The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Low, P. J.
Right arrow Articles by Ogata, R. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Low, P. J.
Right arrow Articles by Ogata, R. T.
The Journal of Immunology, 1999, 162: 6580-6588.
Copyright © 1999 by The American Association of Immunologists

Active Sites in Complement Components C5 and C3 Identified by Proximity to Indels in the C3/4/5 Protein Family1

Pamela J. Low, Rong Ai and Ronald T. Ogata2

Torrey Pines Institute for Molecular Studies, San Diego, CA 92121


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
We recently suggested that sites of length polymorphisms in protein families (indels) might serve as useful guides for locating protein:protein interaction sites. This report describes additional site-specific mutagenesis and synthetic peptide inhibition studies aimed at testing this idea for the paralogous complement C3, C4, and C5 proteins. A series of C5 mutants was constructed by altering the C5 sequence at each of the 27 indels in this protein family. Mutants were expressed in COS cells and were assayed for hemolytic activity and protease sensitivity. Mutants at five indels showed relatively normal expression but substantially reduced sp. act., indicating that the mutations damaged sites important for C5 function. Twenty-three synthetic peptides with C5 sequences and 10 with C3 sequences were also tested for the ability to inhibit C hemolytic activity. Three of the C5 peptides and one of the C3 peptides showed 50% inhibition of both C hemolytic and bactericidal activities at a concentration of 100 µM. In several cases both the mutational and peptide methods implicated the same indel site. Overall, the results suggest that regions important for function of both C3 and C5 lie proximal to residues 150–200 and 1600–1620 in the precursor sequences. Additional sites potentially important for C5 function are near residue 500 in the ß-chain and at two or three sites between the N-terminus of the {alpha}'-chain and the C5d fragment. One of the latter sites, near residue 865, appears to be important for proteolytic activation of C5.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Complement components C3,3 C4, and C5 are paralogous approximately 190-kDa glycoproteins (1). Together they play a central role in complement function, interacting with numerous other components during the activation and regulation of this complex system (reviewed in 2). While the complete sequences of C3, C4, and C5 from several species have been known for 10 yr or more, the locations of many interactive sites within these sequences are still unknown. Until the recent report of the x-ray crystal structure of the 35-kDa C3d fragment of C3 (3), the only three-dimensional structural information available was for the small, approximately 9-kDa C3a and C5a fragments of C3 and C5, respectively, which are released from the parent proteins during C activation (4, 5, 6, 7).

We previously suggested that a search for interactive sites in C3, C4, and C5 might be guided by their primary structures alone, through the locations of indels in this protein family. Indels are the insertions or deletions of amino acid residues that result in length polymorphisms among members of a protein family. They are called indels because an insertion in one member of a family is equivalent to a deletion in another (8). We reasoned that regions of a protein near indels are good candidates for protein:protein interaction sites because indels are usually found among those amino acid residues that form loops at the protein surface, usually coils or reverse turns (9, 10, 11, 12), which may be ideal sites for receptor recognition because they present side chains in a highly accessible arrangement around a compact folding of the peptide backbone (13).

We have recently used two approaches to test the indel strategy with C3. First, peptides with sequences corresponding to indel-proximal segments of C3 were tested for their ability to inhibit complement hemolytic and bactericidal activities (14). This approach assumed that a peptide with a sequence recognized by a C3 binding protein could compete for binding to that protein, prevent its interaction with C3, and consequently inhibit complement function. In the second approach, we engineered indel-proximal mutations into C3 and measured the effects of these mutations on C3 activity (15). Four peptides with complement inhibitory activity and a number of mutants with diminished function were identified in these two studies.

Here we describe the results of analogous studies of a second member of the protein family, component C5. This study provides an independent test of the ability of the indel strategy to identify residues directly involved in protein:protein interactions. It also provides an opportunity to assess the extent to which important functional sites occupy similar positions in C3 and C5. These two proteins, together with C4, have similarities in their sequences, subunit and precursor structures, protease sensitivities, and other properties, which suggest that they share very similar three-dimensional structures while having distinct functions and binding specificities. A simple view is that the three proteins share a common structural framework, and that the distinctive features of each protein are due to unique sequences positioned at corresponding sites on the surface of this core structure. The results with C3 and C5 will test this view, because the locations of indels are by the nature of the sequence alignment at equivalent positions for all family members.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Reagents

Most materials for assaying hemolytic activity and protease sensitivity, including specialized buffers, purified proteins, sensitized erythrocytes, and serum reagents, were purchased from Advanced Research Technologies (San Diego, CA). Goat antiserum against human C5 and a monoclonal anti-human C5 were obtained from Advanced Research Technologies and Quidel (Lafayette, CO), respectively. Oligonucleotides were purchased from Genosys Biotechnologies (The Woodlands, TX).

Synthetic peptides were purchased from Chiron Mimotopes (San Diego, CA). Peptides were purified by reverse phase HPLC and were assessed for purity by analytical HPLC and mass spectral analysis by the supplier. Except as noted in Results, all peptides were supplied at >95% purity. Stock solutions were prepared in sterile distilled water at 5 or 10 mg/ml and neutralized when necessary with 1 M NH4HCO3 or 1 M HOAc.

C5 mutant construction

Mutants were constructed by altering the hC5 cDNA sequence in pHC5.D (16) by overlap extension PCR (17) employing Pfu DNA polymerase (Stratagene, La Jolla, CA). Sequences were confirmed by manually sequencing by standard methods the entire amplified segment after reinsertion into the cDNA. Plasmid DNAs for transfection were isolated with the purification kit from Qiagen (Santa Clarita, CA). In all cases two independent clones for each mutant were isolated, and DNA from each was used in duplicate transfections.

