|Home | About | Journals | Submit | Contact Us | Français|
Herpesvirus saimiri encodes a functional homolog of human regulator-of-complement-activation proteins named CCPH that inactivates complement by accelerating the decay of C3 convertases and by serving as a cofactor in factor I-mediated inactivation of their subunits C3b and C4b. Here, we map the functional domains of CCPH. We demonstrate that short consensus repeat 2 (SCR2) is the minimum domain essential for classical/lectin pathway C3 convertase decay-accelerating activity as well as for factor I cofactor activity for C3b and C4b. Thus, CCPH is the first example wherein a single SCR domain has been shown to display complement regulatory functions.
The complement system is an ancient and yet highly evolved effector mechanism of immune defense that forms an imperative branch of innate immunity (23, 46). In addition, recent findings have clearly revealed its role as a vital viaduct between the innate and acquired immune systems (6, 18). Thus, it is not surprising that the system helps in purging a wide array of invaders, including viruses. Consequently, for their successful survival, many viruses have developed mechanisms to subvert the host complement system (7, 24, 26, 29, 39, 45). Herpesviruses and poxviruses, in particular, subvert host complement by encoding structural and/or functional homologs of human complement regulators belonging to the regulator-of-complement-activation (RCA) family, by capturing host membrane complement regulators and by using cellular receptors for entering cells (1, 8, 15, 23).
The RCA proteins are formed by multiple tandem repeats of bead-like complement control protein (CCP) domains or short consensus repeats (SCRs) separated by short linkers. It has been suggested that the sequence variations enforced upon these SCR domain folds and the interdomain dynamics dictate the functionality of the complement regulators (17, 19, 44, 49). Because sequence similarity in herpesviral complement regulators varies between 43% and 89% and in poxviral complement regulators exceeds 91%, it is likely that the structural diversity in herpesviral complement regulators may have resulted in functional differences in these proteins and/or have resulted in variation in structural requirements for complement regulation. In the herpesviridae family, detailed functional characterization has been performed for complement regulators of Kaposi's sarcoma-associated herpesvirus (Kaposica/KCP) (28, 42), herpesvirus saimiri (HVS) (CCPH) (10, 38), and rhesus rhadinovirus (RCP) (31). All these proteins showed conservation of complement regulatory activities, indicating thereby that structural diversity has not resulted in loss of complement regulatory functions in these proteins. However, it is not clear whether sequence variations within the herpesviral complement regulators have resulted in differences in the domain requirements for complement regulatory activities, since mapping of functional domains has been performed only for Kaposica (30, 43). In the present study, we therefore have mapped the complement regulatory domains of HVS CCPH to get further insight into diversity in domain requirements for functional activities.
HVS is a classical prototype of the gamma 2-herpesviruses or rhadinoviruses. It causes rapidly progressing fulminant lymphoma, lymphosarcoma, and leukemia of T-cell origin in marmosets, owl monkeys, and other species of New World primates but not in its natural host, the squirrel monkey (9, 16). Unlike other herpesviruses, it encodes two complement regulators: an RCA homolog (ORF 4; CCPH) that regulates the early steps of complement activation (2, 10) and a CD59 homolog (ORF 15) that inhibits the late steps of complement activation (4, 36). The RCA homolog is formed of four SCR modules (Fig. (Fig.1).1). As a result of alternative splicing, the protein is expressed as a full-length membrane-bound form (mCCPH) containing the transmembrane region as well as a spliced secretory form (sCCPH) lacking the transmembrane region (2). Earlier, we showed that sCCPH inhibits complement by targeting C3 convertases: (i) it supports serine protease factor I-mediated inactivation of C3b and C4b, the subunits of C3 convertases (cofactor activity), and (ii) it accelerates the irreversible decay of the classical pathway (CP)/lectin pathway and to a limited extent the alternative pathway (AP) C3 convertases (decay-accelerating activity [DAA]) (38).
(This work was done in partial fulfillment of the Ph.D. thesis requirements of A.K.S., University of Pune, Pune, India.)
In order to map the functional domains of sCCPH, we have generated a series of soluble triple, double, and single SCR deletion mutants. In brief, the deletion mutants of sCCPH comprising SCR1-3, -2-4, -1-2, -2-3, and -3-4 as well as SCR1, -2, -3, and -4 were constructed from the full-length HVS sCCPH clone (38) by PCR amplification and cloning into the bacterial expression vector pET29. The authenticity of each of the clones was confirmed by DNA sequencing, and then they were transformed into the Escherichia coli BL21 strain for expression. The mutants carried the histidine tag at the C terminus and hence were purified to homogeneity by using histidine affinity chromatography. Refolding of the purified proteins was performed by using the rapid dilution method as previously described (38, 47, 48), and the refolded proteins were loaded onto a Superose 12 gel filtration column (Pharmacia) to obtain monodisperse populations of the expressed mutants (38, 48). The preservation of various functions in mutants (see below) suggests that the mutants have maintained their proper conformation. The expressed proteins were >95% pure as judged by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis (Fig. (Fig.11).
