol/L Tris HCl, pH 7.5; 0.15 mol/L NaCl; 5 mmol/L EDTA; 1 mg/mL BSA; 1% Triton X-100) followed by two additional washes with cold RIPA buffer (50 mmol/L Tris HCl, pH 7.5; 0.65 mol/L NaCl; 10 mmol/L EDTA; 1% Triton X-100; 1% sodium deoxycholate; 0.1% SDS) and a final wash with cold water. Laemmli sample buffer17 was then added and the samples were boiled for 5 minutes in the presence of 5% β-ME (Bio-Rad). The solubilized proteins were separated by 15% SDS-PAGE. The gels were treated with autoradiography enhancer, dried at 60°C under vacuum, and exposed to film at −70°C.
Lys-Sepharose Binding Assay
E coli–Expressed Kringles
Cyanogen bromide–activated Sepharose (Pharmacia) was coupled to the α-amino group of l-Lys (Sigma) essentially according to the company’s instructions. The amount of Lys cross-linked to the beads ranged between 16 and 21 μmol/mL of bead suspension. Kringle IV-10 (50 μg in 0.2 mL PBS) was batch incubated with 0.2 mL Lys-Sepharose for 15 minutes at 22°C. The Sepharose was then washed three times with five bed volumes of PBS. The bound protein was eluted with two bed volumes of 0.2 mol/L EACA in PBS, and the resulting fractions were analyzed by 15% SDS-PAGE. The proteins were visualized by staining the gels with Coomassie Brilliant Blue R-250.
CHO Cell–Expressed Kringles
Aliquots (1 mL) of labeled, cell-culture supernatants (see “Steady-State Labeling Experiments”) were incubated with 0.2 mL Lys-Sepharose, as described for the E coli–expressed kringles. The fractions were immunoprecipitated with anti–kringle IV-10 serum and analyzed by SDS-PAGE, followed by fluorography as described in the previous section.
PM Fibrinogen Binding Assay
Conditons for our binding assays were adapted from those of Harpel et al18 and LoGrasso et al.5 Ninety-six–well plates were incubated with 1 μg fibrinogen per well (Sigma) in TBS buffer (50 mmol/L Tris HCl, pH 7.5; 0.15 mol/L NaCl; and 0.2% BSA) for 2 hours at 37°C. After the wells were emptied, 2% BSA in TBS buffer was added to the plates for 1.5 hours at 22°C. The wells were washed three times with TBST and further treated with 3 ng plasmin per well (Enzyme Research Laboratories) for 40 minutes at 37°C. Plasmin was inactivated by incubating the wells with TBST containing the protease inhibitor p-nitrophenyl p′-guanidinobenzoate (Sigma) at a final concentration of 0.1 mmol/L for 20 minutes at 22°C. After two additional washes with TBST, various concentrations of either purified kringle or cell culture medium that had been diluted with TBS with or without 0.2 mol/L EACA were added and incubated overnight at 22°C. Thereafter, the wells were washed three times with TBST, and after addition of rabbit anti–apo(a) kringle IV-10 serum (1:500), incubated for an additional hour at 22°C. The wells were then washed four times with TBST, and the secondary antibody (goat anti-rabbit IgG alkaline phosphatase conjugate [1:2000, Sigma]) was added for 1 hour at 22°C. After another four washes with TBST, the wells were incubated with 1 mg/mL p-nitrophenylphosphate (Sigma), and color development was monitored at 405 nm on a microplate reader (Biomek 1000, Beckman). We derived Kd values from the Langmuir equation, assuming single-site binding according to Fleury and Anglés-Cano.19
Expression of wt and mut Apo(a) Kringle IV-10 in E coli