Expression and analysis of C5 mutants

Recombinant C5 was obtained by transiently transfecting COS cells as previously described (16). Culture supernatants were collected approximately 72 h after transfection, placed on ice, and assayed immediately for C5 concentration and within 2 h for C5 hemolytic activity as described below. Supernatants were replaced with methionine-free medium containing [35S]methionine, and cell cultures were incubated for an additional 7 h to obtain metabolically radiolabeled proteins. Radiolabeled supernatants were chilled and tested within 30 min for susceptibility of radiolabeled proteins to CVF,Bb and trypsin. All transfections included a parallel transfection with wild-type (wt) pHC5.D as a standard. Up to 16 transfections (15 mutants plus wtC3) were conducted in duplicate simultaneously.

ELISA quantitation of recombinant C5 in transfected COS supernatants was conducted by a competition ELISA essentially as described for C3 (15), with the substitution of serum C5 (Advanced Research Technologies) for C3 and of monoclonal anti-C5 for anti-C3c.

C5 hemolytic activity was measured by a hemolytic assay (15), with substitution of C5 for C3 and of C5-depleted serum for C3-depleted serum (both from Advanced Research Technologies).

Proteolysis of radiolabeled C5

C5 {alpha}-chain cleavage by the CVF,Bb convertase, immunoprecipitation, SDS-gel electrophoresis, autoradiography, and densitometry were conducted as previously described (15, 16, 18). After an initial 15-min incubation of CVF with factors B and D (16), incubation of the resulting convertase with COS supernatants (at 35 µg/ml CVF) for 15 min at 37°C gave 50–65% conversion of the C5 {alpha}-chain to the {alpha}'-chain. Proteolysis by trypsin was conducted in 50 µg/ml trypsin at 37° for 30 min. For trypsin cleavage of C5b, COS supernatants were incubated with CVF,Bb (35 µg/ml CVF) for 60 min at 37°C before incubation with various concentrations of trypsin.

Peptide inhibition of C hemolytic and bactericidal activities was measured in approximately 0.15 and 1% human serum as previously described (14). Under these conditions, the classical C pathway mediates both hemolysis and bacterial killing.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Design and sequences of indel mutants

The mutational strategy was to reverse the deletion or insertion in C5 relative to C3 or C4 at each of the 27 indels in the C3/4/5 family, as conducted previously with C3 (15). For example, if the indel were an insertion in C5 relative to C3 and C4, those residues were deleted in the mutant; for a deletion in C5 relative to C3 or C4, the corresponding residues from C3 or C4 were inserted. A total of 29 mutants were constructed. Fig. 1Go shows the sequences of the mutants at each indel aligned with the wt hC5 sequence. Two mutants were constructed at indels 21 and 26; the additional mutant at indel 26 was the only purely substitutional mutant that was constructed. A complete sequence alignment of human and murine C3, C4, and C5, showing exact indel locations, is given in Ref. 14.



View larger version (43K):
[in this window]
[in a new window]
  		FIGURE 1. Sequences of human C5 indel mutants. Dashes indicate identity with the wt (hC5) sequence, and asterisks indicate absent residues (gaps) relative to the aligned sequence. The C-terminal residue of each segment is numbered according to Ref. 32. Sequences of the C5 peptides described in this study are underlined; double underlining indicates peptide overlaps. Lys861, which is the primary trypsin cleavage site in C5, is labeled with a near indel 15.

 
Expression of indel mutants

The wt and mutant C5 proteins were assayed directly in supernatants of transiently transfected COS cells. C5 concentrations were measured by ELISA using an anti-C5 mAb and by autoradiographic quantitation of the [35S]methionine-radiolabeled C5 {alpha}-chain after immunoprecipitation with a polyclonal anti-C5 Ab and gel electrophoresis.

Fig. 2Go shows representative results of immunoprecipitation and gel electrophoresis of radiolabeled recombinant wtC5 and the C5 indel 22 mutant, designated hC5/Id22. Note that the immunoprecipitated products contain the approximately 115- and 75-kDa {alpha}- and ß-chains, respectively, of the mature C5 heterodimer as well as the approximately 190-kDa biosynthetic precursor, pro-C5, which makes up about 25% of the total expressed C5 protein. C3 expressed in COS cells shows similar features, but a higher proportion of pro-C3 (~35%) (15).



View larger version (72K):
[in this window]
[in a new window]
  		FIGURE 2. Autoradiogram of a SDS-polyacrylamide gel showing representative results of expression and convertase cleavage of radiolabeled recombinant wtC5 and C5 mutant C5/Id22 from transiently transfected COS cells. Culture supernatants were immunoprecipitated directly with no treatment (N), or after treatment with cobra venom factor C5 convertase CVF,Bb or trypsin at 50 µg/ml (T). The positions of m.w. markers are given at the right of the gel.

 
Fig. 3Goa summarizes the concentrations of C5 mutant proteins in COS supernatants measured by ELISA and radiolabel quantitation. Concentrations are expressed as amounts relative to wtC5 in a parallel transfection to allow comparisons of radiolabel results and to account for experimental variations in transformation efficiencies. By ELISA, the absolute concentration of wtC5 in 12 independent transfections was 110 ± 35 ng/ml, which is slightly lower than the 140 ± 40 ng/ml obtained previously for C3.