To identify the domains required for cofactor activities of sCCPH against C3b and C4b, we utilized a fluid phase assay wherein C3b or C4b was incubated with each of the deletion mutants and factor I, and inactivation of C3b/C4b (cleavage of the α′-chain) was determined by running the samples on SDS-PAGE gels. It is clear from the data presented in Fig. Fig.22 that sCCPH and the mutants SCR1-3, -2-4, and -1-2 supported the cleavage of the α′-chain of C3b. A very weak cleavage was also supported by SCR2-3 and -3-4. The cleavage of the α′-chain of C4b, however, was supported by sCCPH and the mutants SCR1-3, -2-4, -1-2, and -2-3 but not by SCR3-4 (Fig. (Fig.2).2). Together, these data point out that SCR1 and -2 considerably contribute to the C3b and C4b cofactor activities of sCCPH but that SCR3 and SCR4 in the case of C3b cofactor activity and SCR3 in the case of C4b cofactor activity contribute to its optimal activity. These results, however, did not elucidate whether a single domain(s) could impart the cofactor activities. We therefore expressed the single-domain mutants (SCR1, SCR2, SCR3, and SCR4) and analyzed their cofactor activities. The results presented in Fig. Fig.33 indicate that SCR2, by itself, possesses the ability to support factor I-mediated inactivation of C3b and C4b; SCR3 also displayed very weak cofactor activity against C3b when used at higher concentrations (88 μM; data not shown). These results suggest that structural elements involved in the interaction of sCCPH with factor I are primarily located within SCR2 and -3. Admittedly, the single-domain mutants possess very weak cofactor activities and other domains too contribute to the optimal activity; the cofactor activities of SCR2 for C3b and C4b were 781- and 212-fold lower than that for sCCPH (Fig. (Fig.3).3). It should be mentioned here that earlier observations on mapping of the human RCA proteins (factor H, C4b-binding protein, membrane cofactor protein, and complement receptor 1) (3, 11-13, 21), Kaposica (30), and vaccinia virus CCP (VCP) (27) indicated that a minimum of two (in Kaposica) or three (in all other RCA proteins) successive SCR domains are necessary for factor I cofactor activities. Thus, sCCPH is the first complement regulator in which a single SCR domain has been shown to display the factor I cofactor function.
As discussed above, in addition to the inactivation of subunits of C3 convertases (C3b and C4b), sCCPH also regulates C3 convertases by accelerating their decay. It possesses considerable DAA for the CP/lectin pathway C3 convertase (C4b,2a) and a poor decay activity for the AP C3 convertase (C3b,Bb). Thus, we next examined the DAAs of the various sCCPH mutants to map the domains required for this function. To measure the CP C3 convertase decay activity, the C4b,2a enzyme was formed on sheep erythrocytes and allowed to decay in the presence of various mutants. The remaining enzyme activity was then measured by incubating the reaction mixture with EDTA sera (a source of C3 to C9) and measuring hemolysis. Apart from sCCPH, mutants SCR1-3, -1-2, and -2-3 showed substantial DAA for the CP C3 convertase (Fig. (Fig.4).4). These data suggested that SCR1-3 is primarily responsible for this activity. On a molar basis, SCR1-3 was 1.6-fold less efficient than sCCPH. Because both SCR1-2 and SCR2-3 possessed the decay activity, it was likely that similar to the cofactor activities, a single SCR domain of sCCPH might also possess the DAA for the CP C3 convertase. Hence, we also assessed the DAAs of the single-domain mutants. Interestingly again, SCR2 was the only single domain that distinctly displayed CP DAA (Fig. (Fig.4);4); however, on a molar basis, it was 26-fold less active than sCCPH. Previous data on the involvement of SCR domains in decay acceleration of CP C3 convertase in human RCA proteins (decay-accelerating factor, complement receptor 1, and C4b-binding protein) (3, 5, 20) and viral RCA homologs (Kaposica and VCP) (27, 30) have shown that a minimum of two or three consecutive domains are necessary for the activity. Thus, sCCPH is the only prototype to date in which a single SCR is adequate to impart the CP DAA.