Both GST-Trp72 kringle IV-10– and GST-Arg72 kringle IV-10–expressed fusion proteins were found in the soluble fraction; thus, they could be readily purified by one-step chromatography on GST-Sepharose as described in “Methods.” wt and mut kringles were then separated from the GST-protein by thrombin cleavage of the fusion protein bound to the column matrix. The product, containing the 112 amino acids of apo(a) kringle IV-10 and the 12 amino acids of GST-protein (GSPGISGGGGGI), was eluted in the flow-through fraction. The yields of purified kringles ranged from 0.4 to 0.8 mg/L of medium. By SDS-PAGE we estimated that the purity of the isolated kringles was >90% (Fig 1A⇓). On the SDS gel, both wt and mut forms of kringle IV-10 migrated as proteins with an approximate molecular mass of 14 500 D (the predicted molecular masses for Trp72 kringle IV-10 and Arg72 kringle IV-10 were 15 861 and 15 831 D, respectively). There were no detectable free sulfhydryl groups in the expressed kringles on the basis of their failure to react with 4,4′-dithioldipyridine, a Cys-specific modifying reagent.16
In Vitro Translation Experiments
Direct expression of pcDNA3/Trp72 KIV-10 and pcDNA3/Arg72 KIV-10 plasmids in the in vitro translation system resulted in the synthesis of proteins with apparent molecular masses of 17 000 D (predicted molecular masses for Trp72 kringle IV-10 and Arg72 kringle IV-10 containing the signal peptide were 17 239 and 17 209 D, respectively; Fig 2⇓, lane 1). To determine the efficiency of the processing events (ie, signal peptide cleavage and core N glycosylation of the kringles), in vitro translation experiments were performed in the presence of different amounts of microsomal vesicles (from 0 to 1.8 μL per 25 μL of reaction mix). Analysis of the translation products by SDS-PAGE revealed two additional proteins with apparent molecular masses of 18 500 and 19 000 D (Fig 2⇓, lanes 2 through 4). Synthesis of these two proteins was completely abolished by the addition of tunicamycin (data not shown) indicating that they represented the N-glycosylated wt and mut forms of kringle IV-10.
Synthesis and Secretion of wt and mut Apo(a) Kringle IV-10 by CHO Cells
By Northern blot analyses, the sizes of the transcripts corresponding to Trp72 kringle IV-10 and Arg72 kringle IV-10 were estimated to be ≈0.9 kb, as predicted (data not shown). Our polyclonal antibody raised against E coli–expressed Trp72 kringle IV-10 exhibited a high affinity for both wt and mut glycosylated kringles during the in vitro translation experiments. Thus, we used this antibody for direct immunoprecipitation of kringles secreted by CHO cells. The cells were metabolically labeled and the conditioned medium immunoprecipitated as described in “Methods.” Analysis of the immunoprecipitates by SDS-PAGE under reducing conditions (Fig 1B⇑) gave an apparent molecular mass of ≈22 000 D for the secreted kringles. The calculated molecular masses for Trp72 kringle IV-10 and Arg72 kringle IV-10 without signal peptide were 14 632 and 14 602 D, respectively; those of the kringles that had been glycosylated during in vitro translation experiments were 18 500 and 19 000 D, respectively. These data suggest that the kringle IV-10 secreted by CHO cells into the conditioned medium was also O-glycosylated.
To determine the effect of the Trp72→Arg mutation on the synthesis and secretion rate of apo(a) kringle IV-10, cells from a representative pool of three independent clones were pulse-labeled for 10 minutes with [35S]Met and [35S]Cys and chased for 0 to 1 hour. The cell lysates and media were then immunoprecipitated with anti–apo(a) kringle IV-10 and analyzed by SDS-PAGE and fluorography. No difference in secretion rate between wt and mut kringles was observed (Fig 3⇓). After a 30-minute chase, both kringles were detected in the medium. The two intracellular proteins (18 500 and 19 000 D) presumably represent the N-glycosylated precursors of kringle IV-10, by comparison of their masses with those of the in vitro translated kringles (Fig 2⇑).