View larger version (72K):
[in this window]
[in a new window]
  		FIGURE 3. Properties of C5 indel mutants expressed by transiently transfected COS cells; errors are SDs. Indels 1–12 are in the ß-chain, and indels 13–27 are in the {alpha}-chain. a, Expression. Bars give the concentration of mutant protein in culture supernatants, relative to the amount of wtC5 in a parallel transfection, about 72 h after transfection. Protein levels were measured by ELISA and by densitometric quantitation of the radiolabeled {alpha}-chain after immunoprecipitation, polyacrylamide gel separation, and autoradiography. b, Relative sp. act. This is the mutant sp. act. divided by the sp. act. of wtC5 in a coincident transfection. c, Sensitivity to proteolytic activation by the cobra venom convertase CVF,Bb, expressed as the extent of cleavage relative to wtC5.

 
The data in Fig. 3Goa show that COS cells expressed most of the mutants at levels comparable to wtC5, consistent with the idea that indels are at the protein surface. Only mutants at indels 2, 4, 6, 8, 9, and 11, were expressed at <=30% the wild-type level by both quantitation methods, while mutants at indels 7 and 23 showed this low expression only in the radiolabel assay. In our earlier study of C3, five mutants were expressed at similar low levels. All low expressing mutants were altered in the ß-chain in both C3 and C5, with the exception of the C5 indel 23 mutant.

None of the mutations appeared to affect recognition by the anti-C5 mAb used in the ELISA. If this were the case, we expect for that mutant that the protein concentration measured by ELISA would be much lower than the concentration measured by radiolabeling. However, ELISA and radiolabeling gave similar levels in all cases except the indel 7 and 23 mutants, where ELISA showed a higher level. We cannot fully account for this effect; it may be due in part to an increased proportion of the precursor form, since radiolabel quantitation shows that the proportion of the precursor is 50% greater than that of the wt form for both mutants (data not shown). The epitope recognized by the commercial C5 mAb that we used is unknown.

In general, quantitation of C5 showed more variability, especially by ELISA, and the ELISA and radiolabel methods disagreed more often than in the previous study of C3 (15). These uncertainties may be due to the lower levels of C5 expressed by COS cells and perhaps also to the reported high affinity of C5 for vessel surfaces (19, 20). Radiolabel quantitation is probably more reliable for measuring relative expression levels, since it involves a polyclonal Ab and specific quantitation of the radiolabeled {alpha} subunit of the mature C5 protein.

Hemolytic activities of indel mutants

Again, all assays were conducted directly on supernatants from transiently transfected COS cells. Fig. 3Gob summarizes our results. Activities of the mutants are expressed as a relative sp. act., which is the mutant sp. act. divided by the sp. act. of wtC5 from a parallel transfection. The sp. act. of recombinant wtC5 from COS cells was 130% ± 30% of the sp. act. of the reference standard human C5 from serum.

Mutants C5/Id3, C5/Id10, C5/Id15, C5/Id16, and C5/Id26B consistently showed reduced sp. act. without substantial changes in other properties. The sp. act. of mutants at indels 2, 4, 6, 9, 11, and 23 were also low, at <=20% the wt level, but we regard these results as well as those for the indel 8 mutant to be equivocal. For these latter mutants, reliable calculations of sp. act. were difficult because expression levels were very low, highly variable among independent transfections, and/or disagreed substantially by ELISA and radiolabeling for the same transfection. For example, as shown in Fig. 3Goa, C5/Id2, C5/Id4, and C5/Id11 were nearly undetectable by the radiolabel assay, and the results of ELISA and radiolabeling differed by 4-fold for C5/Id23.

Susceptibility of indel mutants to activation by C5 convertase

C5 is activated when a C5-specific convertase cleaves the {alpha}-chain at a single site, yielding the C5a fragment and the approximately 107-kDa {alpha}'-chain associated with the ß-chain. Indel mutants were assayed for convertase cleavage to test for perturbations of convertase recognition sites. For convenience, we used the soluble, relatively stable convertase, CVF,Bb, formed by the association of cobra venom factor with the Bb fragment of factor B. Although convenient, CVF,Bb differs in substrate specificity from the natural C5 convertases, since it activates both C3 and C5 (21).

Our results, displayed in Fig. 3Goc, revealed that only C5 indel 14 and 15 mutants exhibited substantial resistance to CVF,Bb. Expression levels of indel 2, 4, and 11 mutants were too low to assess cleavage. These results suggest that the low sp. act. of C5/Id15 may be due to resistance to proteolytic activation. However, resistance to activation in this assay is apparently not adequate to cause decreased hemolytic activity, since C5/Id14 shows essentially normal hemolytic activity. The latter result may be an artifact specific to CVF,Bb, however, and C5/Id14 may have normal susceptibility to the surface-bound classical pathway C5 convertase, C4b,2a,3b, that is functional in the hemolytic assay.

C5 conformation assessed by trypsin sensitivity

All mutants were tested for sensitivity to trypsin to assess their conformational state, as proteolysis of C5 is very conformation dependent (20). The conformations of the convertase-resistant indel 14 and 15 mutants were of particular interest because in our previous study (15) we found that a mutation in C3 at indel 14 impaired activity and susceptibility to C3 convertase by shifting the overall conformation of the mutant to a C3b-like form.

As illustrated in Figs. 2Go and 4Go, treatment of wtC5 with low levels of trypsin results in scission primarily at a single site in the {alpha}-chain, probably at Lys861 (20), yielding 90- and 25-kDa peptides designated {alpha}1 and {alpha}5, respectively (22). In comparison, wtC5b is much more sensitive to trypsin. As shown in Fig. 4Go, at 90 µg/ml trypsin, the parent {alpha}'-chain of C5b is completely digested, and lower m.w. digestion products predominate.