Although sCCPH is known to possess limited AP C3 convertase DAA, we sought to determine whether this limited activity is localized in a specific region or the full-length protein. To measure the AP DAA, the C3 convertase C3b,Bb was formed on the sheep erythrocytes and incubated with sCCPH or with each of its deletion mutants. The decay of the AP C3 convertase was assessed by adding EDTA sera and measuring hemolysis. Although the full-length protein displayed a limited AP C3 convertase, none of the deletion mutants exhibited any activity (Fig. (Fig.44).
Inactivation of C3 convertases by the RCA proteins, owing to their cofactor and decay activities, requires interaction of these proteins with C3b and C4b. The ligand binding activity of the RCA proteins, however, does not always correlate with their cofactor and decay activities (12, 34), as apart from ligand binding, cofactor activity involves interaction of the RCA protein with factor I (40), and decay activity involves interaction of the RCA protein with C2a or Bb (22, 25). In order to determine whether cofactor and decay activity data of sCCPH and the various mutants correlate with the ligand binding data, we measured binding of these proteins to C3b and C4b by using a surface plasmon resonance-based assay (38). As observed earlier (38), sCCPH displayed higher affinity for C4b than for C3b (Fig. (Fig.55 and Table Table1).1). When we measured binding of various deletion mutants to C3b and C4b, only SCR2-4 showed binding to C3b, and SCR1-3 showed binding to C4b (Fig. (Fig.5).5). However, there were reductions of about 16- and 14-fold in the affinities of these deletion mutants for C3b and C4b, respectively, compared to that for sCCPH (Table (Table1),1), suggesting that all the four domains contribute to binding to C3b and C4b. Because most of the deletion mutants that displayed complement regulatory activities possessed negligible binding to C3b and C4b, it is clear that binding of the mutants does not correlate with their cofactor and decay activities. It is likely that during cofactor activity, interaction of the mutants with C3b and C4b is stabilized by the interaction of factor I with C3b/C4b and the mutants. Similarly, during DAAs, the mutants may possess better affinity for the convertases than their subunits C3b and C4b. Consistent with this argument, decay-accelerating factor has previously been shown to bind to CP C3 convertase with 1,000-fold higher affinity than to C4b (33).
The presence of SCR domains is not restricted to complement regulators, as SCR domains are also present in other complement proteins (e.g., C1r, C1s, MASP-1, MASP-2, factor B, C2, C6, and C7) and noncomplement proteins (e.g., β2-GPI, interleukin-2 and -15 receptors, GABAB receptor type 1a, E-selectin, brevican, CSMD-1, and polydom) (41). The SCR domains are always present as a pair or more, and the presence of a single SCR domain in proteins is rare (e.g., interleukin-15R and brevican). Further, data obtained thus far from domain mapping studies indicate that a minimum of two successive SCR domains are required for imparting any function. Together, these findings led to a paradigm: a two-SCR structure is the smallest basic structural unit required for exhibiting any function (44). In the present study, data obtained for HVS sCCPH elucidate for the first time that a single SCR domain (SCR2) is able to impart factor I cofactor activities as well as DAA. Therefore, clearly, the current belief regarding the requirement of multiple domains for displaying any functional activity requires revision. We would like to point out here that though earlier studies of viral complement regulators have used comparable molar excess of regulators for domain mapping studies, similar studies performed for human complement regulators utilized 5- to 50-fold less molar excess of regulators than the present study. Thus, it is likely that single domains in human complement regulators too may possess the complement regulatory activities.
In summary, our findings demonstrate that though three SCR domains of HVS CCPH are necessary for displaying the optimum complement regulatory activities, a single domain is sufficient to impart the various complement regulatory activities. These data therefore point out that sequence variations in herpesviral complement regulators have resulted in a notable difference in domain requirements for the functional activities in these proteins.
We thank John D. Lambris (Department of Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, PA) and Michael K. Pangburn (Department of Biochemistry, University of Texas Health Center, Tyler, TX) for their continuous support. We also thank John Bernet for site-specific labeling of C3b and C4b with biotin and express appreciation to Yogesh Panse and Sarang Satoor for their excellent technical assistance.
This work was supported by the Wellcome Trust Senior Research Fellowship in Biomedical Science in India and a project grant from the Department of Biotechnology, New Delhi, India, to A.S. We also acknowledge the financial assistance to A.K.S., K.P., and V.N.Y. from the Council of Scientific and Industrial Research, New Delhi, India, and to J.M. from the Department of Science and Technology, New Delhi, India.
Published ahead of print on 29 July 2009.