Lys-Binding Properties of Expressed Kringles
The Lys-binding capacities of the recombinant Trp72 and Arg72 IV-10 kringles were examined by affinity chromatography on Lys-Sepharose as described in “Methods.” Both E coli– and CHO-expressed wt kringle IV-10 bound avidly to Lys-Sepharose, so only small amounts (<5%) of the kringles were found in the flow-through fraction (Fig 4A⇓ and 4B⇓, lane 1); the bound kringles were then eluted with 0.2 mol/L EACA (Fig 4A⇓ and 4B⇓, lane 3). In contrast to the wt kringle, both the E coli– and CHO-expressed Arg72 mut had no affinity for Lys-Sepharose and was detected exclusively in the flow-through fraction (Fig 4A⇓ and 4B⇓, lane 1). These results provide direct evidence for the crucial role of Trp72 in the binding of human apo(a) kringle IV-10 to Lys, indicating that the presence of carbohydrates has no effect on this binding.
Binding of Recombinant Kringles to PM-Fibrinogen
wt kringle IV-10 expressed in E coli readily bound to immobilized PM-fibrinogen (Fig 5A⇓). This binding had two components: one Lys mediated, inhibited by 0.2 mol/L EACA and one non–Lys-mediated, unaffected by the presence of the Lys analogue. The Lys-mediated binding was found to be saturable, with an apparent Kd of 23±8 μmol/L. The non–Lys-mediated component represented ≈60% of the total binding of Trp72 kringle IV-10 to fibrinogen. No Kd could be determined for this binding component. In contrast, the Arg 72 mut expressed in E coli exhibited only the non–Lys-mediated binding component (Fig 5B⇓).
The fibrinogen-binding test for Trp72 and Arg72 IV-10 kringles expressed in CHO cells was performed in the presence of conditioned media. The range of kringle concentrations in the fibrinogen-binding assay was estimated to be the same as that for purified kringles expressed in E coli, as determined by an ELISA using anti–apo(a) kringle IV-10 antibody. Neither the wt nor mut form of kringle IV-10 expressed in CHO cells exhibited binding to immobilized PM-fibrinogen. We postulated that other proteins in the CHO cell medium could have affected binding. In fact, addition of medium from mock-transfected CHO cells to the purified Trp72 kringle IV-10 expressed in E coli completely abolished the binding of this kringle to immobilized PM-fibrinogen (data not shown). Of interest, presence of the medium did not influence the binding of wt kringle IV-10 expressed in CHO cells to Lys-Sepharose. On the basis of this latter observation, we purified Trp72 kringle IV-10 from the conditioned medium by affinity chromatography on Lys-Sepharose (see “Methods”) and used this preparation in the fibrinogen-binding studies. However, we were unable to use this method of purification for mut Arg72 kringle IV-10 because of its inability to bind to Lys. The purity of the isolated wt kringle was estimated to be ≈50% by SDS-PAGE on silver-stained gels (data not shown). Kringle concentration was measured by an ELISA using the anti–apo(a) kringle IV-10 antibody against known concentrations of wt kringle IV-10 expressed in E coli. Under our assay conditions, wt kringle IV-10 that had been partially purified from the conditioned medium bound to immobilized PM-fibrinogen (Fig 6⇓). As with wt kringle IV-10 expressed in E coli, the fibrinogen binding of Trp72 kringle IV-10 expressed in CHO cells exhibited two components: one Lys mediated, inhibited by 0.2 mol/L EACA, and one non–Lys-mediated, unaffected by the presence of the Lys analogue. The Lys-mediated binding component was saturable, with an apparent Kd of 3.9±1.2 μmol/L, on the same order of magnitude as that for Trp72 kringle IV-10 expressed in E coli (23±8.7 μmol/L). The non–Lys-mediated component represented ≈40% of the total binding of this kringle to fibrinogen. No Kd could be determined for this binding.
To investigate the nature of the second, non–Lys-mediated binding component in apo(a) kringle IV-10, we performed PM-fibrinogen binding experiments in the presence of Pro, which was shown by Trieu et al20 to inhibit Lp(a) assembly and the binding of apo(a) to apoB-containing lipoproteins. In our studies at concentrations as high as 0.2 mol/L, Pro had no effect on the binding of apo(a) kringle IV-10 to PM-fibrinogen (data not shown).