View larger version (58K):
[in this window]
[in a new window]
  		FIGURE 4. Autoradiogram of an SDS-polyacrylamide gel showing the results of treating recombinant wt hC5, C5b (wt hC5b), and indel mutant hC5/Id21B with trypsin at 10, 90, and 230 µg/ml for 30 min at 37°C. Peptides {alpha}1, {alpha}2, {alpha}4, and {alpha}5 are as designated in Ref. 22.

 
Mutant C5/Id14 displayed the wild-type trypsin cleavage pattern (not shown); therefore, its resistance to cleavage by CVF,Bb is apparently not due to an altered conformation. In contrast, C5/Id15 exhibited almost no trypsin cleavage under the same conditions. We believe that the failure to cleave at Lys861 is probably due to a local structural perturbation, not to a widespread conformational change, because the six-residue deletion in this mutant is only two residues away from Lys861 (labeled {dagger} in Fig. 1Go). The complete absence of trypsin cleavage indicates that C5/Id15 does not have the generally increased protease sensitivity displayed by C5b. Therefore, both C5/Id15 and C5/Id14 appear to have normal conformations, and their resistance to CVF,Bb is apparently due to localized, more specific structural changes.

Only five other mutants, C5/Id16, C5/Id17, C5/Id21B, C5/Id22, and C5/Id23, showed trypsin fragmentation patterns that differed from the wt pattern. C5/Id21B had the most divergent pattern, with an increased sensitivity to trypsin that is characteristic of C5b and a fragmentation pattern that had characteristic of both C5 and C5b (Fig. 4Go). The remaining mutants were more similar to wtC5 in protease sensitivity, but they all showed a doublet at the {alpha}1 position, illustrated for C5/Id22 in Fig. 2Go. Mutants at indels 16, 17, and 23 also showed a greater proportion of the {alpha}36-kDa peptide characteristic of C5b cleavage.

These results suggest that C5/Id21B and, to a lesser extent, C5/Id16, C5/Id17, C5/Id22, and C5/Id23 have altered conformations, which may resemble that of C5b. However, whatever the nature of the conformational changes, they do not appear to affect C5 function, since only C5/Id16 and perhaps C5/Id23 show decreased activities. The relative sp. act. of C5/Id21B, the mutant showing the greatest change in trypsin cleavage, is 0.3 ± 0.1. This is close to the wild-type activity and is identical with the activity of the other indel 21 mutant, C5/Id21A (Fig. 3Gob), which exhibits the normal trypsin sensitivity and fragmentation pattern.

C3 and C5 indel peptides inhibit C hemolytic activity

To test the indel strategy by a method independent of mutagenesis, we used synthetic peptides with indel-proximal sequences as putative interface peptides (23, 24) to inhibit specific protein:protein interactions. We previously reported the results of testing the inhibitory activities of 21 such peptides with C3 sequences proximal to 15 of the 27 indels (14). Here, we have extended the earlier study to nine additional indels in C3 to give a total of 24 indels examined and to 18 indels in C5. Unlike the mutagenesis studies in which all 27 indel sites were altered, the peptide scans tested peptides from only a subset of indels, primarily because of the high cost of synthetic peptides.

The strategy for selecting peptides was described in detail in the initial study. Briefly, peptide sequences were chosen for their proximity to an indel; sequences spanning an indel were usually chosen. Peptides were arbitrarily made 14 residues in length, with the precise length, sequence, and terminal amidation/acetylation selected primarily to optimize peptide solubility. Some preference was given to peptides with intermediate hydropathy. No systematic attempt was made to determine the effect of terminal modifications, although such modifications can affect peptide activity (25).

The sequences of the C3 and C5 peptides and their positions in the corresponding protein sequences are listed in Tables I and II, respectively. The positions of the C5 peptide sequences in relation to the C5 indel mutations are shown in Fig. 1Go. Peptides were tested for the ability to inhibit C hemolytic and bactericidal activities. Tables I and II list peptide ICH50 and ICB50, defined as the peptide concentrations causing a 50% reduction in serum hemolytic and bactericidal activities, respectively, relative to serum without peptide. The concentrations of serum used in the hemolytic and bactericidal assays (0.15 and 1%, respectively) gave lysis of about 30% of the input erythrocytes and 70–80% killing of input bacteria in the absence of peptide.

Fig. 5Go summarizes the peptide results, including those of the earlier study, which identified C3 peptides from indels 11, 22, and 23 and a segment between the closely spaced indels 13 and 14 (13/14) as having ICH50 <=100 µM. Only one of these, at indel 23, inhibited both hemolytic and bactericidal C activities. The present results reveal that a second C3 peptide, from indel 3, and three C5 peptides, from indels 4, 26, and the 13/14 region, inhibited both C activities. Inhibition in both assays decreases the likelihood of target cell-specific artifacts.



View larger version (54K):
[in this window]
[in a new window]
  		FIGURE 5. Summary of C hemolytic and bactericidal inhibition data for C3 (a) and C5 (b) peptides. More detailed data are given in Tables I and II and in Ref. 14. no eff, No effect of the peptide on hemolysis; enh, enhancement of lysis as described in Table IIGo.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The present studies were designed to test the idea that indels can guide a search for intermolecular interactive sites in C components C3, C4, and C5. Two methods were used: site-specific mutagenesis and inhibition with synthetic peptides. As discussed below, both methods have substantial limitations, and hence neither represents a definitive test of the indel approach. That is, indels may indeed harbor binding sites, but our mutagenesis and peptide inhibition studies as designed may not be adequate to uncover them. Nevertheless, we believe that the use of both methods, on multiple members of this family, will provide a reasonable test of this idea. These studies of C5 extend previous work on synthetic peptides and mutants of component C3 (14, 15).

Potential binding sites on C5 identified by indel mutations

C5 mutants at all 27 of the indels in the C3/4/5 protein family were constructed and tested. The mutations themselves were arbitrarily designed to reverse the insertion or deletion in C5 relative to C3 and C4; this limited systematically the choice of mutations among the thousands of possibilities. Hence, as applied to site-specific mutagenesis, the indel approach simplified the search strategy by limiting both the location and nature of the sequence change.

We assumed as a first approximation that because indels usually occur at the protein surface, most mutations would not affect a protein’s structure or function. We assumed further, that a few mutations, those at intermolecular binding sites, will damage these sites, inhibit binding, and thereby reveal their locations. In practice, indel mutations often result in null or low expression of the mutant protein, and even for proteins expressed at normal levels, loss of function is often incremental, not total (15). Therefore, these initial mutational data alone cannot identify binding sites unequivocally. Rather, they can only point out candidate sites for further investigation. Sites at which the mutation causes complete loss of function while not affecting protein expression are probably the best candidates, but we regard sites where the mutation causes partial loss of activity and/or decreased protein expression to be of potential importance as well.

Table IIIGo lists the C5 mutants showing reduced sp. act. (arbitrarily defined as <=30% of the wild-type activity) and their locations in the pro-C5 amino acid sequence. The mutants are divided into two major groups based on expression, with type 1 mutants having null or low expression, and type 2 mutants having expression levels comparable to that of wtC5. Within types, the mutants are further segregated on the basis of their susceptibilities to activation by the CVF,Bb convertase and their tryptic fragmentation patterns; these properties provide clues to the molecular mechanism responsible for low activity. With the exception of type 2b, which is unique to C5, the types listed are similar to those defined earlier for the C3 mutants.


View this table:
[in this window]
[in a new window]
  		Table III. hC5 mutants with reduced specific activities

 
Based on our simplistic assumptions about the effects of indel mutations, we infer from these results that indels 3, 10, 15, 16, and 26 are the best candidates for active sites in C5. C5/Id16 showed a weak C5b-like tryptic fragmentation pattern, but this change is probably not responsible for the reduced activity of this mutant, since C5/Id17 and C5/Id22 show essentially the same pattern with no loss of activity. Regions near these indels have not previously been reported to be active sites in C5. However, the indel 10 deletion in C3 resulted in essentially complete loss of protein expression (15), and C5 peptides from indels 3 and 26 show complement inhibitory activities.

As discussed earlier, the sp. act. measurements for mutants C5/Id6, C5/Id9, and C5/Id23 were quite equivocal because of large uncertainties in quantitating the recombinant proteins. Therefore, our results are inconclusive with respect to the possible involvement of C5 sequences proximal to indels 6, 9, and 23 in intermolecular interactions. Likewise, C5/Id2, C5/Id4, and C5/Id11 were essentially not expressed in COS cells, and hence the corresponding indels are possible, but unlikely, sites of intermolecular interactions.

Several C3 indel 11 mutants were also poorly expressed (15). Indel 11 is close to an intersubunit disulfide linkage in C3, at Cys559 (26, 27), and inferring from sequence alignment at Cys549 in C5. Therefore, mutations at indel 11 may disrupt the disulfide bridge and/or the {alpha}-ß subunit interface in both proteins. Two other indel mutations are also near intramolecular disulfide bonds. The mutation in C5/Id19 is a 5-aa residue insertion two residues away from Cys1141 that, by analogy with C3, is linked to Cys1083 (27), and the mutation in C5/Id27 is a two-residue insertion that doubles the spacing between two other, by analogy linked, cysteine residues. Both of these mutants are indistinguishable from wtC5 in expression and activity and, hence, apparently do not affect the protein structure.

The region near indel 15 appears to harbor a convertase recognition site

C5/Id15 is unique among the low activity mutants in showing limited susceptibility to CVF,Bb. This suggests that the loss of activity is due to impaired convertase cleavage, and therefore, that the indel 15 region is involved in convertase recognition. Indel 15 is quite distant from the convertase cleavage site, lying approximately 130 residues downstream of Arg733 at the cleavage site. We previously proposed the existence of such a distal recognition site to explain the observation that convertase cleavage of C5 is not blocked by mutations at the cleavage site itself (16).

Consistent with the idea that they form an intermolecular recognition site, the residues near indel 15 appear to be uniquely accessible in comparison with C3 and C4. Limited trypsin digestion of both C3 and C4 results in cleavage at their respective activation sites (28, 29). In contrast, as discussed earlier and illustrated in Figs. 2Go and 4Go, C5 is not cleaved at its convertase activation site under the same conditions, but is instead cleaved almost exclusively at Lys861 (20, 22), which is a single amino acid residue away from indel 15 (see Fig. 1Go). C4 does not have any basic residues in the corresponding region and so should not be similarly cleaved by trypsin, but C3 has a LysArgArg sequence in this region, which should be a good target for trypsin if it were similarly accessible. This region of C5 might also be expected to be highly accessible in comparison with C3 and C4, because indel 15 is a large insertion in C5 relative to both C3 and C4 (seven and nine residues, respectively) (14). This insertion includes a cysteine at residue 865 that is unique to C5. The C5/Id15 mutation may perturb more than convertase recognition, since the indel 14 mutant C5/Id14 shows the same low sensitivity to the convertase, but normal activity. However, as discussed above, the latter could be an artifact of the CVF,Bb convertase, because CVF,Bb recognizes both C3 and C5, whereas the natural classical pathway of C and alternative pathway of C convertases are specific for one or the other protein.