The current investigation directly supports our previous proposal that human kringle IV-10 has the capacity to bind to Lys and fibrinogen and that Trp72 plays an essential role in this binding.2 4 In terms of Lys binding, our data are in accord with the crystallographic studies of plasminogen kringle 4 by Mulichak et al21 and with the work on recombinant apo(a)/Lp(a) expressed in HepG2 cells by Ernst et al.7 We can also conclude that neither Lys nor fibrinogen binding appears to be influenced by the Met66→Thr mutation, because binding was observed with the Met66 wt kringle IV-10 in the current investigation and with the Thr66 form studied by LoGrasso et al.5
Fless and Snyder22 recently observed that the binding of plasminogen to thrombin-treated fibrin was mediated almost entirely by Lys residues, contrary to Lp(a) binding, which exhibited two components: one LBS mediated and the other LBS independent. In those studies the contribution of the LBS-independent component to the total binding varied from 33% to 80%, depending on the Lp(a) polymorph used.22 Our work now shows that the individual wt kringle IV-10 also binds to PM-fibrinogen via two components, one LBS mediated and another LBS independent. A similar observation concerning apo(a) kringle IV-10 was recently reported by LoGrasso et al,5 although the proportion of the LBS-independent component (referred to by the authors as “nonspecific”) was significantly less than in our studies (15% versus 40% to 60%). In addition, our current investigation has documented the occurrence of this LBS-independent binding component in mut (Arg72) kringle IV-10, indicating that kringle IV-10 can bind to PM-fibrinogen even without a functional LBS. We currently have no knowledge of the mechanism(s) responsible for this “nonspecific” binding. We must note, however, that although the bimodal binding of apo(a) kringle IV-10 to PM-fibrinogen is qualitatively similar to that reported for Lp(a) binding to thrombin-treated fibrin,22 there are important quantitative differences in Kd values: micromolar for kringle IV-10 versus nanomolar for Lp(a). This disparity suggests that either there is a cooperative effect among kringles in terms of apo(a) binding to fibrin(ogen) and/or an additional site or sites on apo(a) are implicated in this binding. Recently, Huby et al23 reported that a “mini-Lp(a),” which was prepared by limited proteolysis of Lp(a) and composed of kringles IV-5 through IV-10, kringle V, and the protease region, bound to either intact or PM-degraded fibrin with a Kd similar to that of intact Lp(a). This finding contrasted with results obtained with the N-terminal domain of apo(a), which comprised kringle IV-1, several kringle IV-2 repeats, kringle IV-3, and kringle IV-4 and did not bind to fibrin either in the presence or absence of EACA.23 Therefore, the location of the additional fibrin-binding site(s) can be restricted to the C-terminal region of apo(a). In addition, Ernst et al7 have recently identified two functionally distinct LBSs on human apo(a), in kringle IV-10 (LBS I) and kringle IV-5–IV-9 (LBS II). However, the authors did not examine whether LBS I and/or LBS II was involved in fibrin binding.
In the current study we also demonstrated that glycosylation had no effect on the binding of recombinant kringle IV-10 to Lys and PM-fibrinogen. On the other hand, Sangrar et al6 observed no fibrinogen binding with recombinant apo(a) kringle IV-10 expressed in human embryonic kidney cells and examined in the presence of conditioned medium. It is likely that in those studies, the expressed kringle was functionally competent but unable to exhibit its binding capacity owing to interfering components in the cell medium. This possibility is supported by our results, which show that the medium of mock-transfected CHO cells inhibits the fibrinogen binding of wt kringle IV-10 expressed in E coli. Moreover, wt kringle IV-10 that had been partially purified from CHO cell medium readily bound to PM-fibrinogen. From these observations it is apparent that caution should be used in interpreting negative binding data from the study of impure products.
In a study concerning the interaction of a purified recombinant 17-kringle apo(a) with a PM-fibrin surface, Rouy et al8 found that the binding was not inhibited by plasminogen kringle 4 and interpreted their observation to suggest that apo(a) kringle IV-10 lacks fibrin-binding capacity. On the other hand, LoGrasso et al5 observed that recombinant apo(a) kringle IV-10 expressed in E coli did inhibit the binding of Lp(a) to immobilized PM-fibrinogen. Based on those observations and our own, it is evident that data obtained with plasminogen may not be readily extrapolated to the apo(a) system.
Taken together, our results indicate that the individual apo(a) kringle IV-10 has Lys- and PM-fibrinogen–binding capacities that are independent of glycosylation and require Trp72, one of the key amino acids for the LBS of this kringle. Our results have also shown for the first time that mut Arg72 kringle IV-10 lacks a functional LBS and therefore binds to PM-fibrinogen through a non–Lys-mediated component outside the LBS domain. This finding may be clinically relevant because it indicates that Lp(a), by binding to fibrin(ogen), may exhibit a prothrombotic action even in the absence of a functional kringle IV-10 LBS.
Selected Abbreviations and Acronyms
|BSA||=||bovine serum albumin|
|CHO||=||Chinese hamster ovary|
|DMEM-F12||=||Dulbecco’s modified Eagle’s medium nutrient mixture F-12 Ham|
|PAGE||=||polyacrylamide gel electrophoresis|
|PCR||=||polymerase chain reaction|
|TBST||=||TBS with 0.02% Tween-20|
This work was supported by National Institutes of Health–National Heart, Lung, and Blood Institute program project grant 18577 (Bethesda, Md). We thank Celina Edelstein for fruitful discussions during the course of the studies and Dr Glyn Dawson for providing the CHO cells.
Eaton DL, Fless GM, Kohr W, McLean J, Qin-Y X, Miller CG, Shively J, Lawn RM, Scanu AM. Partial amino acid sequence of apolipoprotein(a) shows that it is homologous to plasminogen. Proc Natl Acad Sci U S A. 1987;84:3224-3228.
Scanu AM, Miles LA, Fless GM, Pfaffinger D, Eisenbart J, Jackson E, Hoover-Plow JL. Rhesus monkey lipoprotein(a) binds to lysine Sepharose and U937 monocytoid cells less efficiently than human lipoprotein(a): evidence for the dominant role of kringle 437. J Clin Invest. 1993;44:283-291.
LoGrasso PV, Cornell-Kennon S, Boettcher BR. Expression and characterization of human apolipoprotein(a) kringle IV37. J Biol Chem. 1994;269:21820-21827.
Sangrar W, Marcovina SM, Koschinsky ML. Expression and characterization of apolipoprotein(a) kringle IV types 1, 2, and 10 in mammalian cells. Protein Eng. 1994;7:723-731.
Ernst A, Helmhold M, Brunner C, Petho-Schramm A, Armstrong VW, Müller H-J. Identification of two functionally distinct lysine-binding sites in kringle 37 and in kringles 32-36 of human apolipoprotein(a). J Biol Chem.. 1995;270:6227-6234.
Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory; 1989.
Scanu AM, Pfaffinger D, Klezovitch O, Edelstein C. Molecular basis for the lysine binding polymorphism of lipoprotein(a). In: Gallo LL, ed. Cardiovascular Disease 2. New York, NY: Plenum Press; 1995:45-47.
Higuchi R. Recombinant PCR. In: Innis MA, Gelfand DH, Sninsky JJ, White TJ, eds. PCR Protocols: Guide to Methods and Applications. San Diego, Calif: Academic Press Inc; 1990:177-183.
Harpel PC, Gordon BR, Parker TS. Plasmin catalyzes binding of lipoprotein(a) to immobilized fibrinogen and fibrin. Proc Natl Acad Sci U S A.. 1989;86:3847-3851.
Trieu VN, Zioncheck TF, Lawn RM, McConathy WJ. Interaction of apolipoprotein(a) with apolipoprotein B-containing lipoproteins. J Biol Chem. 1991;266:5480-5485.
Fless GM, Snyder M. Polymorphic forms of Lp(a) with different structural and functional properties: cold-induced self-association and binding to fibrin and lysine-Sepharose. Chem Phys Lipids. 1994;67/68:69-79.