Potential binding sites identified by indel peptides

The peptide method is based on the assumption that a synthetic peptide with the sequence of a protein binding site can in isolation have the same binding capacity. Hence, it can act as a competitive inhibitor of binding to that protein. Successful inhibition of the association of proteins or protein subunits by such interface peptides has been reported in a number of cases, including binding of IgE to its high affinity receptor (30) and the association of the subunits of the EcoRI restriction endonuclease (31).

The present results identified three peptides from C5 with inhibitory activities in both the hemolytic and bactericidal assays. These three peptides had sequences from regions near indel 4, between indels 13 and 14, and near indel 26. One of the C3 peptides tested here, from indel 3, was found to be similarly active. Therefore, including the previous report (15), two C3 peptide inhibitors have now been identified, with sequences from indels 3 and 23.

Neither of the indel 15 peptides, P5–15A nor P5–15B, showed any inhibition of hemolytic activity. As discussed above, the mutational screen had implicated indel 15 as a likely site for C5 convertase recognition. Failure of a peptide to inhibit C activity in these assays does not preclude the corresponding protein sequence as being part of an interactive site, however. The required binding conformation may, for example, be stable only in the context of the native protein. Conversely, inhibition by a peptide does not constitute unequivocal evidence of a binding site in the intact protein. While we infer that peptides inhibit C by an interface peptide mechanism, we do not have direct evidence that this is the case or even of which protein(s) interacts with the peptides. For example, an inhibitory C3 peptide could act by a mechanism unrelated to any normal interaction of C3 with other proteins. Hence, these peptide data are only as suggestive and complementary to the mutational data with respect to the involvement of a particular sequence in a protein:protein interaction.

Comparison of active sites in C3 and C5

Fig. 6Go summarizes the cumulative results of this and previous (14, 15) mutational and peptide studies. Indels where indications of active sites have been found are labeled as • or {circ}, for strong and equivocal indications, respectively. Nineteen different indels are marked, and it is unlikely that all of them are intermolecular binding sites. Nevertheless, they furnish starting points for more detailed studies.



View larger version (9K):
[in this window]
[in a new window]
  		FIGURE 6. Locations of potential active sites near indels in C3 and/or C5 as indicated by indel mutagenesis or indel peptide C inhibition. The horizontal line represents the linear sequences of pro-C3 and pro-C5, which share the same overall features. Positions of indels and the extent of the {alpha} and ß subunits are as noted. TE and {gamma} mark, respectively, the corresponding positions of the thioester in C3 and the beginning of the {gamma}-chain in C4. For C3 and C5 peptide inhibition, • indicates 50% inhibition of both hemolytic and bactericidal activities at <=100 µM; {circ} indicates strong inhibition of hemolytic activity only. Half-symbols ( and ) ) indicate that the peptides had sequences from the segment spanning the closely spaced indels 13 and 14. N, not tested. For C3 and C5 mutants, • indicates mutants with sp. act. <=10% of the wt activity with <=50% of the wt level of expression by COS cells; {circ} indicates either a poorly expressed mutant with <=10% the wt activity or a mutant with 15–30% of the wt activity and near normal expression. N, no expression.

 
Positive indications by both the peptide and mutational methods provide supporting evidence of an active site at that location. Their coincidence at the same indel in both C3 and C5 suggests placement of an active site at a corresponding position in the two proteins. Most suggestive of the latter are the regions proximal to indels 3 and 4, near residues 150–200 and the region flanking indel 26, at residues 1600–1620; these sites are near the N- and C-terminal ends, respectively, of pro-C3 and pro-C5. The region extending from the N-terminus of the {alpha}'-chains to the N-terminus of the C3d/C5d peptide (spanning indels 13–17) also appears to harbor several active sites in both proteins. With the exception of the region between indels 13 and 14, these all represent novel locations for active sites in C3 and C5.

The peptide and mutational results are consistent for indel 26 in C5, where two peptides with complement inhibitory activity, P5–26 and P5–26a, overlap the mutation in the low activity C5/Id26B (Fig. 1Go). These peptides do not overlap the mutation in a second indel 26 mutant, C5/Id26A, which had a near normal sp. act. of 0.4 ± 0.04 and normal expression. Unlike all other mutants, C5/Id26B has substitution and not insertion/deletion mutations; the sequence substituted is from C3. This mutant was constructed only because the peptide results with P5–26 and P5–26A suggested that this segment of C5 is important for activity. This illustrates the potential benefit of using both the peptide and mutational approaches in evaluating the roles of individual indels in protein function.

Both of the C3 indel 26 mutants previously tested also showed impaired hemolytic activities (15) (Fig. 7Go); this indicates that this region is important for the activity of C3 as well, although the corresponding C3 indel 26 peptide P-26 had no inhibitory activity (14). As shown in Fig. 7Go, the sequences of C3, C4, and C5 diverge substantially just upstream of indel 26. This upstream region is flanked on both sides by length polymorphisms: indel 26 and a small indel that we originally (14) excluded from our analysis because of its small size. We have noticed in sequence alignments that closely spaced indels often flank highly polymorphic regions, where the alignment program apparently finds only weak preferences in the precise placement of indels. Hence, closely spaced indels may be a feature of highly polymorphic regions at the protein surface that are important for specific functions.



View larger version (16K):
[in this window]
[in a new window]
  		FIGURE 7. Alignment of indel 26-proximal sequences of human C3, C4, and C5 and the C3 and C5 mutants described in Ref. 15 and this study, respectively. Dashes indicate identity with the wt hC3 or wt hC5 sequence, and asterisks indicate absent residues (gaps) relative to the aligned sequences. The C3 sequence underlined is that of peptide P-26 described in Ref. 14, and the underlined C5 sequences are those of peptides P5–26 and P5–26a, as shown in Fig. 1Go. The terms none and inhibits refer to the complement inhibitory activities of the peptides with the underlined sequences. The + and - refer to the hemolytic activities of the corresponding mutants.

 
Finally, Nagar et al. (3) have recently reported the x-ray crystal structure of the C3d fragment of C3. They identified an acidic pocket in this structure that they suggest is an "attractive candidate" as the binding site for complement receptor CR2. With regard to the utility of the indel approach, it is notable that all three indels in the C3d fragment (indels 18, 19, and 20) (15) lie within loops at the C3d surface, consistent with our basic assumptions. In addition, four of the nine acidic residues in the pocket lie within the loops associated with indel 18 or 19, and four others are in adjoining {alpha}-helixes, within 10 residues of these indels. Therefore, assuming that this pocket is indeed the CR2 binding site, it is interesting that even for binding sites composed of discontinuous segments, the component segments themselves are proximal to indels.


View this table:
[in this window]
[in a new window]
  		Table II. Inhibitory activity of C5 peptides in hemolytic and bactericidal assays1

 

View this table:
[in this window]
[in a new window]
  		Table I. Inhibitory activity of C3 peptides in hemolytic and bactericidal assays1

 

   Acknowledgments
 
We thank W. Kolb for advice on complement assays and comments on the manuscript, R. DiScipio and R. Wetsel for discussions, and D. Li for expert assistance. We also thank Lidak Pharmaceuticals for gifts of synthetic peptides.


   Footnotes
 
1 This work was supported by National Institutes of Health Grant GM29831. Back

2 Address correspondence and reprint requests to Dr. Ronald T. Ogata, Torrey Pines Institute for Molecular Studies, 3550 General Atomic Ct., San Diego, CA 92121. E-mail address: Back

3 Abbreviations used in this paper: C3, C4, and C5, third, fourth, and fifth components of C, respectively; hC5, human C5; wt, wild type; CVF, cobra venom factor. Back

Received for publication November 12, 1998. Accepted for publication March 10, 1999.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Kasahara, M., J. Nakaya, Y. Satta, N. Takahata. 1997. Chromosomal duplication and the emergence of the adaptive immune system. Trends Genet. 13:90.[Medline]
  2. Law, S. K. A., K. B. M. Reid. 1995. Complement 2nd Ed. IRL Press, Oxford, U.K.
  3. Nagar, B., R. G. Jones, R. J. Diefenbach, D. E. Isenman, J. M. Rini. 1998. X-ray crystal structure of C3d: a C3 fragment and ligand for complement receptor 2. Science 280:1277.[Abstract/Free Full Text]
  4. Huber, R., H. Scholze, E. P. Péques, J. Deisenhofer. 1980. Crystal structure analysis and molecular model of human C3a anaphylatoxin. Hoppe-Seyler’s Z. Physiol. Chem. 361:1389.[Medline]
  5. Chazin, W. J., T. E. Hugli, P. E. Wright. 1988. 1H NMR studies of human C3a anaphylatoxin in solution: sequential resonance assignments, secondary structure and global fold. Biochemistry 27:9139.[Medline]
  6. Zarbock, J., R. Gennaro, D. Romeo, G. M. Clore, A. M. Gronenborn. 1988. A proton nuclear magnetic resonance study of the conformation of bovine anaphylatoxin C5a in solution. FEBS Lett. 238:289.[Medline]
  7. Zuiderweg, E. R. P., D. G. Nettesheim, K. W. Mollison, G. W. Carter. 1988. Tertiary structure of human complement component C5a in solution from nuclear magnetic resonance data. Biochemistry 28:172.
  8. Kruskal, J. B.. 1983. An overview of sequence comparison. D. Sankoff, and J. B. Kruskal, eds. Time Warps, String Edits, and Macromolecules: Theory and Practice of Sequence Comparison 1. Addison-Wesley, Reading, MA.
  9. Freimuth, P. I., J. W. Taylor, E. T. Kaiser. 1990. Introduction of guest peptides into Escherichia coli alkaline phosphatase. J. Biol. Chem. 265:896.[Abstract/Free Full Text]
  10. Pascarella, S., P. Argos. 1992. Analysis of insertions/deletions in protein structures. J. Mol. Biol. 224:461.[Medline]
  11. Betton, J.-M., P. Martineau, W. Saurin, M. Hofnung. 1993. Location of tolerated insertions/deletions in the structure of the maltose binding protein. FEBS Lett. 325:34.[Medline]
  12. Sibanda, B. L., J. M. Thornton. 1993. Accommodating sequence changes in ß-hairpins in proteins. J. Mol. Biol. 229:428.[Medline]
  13. Rizo, J., L. M. Gierasch. 1992. Constrained peptides: models of bioactive peptides and protein substructures. Annu. Rev. Biochem. 61:387.[Medline]
  14. Ogata, R. T., P. J. Low. 1997. Complement-inhibiting peptides identified by proximity to indels in the C3/4/5 protein family. J. Immunol. 158:3852.[Abstract]
  15. Ogata, R. T., R. Ai, P. J. Low. 1998. Active sites in complement component C3 mapped by mutations at indels. J. Immunol. 161:4785.[Abstract/Free Full Text]
  16. Ogata, R. T., P. J. Low. 1995. Complement component C5: engineering of a mutant that is specifically cleaved by the C4-specific C1s protease. J. Immunol. 155:2642.[Abstract]
  17. Ho, S. N., H. D. Hunt, R. M. Horton, J. K. Pullen, L. R. Pease. 1989. Site-directed mutagenesis by overlap extension using the polymerase chain reaction. Gene 77:51.[Medline]
  18. Ogata, R. T., P. J. Low, M. Kawakami. 1995. Substrate specificities of the protease of mouse serum Ra-reactive factor. J. Immunol. 154:2351.[Abstract]
  19. Dodds, A. W., S. K. A. Law, R. R. Porter. 1985. The origin of the very variable haemolytic activities of the common human complement component C4 allotypes including C4–A6. EMBO J. 4:2239.[Medline]
  20. DiScipio, R. G.. 1992. Formation and structure of the C5b-7 complex of the lytic pathway of complement. J. Biol. Chem. 267:17087.[Abstract/Free Full Text]
  21. DiScipio, R. G., C. A. Smith, H. J. Müller-Eberhard, T. E. Hugli. 1983. The activation of human complement component C5 by a fluid phase C5 convertase. J. Biol. Chem. 258:10629.[Abstract/Free Full Text]
  22. Wetsel, R. A., W. P. Kolb. 1982. Complement-independent activation of the fifth component (C5) of human complement: limited trypsin digestion resulting in the expression of biologic activity. J. Immunol. 128:2209.[Medline]
  23. Duncan, A. R., G. Winter. 1988. The binding site for C1q on IgG. Nature 332:738.[Medline]
  24. Schramm, H. J., J. Boetzel, J. Büttner, E. Fritsche, W. Göhring, E. Jaeger, S. König, O. Thumfart, T. Wenger, N. E. Nagel, et al 1996. The inhibition of human immunodeficiency virus proteases by "interface peptides.". Antiviral Res. 30:155.[Medline]
  25. Prigge, S. T., A. S. Kolhekar, B. A. Eipper, R. E. Mains, L. M. Amzel. 1997. Amidation of bioactive peptides: the structure of peptidylglycine {alpha}-hydroxylating monooxygenase. Science 278:1300.[Abstract/Free Full Text]
  26. Janatova, J.. 1986. Detection of disulphide bonds and localization of interchain linkages in the third (C3) and fourth (C4) components of human complement. Biochem. J. 233:819.[Medline]
  27. Dolmer, K., L. Sottrup-Jensen. 1993. Disulfide bridges in human complement component C3b. FEBS Lett. 315:85.[Medline]
  28. Bokisch, V. A., H. J. Müller-Eberhard, C. G. Cochrane. 1969. Isolation of a fragment (C3a) of the third component of human complement containing anaphylatoxin and chemotactic activity and description of an anaphylatoxin inactivator of human complement. J. Exp. Med. 129:1109.[Abstract]
  29. Bolotin, C., S. Morris, B. Tack, J. Prahl. 1977. Purification and structural analysis of the fourth component of human complement. Biochemistry 16:2008.[Medline]
  30. McDonnell, J. M., A. J. Beavil, G. A. Mackay, B. A. Jameson, R. Korngold, H. J. Gould, B. J. Sutton. 1996. Structure based design and characterization or peptides that inhibit IgE binding to its high-affinity receptor. Nat. Struct. Biol. 3:419.[Medline]
  31. Brickner, M., J. Chmielewski. 1998. Inhibiting the dimeric restriction endonuclease EcoRI using interfacial helical peptides. Chem. Biol. 5:339.[Medline]
  32. Haviland, D. L., J. C. Haviland, D. T. Fleischer, A. Hunt, R. A. Wetsel. 1991. Complete cDNA sequence of human complement pro-C5: evidence of truncated transcripts derived from a single copy gene. J. Immunol. 146:362.[Abstract]
  33. DeBruijn, M. H. L., G. H. Fey. 1985. Human complement component C3: cDNA coding sequence and derived primary structure. Proc. Natl. Acad. Sci. USA 82:708.[Abstract/Free Full Text]
  34. Fauchere, J.-L., V. Pliska. 1983. Hydrophobic parameters of amino-acid side chains from the partitioning of N-acetyl-amino-acid amides. Eur. J. Med. Chem. 18:369.



This article has been cited by other articles:


Home page
 J. Biol. Chem.Home page
J. Bramham, C.-T. Thai, D. C. Soares, D. Uhrin, R. T. Ogata, and P. N. Barlow
Functional Insights from the Structure of the Multifunctional C345C Domain of C5 of Complement
J. Biol. Chem., March 18, 2005; 280(11): 10636 - 10645.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
C.-T. Thai and R. T. Ogata
Expression and Characterization of the C345C/NTR Domains of Complement Components C3 and C5
J. Immunol., December 15, 2003; 171(12): 6565 - 6573.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
J. M. Inal and J. A. Schifferli
Complement C2 Receptor Inhibitor Trispanning and the {beta}-Chain of C4 Share a Binding Site for Complement C2
J. Immunol., May 15, 2002; 168(10): 5213 - 5221.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
A. Sandoval, R. Ai, J. M. Ostresh, and R. T. Ogata
Distal Recognition Site for Classical Pathway Convertase Located in the C345C/Netrin Module of Complement Component C5
J. Immunol., July 15, 2000; 165(2): 1066 - 1073.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Low, P. J.
Right arrow Articles by Ogata, R. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Low, P. J.
Right arrow Articles by Ogata, R. T.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS