>,
`Cl
`0
`0
`C
`::,
`E
`E
`-~
`::,
`'lo
`
`~ 0 u
`~ ::,
`'lo
`C
`~
`:::::
`==
`ii
`::::
`.c:
`a.
`
`::J e C,
`C :c
`.!!? :a
`
`Cl
`
`::J
`Q.
`Q)
`~
`iii z
`
`'<t
`0
`0
`N
`@
`
`~)
`
`BRIDGING INNATE AND ADAPTIVE IMMUNITY
`
`nature
`immunology
`
`REVIEW
`
`The complement system in regulation
`of adaptive immunity
`
`Michael C Carroll
`
`The serum complement system, which represents a chief component of innate immunity, not only participates in inflammation
`but also acts to enhance the adaptive immune response. Specific activation of complement via innate recognition proteins or
`secreted antibody releases cleavage products that interact with a wide range of cell surface receptors found on myeloid, lymphoid
`and stromal cells. This intricate interaction among complement activation products and cell surface receptors provides a basis
`for the regulation of both B and T cell responses. This review highlights fundamental events, explaining how complement links
`innate and adaptive immunity as well as describing more recent studies on how this large family of proteins functions locally in
`peripheral lymph nodes to enhance B and T cell responses.
`
`The complement system was first identified as a heat-sensitive factor
`in fresh serum that 'complemented' the effects of specific antibody in
`the lysis of bacteria and red blood cells1. Its importance as an effector
`of humoral immunity was extended with the early observations of
`opsonization and participation in cellular immunity2. From these
`early studies demonstrating its importance in host protection, the
`complement system was considered a chief component of innate
`immunity. It is now appreciated that this system represents a complex
`pathway of more than 30 serum proteins and cell surface receptors
`that interact in a range of functions from direct cell lysis to the
`enhancement of B and T cell responses3•4. Like the more recently
`identified Toll-like receptors (TLRs), the complement system can be
`activated by 'hard-wired' pattern -recognition receptors that have
`evolved
`to
`recognize pathogen-associated molecular patterns
`(PAMPs) 5·6• Most recognition structures involved in complement
`activation are serum proteins that, in addition to specific antibody,
`include pattern-recognition receptors such as mannan-binding lectin
`ficolins 8, C-reactive protein 9, Clq IO, ! I and natural
`(MBL)7,
`immunoglobulin M (IgM) 12-14.
`Complement is activated by three different pathways: classical,
`lectin and alternative. All three share the common step of activating
`the central component C3, but they differ according to the nature of
`recognition . The classical pathway was the first studied; it is activated
`by antibody released after a humoral response or by natural antibody
`(discussed in more detail below)3. The lectin pathway is activated
`after the recognition and binding of PAMPs by lectin proteins. To
`date, three members of this pathway have been identified: MBL15,
`ficolin H and ficolin L 16. MBL is a C-type lectin and a member of the
`
`collectin family of proteins that includes both collagen and globular
`regions. Structurally MBL is similar to Clq. However, the six globu(cid:173)
`lar heads of MBL form carbohydrate-recognition domains and bind
`N-actetyl glucosamine and mannose, which are common among
`microbes. Given
`the nature of the carbohydrate recognition
`domains, this pathway is probably the oldest in evolutionary terms.
`Ficolins are also lectin-type recognition proteins and, like MBL, they
`recognize N-actetyl glucosamine and mannose structures and
`include both collagen and globular regions. Both MBL and ficolins
`are associated with serine proteases or MEL-associated serine pro(cid:173)
`teases, which act like those members of the classical pathway in acti(cid:173)
`vating components C2 and C4 leading to the central C3 step 16,17.
`Finally, the alternative pathway contrasts with both the classical and
`lectin pathways in that it is continuously 'turned on' because of the
`spontaneous activation of C3 and its promiscuity in binding to a
`wide range of suitable acceptor sites 18, 19.
`Given the multiple pathways of activation and the catalytic nature
`of the many steps, regulation of the complement system is complex.
`To limit host destruction, the system makes use of both serum and
`cell surface regulatory proteins. Almost all mammalian cells express
`regulators of complement to protect against an autologous attack on
`self2°. The finding that some of these 'regulators of complement' also
`participate in adaptive immunity led to the general idea of Fearon and
`Carter that the complement system links innate and adaptive immu(cid:173)
`nity21. It was proposed that the complement system 'instructs' the
`hum oral response in a way that is similar to that of the direction of the
`CD4+ T cell response by dendritic cells (DCs) 22. This review expands
`on this idea of how the complement system is intricately involved in
`regulating not only humoral but T cell immunity.
`
`Th e CBR Institute for Biomedical Research, Harvard Med ica l School, 800
`Hunti ngton Ave., Boston, Massachusetts 02115, USA. Correspondence should
`be addressed to M.C.C. (carrol l@cbr.med.harvard.edu) .
`
`Published on line 28 September 2004; doi:10.1038/nilll3
`
`Regulation of humeral immunity
`Evidence that the complement system was involved in adaptive
`immunity can be traced over three decades to the early reports of
`binding of C3 to circulating lymphocytes23 and follicular dendritic
`
`NATURE IMMUNOLOGY VOLUME 5 NUMBER 10 OCTOBER 2004
`
`981
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1074 - Page 1
`
`

`

`REVIEW
`
`cells (FDCs) within lymphoid follicles24. These findings, combined
`with experimental results showing that a transient reduction in the
`amount of circulating complement C3 led to an impairment in anti(cid:173)
`body response, suggested that the complement system was involved in
`adaptive immunity25.
`A fundamental event following activation of the complement sys(cid:173)
`tem is covalent attachment ofC3b to the foreign (or self) antigen 26,27.
`Conceptually, this is a key event in innate immunity because it pro(cid:173)
`vides a mechanism to mark or identify antigens (both foreign and
`self) for uptake by phagocytic cells or retention on FDCs for recogni(cid:173)
`tion by cognate B lymphocytes (discussed in more detail below). Both
`C3 and C4 components have an internal thioester that upon activa(cid:173)
`tion can form amide or ester linkage28. This usually follows prote(cid:173)
`olytic cleavage via the classical or lectin pathways. However, as
`mentioned above, C3 can spontaneously form a covalent linkage and
`provide a substrate for amplification via the alternative pathway29.
`Complement enhances B cell immunity principally via comple(cid:173)
`ment receptors CD21 (which binds iC3b, C3d,g and C3d) and CD35
`(which binds the same C3 products as CD21 but also binds C3b and
`C4b) 30. In mice, the two receptors seem to be coexpressed and rep(cid:173)
`resent splice products of a single locus (the Cr2 locus)31,32. They are
`expressed mainly on B cells and FDCs. On B cells, CD2 l forms a
`receptor complex with the signaling protein CD19 and tetraspan
`protein CDSl (ref. 33). Uptake ofC3d-coated antigen by cognate B
`cells results in an enhanced signal via the B cell antigen receptor34.
`Thus, coengagement of the CD21-CD19-CD81 coreceptor with B
`cell antigen receptor lowers the threshold of B cell activation and
`provides an important survival signal. A second key mechanism by
`which complement enhances B cell immunity is by localization of
`antigen to FDCs within lymphoid follicles35. FDCs are specialized
`stromal cells
`that secrete the B lymphocyte chemoattractant
`chemokine36 and are important in organizing germinal centers
`
`within the B cell follicles 37. FDCs have relatively high expression of
`CD21 and CD35, and this provides an effective mechanism for
`retention of C3 -coated immune complexes within the lymphoid
`compartment. How immune complexes are taken up by FDCs is not
`clear, but an intact classical pathway and CD2 1 and CD35 receptors
`are thought to be essential38.
`Studies in mice bearing targeted deficiencies of complement Clq39,
`C4 or C3 (ref. 40) or CD2 1-CD35 receptors41 A2 demonstrate the
`involvement of complement in multiple stages of B cell differentia(cid:173)
`tion43 (Fig. 1). B cells first express the CD21-CD19-CD81 coreceptor
`at the transitional stage as they migrate from the bone marrow into the
`periphery44. This stage of development is important in the elimination
`of self-reactive B cells because crosslinking of the B cell antigen recep(cid:173)
`tor results in cell death or anergy rather than activation45. Although
`direct involvement of the CD2 l-CD 19-CDS l co receptor at this stage is
`not apparent, coligation with the B cell antigen receptor by C3- or C4-
`coated self-antigens could lead to enhanced negative signaling.
`All mature B cells have varying expression of CD2 l and CD35. For
`example, follicular B cells and peritoneal Bl cells express low to
`intermediate amounts, whereas marginal zone B cells express rela(cid:173)
`tively high amounts, wh ich they use to transport C3-coated immune
`complexes from the splenic marginal zone into the follicles46A7 (dis(cid:173)
`cussed more below) . The importance of this difference is not appar(cid:173)
`ent, but functional
`involvement of the coreceptor
`in each
`subpopulation has been suggested (Fig. 1). Bl cells, which are the
`chief source of natural antibody, are positively selected during early
`development and are thought to be self-replenishing, unlike conven(cid:173)
`tional B cells48- 50. Complement seems to function in the selection or
`maintenance of Bl cells, as mice deficient in CD21 and CD35 recep(cid:173)
`tors have an altered repertoire51,52.
`Activation of naive mature B cells is enhanced by the presence of an
`intact classical pathway, as mice deficient in early pathway compo-
`nents (Clq, C4 or C3) have
`impaired
`humoral responses
`to
`thymus-dependent
`and thymus-independent antigens4. Notably,
`mice deficient in CD21 and CD35 receptors
`have a similar impairment suggesting that
`overall enhancement by complement
`is
`mediated by these receptors41 •42. However, as
`discussed below, C3 ligands can also affect
`adaptive immunity by other pathways.
`
`Memory B
`
`Spleen, bone marrow
`
`a.
`:I e
`
`C)
`Cl
`
`C :c
`.!!!
`jj
`:I
`CL
`
`Q) ...
`iv z ..,.
`
`:I
`
`0
`0
`C\I
`@
`
`t:t
`
`Peripheral compartment
`
`Transitional
`
`Mature
`
`Activation
`
`~ ©1B1
`
`--(cid:141)
`
`T
`
`Anergy or
`apoptosis
`
`tr
`MHC peptide
`
`!
`C3d
`
`(cid:127)
`
`8
`
`Antigen CD21 , CD19 T
`
`@
`Plasma cell
`
`B1
`
`B2
`
`' >f-
`
`FDC
`
`CD35
`
`Regulation of T cell immunity
`Janeway's conceptualization of the 'adjuvant
`effect' as being due to the influence of the
`innate immune system on T cell immunity
`led to an intense study of how pathogens are
`recognized by DCs and other professional
`antigen-presenting cells (APCs)53. Much of
`the focus has been directed at the growing
`family ofTLRs that recognize a wide range of
`bacterial structures54. Not unexpectedly, the
`complement system also seems to participate
`in T cell immunity through its property of
`marking antigens as foreign and recognition
`by specific receptors. Early studies examining
`the function of complement in T cell priming
`suggested that its involvement was more
`important in B cell immunity (as discussed
`above). C3-deficient mice immunized with
`noninfectious antigens such as haptenated
`
`Figure 1 Complement receptors are important in the regulation of B lymphocyte differentiation at five
`stages. Stage 1: Bl cell s represent a subset of B cel ls that preferentially develop during early life and
`are positively selected by self or microbial antigens, probably at the transitional stage . CD21-CD19-
`CD81 coreceptor engagement by complement-coated antigens enhances positive selection. Stage 2:
`Self-reactive B cells are eli minated in the bone marrow and at the peripheral transitional stage.
`Coligation of the B cel l antigen receptor (BCR) and CD21-CD19-CD81 coreceptor by C3- or C4-coated
`self antigen alters negative signaling. Stage 3: Activation of naive matu re fo llicular B cells by the
`engagement of antigen coupled to complement C3d resu lts in col igation of the B ce ll antigen receptor
`and coreceptor; in the presence of T cell help, th is leads to activation and expansion. Stage 4:
`Activated B cel ls initiate the formation of a germinal center in the splenic fol li cles, which is organ ized
`by FDCs. Complement receptors expressed on germ inal center B cells enhance BCR signaling. In
`addition , complement receptors expressed on FDCs retain antigen and promote the antigen selection
`of high-affinity germinal center B cel ls. Stage 5: Post-GC B cells require complement and antigen
`selection for the efficient maintenance of long-term memory B cel ls.
`
`982
`
`VOLUME 5 N U MBER 10 OCTOBER 200 4 NATURE IMMUNOLOGY
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1074 - Page 2
`
`

`

`REVIEW
`
`keyhole limpet hemocyanin or bacteriophage40,55 developed appar(cid:173)
`ently normal T cell responses, yet humoral responses were impaired.
`Likewise, initial work with infectious herpes simplex virus or inacti(cid:173)
`vated lymphocytic choriomeningitis virus in C3-deficient mice found
`no alteration in the priming of CD4+ T cells, despite the impairment
`ofB cell immunity56•57. In addition, impairment ofB cell responses by
`blocking ofCr2 receptors did not alter priming ofT helper cells58.
`A report that the priming of both CD4+ and CDs+ T cells was
`impaired in C3-deficient mice in an influenza virus model supports
`the idea of more general involvement of the complement system in
`adaptive immunity59. Although the mechanism remains unclear, C3
`is required in the pulmonary T cell response to influenza. Challenge
`of Cr2-deficient mice in a similar model indicates that T cell priming
`is normal, suggesting that the CD21-CD19-CDS1 B cell coreceptor is
`not involved. One possible explanation is that APCs preferentially
`take up C3-coated viral particles for the priming CD4+ and CDs+ T
`cells via complement receptors such as CR3 (CDl lb-CDlS) or CR4
`(CDllc-CDlS). In the absence of C3, APC function could be
`reduced, resulting in a limited priming ofT cells. An alternative possi(cid:173)
`bility that is not mutually exclusive of an APC defect is that the
`chemokine-like C3a and CSa receptors are important in the pul(cid:173)
`monary response.
`A fundamental event following the activation of complement C3
`and CS is release of N-terminal peptides of approximately 9 kilodal(cid:173)
`tons. They are byproducts of all three complement activation path(cid:173)
`ways and serve as potent
`ligands
`for G protein-coupled
`chemoattractant receptors referred to as C3aR and CSaR, respec(cid:173)
`tively60-62. They are expressed on a wide range of inflammatory cells,
`such as mast cells, eosinophils, basophils and lymphocytes, as well as
`smooth muscle cells. They seem to participate in inflammatory events
`via direct cell activation and in the modulation of cytokine release by
`macrophages and/or monocytes. In the influenza model, the absence
`of C3a and CS a ligands could influence the infiltration of CD4 + and
`CDs+ T cells, as is described in more detail below.
`How complement is activated in the influenza model is not
`known, but one strong possibility is that specific natural antibody
`binds virus particles and activates complement via the classical
`pathway. Studies showing involvement of natural antibody in the
`humoral response to influenza support this possibility63• Further
`support for the idea of involvement of natural IgM and complement
`in priming CDS+ T cells comes from a leishmania recombinant anti(cid:173)
`gen hydrophilic acylated surface protein Bl (HASPB-1 ) model64.
`The investigators proposed a pathway in which the recognition of
`HASPB-1 by natural IgM activates classical pathway complement,
`leading to the production of interleukin 4 (IL-4) by a CD11b+
`CD l lc10 population of macrophages and the stimulation of IL-12
`release by DCs. How complement activation stimulates macrophage
`release of IL-4 is not clear. One possibility is that it occurs by direct
`the CR3 receptor with C3-coated antigen.
`opsonization via
`Although crosslinking CR3 is not thought to stimulate the release of
`IL-4 in general, coligation with a HASPB-specific TLR could pro(cid:173)
`vide a suitable stimulus.
`Natural antibody and classical pathway complement also seem to
`be involved in contact sensitivity. In a complex model, involving
`natural killer T cell release of IL-4 and the stimulation of Bl cell
`production of natural IgM, Askenase and colleagues observed that
`the early stage of contact sensitivity is complement-dependent65.
`Notably, complement is activated via the innate natural IgM path(cid:173)
`way, similar to that of the HASPB-1 model discussed above66.
`However, in the example of contact sensitivity, IL-4 production lies
`upstream of Bl cell release of IgM.
`
`In an elegant series of experiments, Wetsel and colleagues first
`demonstrated a role for C3 in a mouse model of asthma. They found
`that mice deficient in C3 were partially protected against airway
`hyper-reactivity, eosinophil infiltration and IL-4 production after
`multiple challenges with a mixture of Aspergillus fumigatus (cell cul(cid:173)
`ture filtrate) and ovalbumin67. The response is characterized in nor(cid:173)
`mal mice by the infiltration of eosinophils, production of IL-4 by
`antigen-specific T cells and airway hyper-reactivity. Thus, this pro(cid:173)
`vided strong support for the idea of involvement of C3 in T helper
`type 2-dependent CD4+ T cell model of asthma. In later studies, the
`same group identified a similar protective phenotype in mice bearing
`a targeted deficiency of C3aR68. This suggests that release of C3a, via
`either direct cleavage by bacterial products or activation of one of the
`three complement pathways, leads to triggering of the C3aR and stim(cid:173)
`ulation of cytokine release. The mechanism in vivo is not clear but as
`both APCs and T cells express C3aR, the results support the idea of
`involvement of C3aR in the linkage of innate and adaptive immunity.
`It will be useful to determine whether C3a release in this model is
`dependent on natural antibody, as found in the leishmania antigen
`study discussed above.
`The other complement-related chemoattractant receptor, CSaR, is
`also potentially involved in the regulation of T cell responses in the
`lung. Mice bearing a targeted deficiency in CSaR were first shown to
`have a reduced response to pulmonary infections with Pseudomonas
`aeruginosa69• Although infiltration of neutrophils seemed normal in
`the infected animals, clearance was substantially impaired in the lung.
`Notably, neutrophil clearance was lung specific, as infection in the
`peritoneum was cleared by the CSaR-deficient mice. The CSa ligand
`and its receptor seem to have a regulatory function in human and
`murine asthma. Karp and colleagues mapped susceptibility to aller(cid:173)
`gen-induced airway hyper-reactivity in humans to the C5 locus70.
`They proposed that CS ligand release induces IL-12 production,
`which has a dampening effect on the T helper type 2-dependent air(cid:173)
`way hyper-reactivity. However, the model is complex: triggering of
`CSaR also downregulates IL-12 (ref. 71).
`The activation ofT cells is tightly regulated and mechanisms have
`been identified both in the thymus (central tolerance) and in the
`periphery. Control of activated CD4+ T cells in the periphery is medi(cid:173)
`ated by both specific elimination and anergy. Recent observations
`support the idea of a third mechanism: that of regulatory T cells. The
`complement system seems to participate in development of human
`regulatory T cells via costimulation of CD3 and CD46. A member of
`the complement regulator receptor family, CD46 (membrane cofac(cid:173)
`tor protein), is associated with this. Early studies had demonstrated
`that crosslinking of CD46 (which is also the measles virus receptor)
`on human monocytes led to an impairment in the expression of IL(cid:173)
`i 2, which is necessary for cell-mediated immunity in measles infec(cid:173)
`tion72. It was proposed that uptake ofC3b-opsonized virus via CD46
`could block IL-12 and that this might explain the T cell suppression
`observed in humans after an infection with measles virus. Later
`experiments supported the idea of an alternative function for CD46.
`Kemper et al. found that crosslinking of CD3 and CD46 on human
`CD4+ T cells led to the induction of a regulatory T phenotype and
`release ofIL-10 (ref. 73). The induced regulatory T cells proliferated
`in culture, blocked activation of bystander T cells and differentiated
`into memory cells. Thus, these findings support the idea of a previ(cid:173)
`ously unknown function for complement in the differentiation of
`regulatory T cells.
`In summary, it is becoming increasingly clear that the comple(cid:173)
`ment system participates in regulation ofT cells by multiple mecha(cid:173)
`nisms such as direct opsonization of foreign antigens by APCs and
`
`NATURE IMMUNOLOGY VOLUME 5 NUMBER 10 OCTOBER 2004
`
`983
`
`a.
`:I e
`
`C)
`Cl
`
`C :c
`.!!!
`jj
`:I
`CL
`
`Q) ...
`iv z ..,.
`
`:I
`
`0
`0
`C\I
`@
`
`t:t
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1074 - Page 3
`
`

`

`REVIEW
`
`marrow-derived C3 deficient (reverse chimeras). Both sets of chimeras
`responded to immunogens, both inert and infectious, when challenged
`intravenously55•77- 79. Thus, despite there being negligible amounts of
`C3 (or C4) in circulation, 'normal' chimeras responded to immuniza(cid:173)
`tion with T cell- dependent antigens with near-normal antibody
`response. These results suggested that local production of classical
`pathway C4 or C3 was sufficient to provide an enhancing effect to
`humoral immunity. The results were further interpreted to suggest that
`the critical site for complement synthesis was the secondary lymphoid
`tissue as complement deposition (and antigen) was observed on FDCs
`in the spleen (for intravenous injection) or draining lymph nodes (for
`peripheral injection).
`Using the same infectious antigen mode described above (herpes
`simplex virus), mice with impaired local production of complement
`('reverse chimeras') but normal levels of complement in circulation
`respond to intravenous immunization, confirming that serum com(cid:173)
`plement proteins are sufficient for enhancement of systemic immune
`responses79 . These results were expected because antigens adminis(cid:173)
`tered intravenously drain into the spleen, where they have contact
`with serum complement proteins. Activation and coupling of C3 to
`antigens injected intravenously facilitates their uptake on marginal
`zone B cells, which participate in transport of antigen into the splenic
`follicles46,47,80• In contrast, 'reverse chimeras' show an impaired
`response to a herpes viral infection introduced intradermally79. This
`finding was somewhat unexpected, because the chimeric animals had
`normal amounts ofC3 in their circulation yet failed to develop a nor(cid:173)
`mal humoral response. Histological examination of the site of viral
`infection identified a vigorous infiltration of monocytic cells similar
`to that found in wild-type control mice. Moreover, leakage ofC3 into
`infected tissues was similar in both reverse chimeras and wild-type
`mice, suggesting that serum-derived complement was not limiting at
`the site of infection. Yet the humoral response to the virus was
`impaired. One explanation for the findings is that the critical site for
`complement enhancement of B cell immunity is in the draining
`peripheral lymph nodes, where activated macrophages produce rela(cid:173)
`tively large amounts of early classical pathway components (Fig. 2).
`These findings raise questions regarding the trafficking of viral anti(cid:173)
`gens into the peripheral lymph nodes and encounter with cognate B
`cells. Complement is not limiting at the site of infection because of
`leakage of serum proteins; therefore, if transport of viral antigens into
`the peripheral lymph nodes is complement dependent, it is unlikely
`that the macrophage components are critical at this step. It seems
`more likely that the limiting step in the function of complement lies
`in the peripheral lymph nodes, where a critical threshold for comple(cid:173)
`ment levels is required for efficient coupling of activated C3 to anti(cid:173)
`gen and the enhancement of adaptive immunity.
`
`Future studies
`Over the past decade our understanding of how the complement
`system regulates B cell immunity has greatly expanded. Yet as mech(cid:173)
`anisms are identified, new questions arise. For example, it is not
`apparent how B cells encounter cognate antigens in either primary
`or secondary responses. Localization of antigens on FDCs in a com(cid:173)
`plement-dependent way after a primary response suggests that
`primed B cells encounter antigen in the follicular zone of secondary
`lymphoid tissues, but how antigen is taken up on FDCs remains a
`mystery. Another important topic over the next decade will no
`doubt focus on unraveling the mechanisms involved in the clear(cid:173)
`ance of self antigens, especially lupus antigens. The finding that
`complement deficiencies are chief susceptibility factors for diseases
`like lupus suggest it is important in clearance. With technological
`
`Figure 2 Complement enhances B cell response to herpes simplex virus 1
`introduced intradermally. Step 1: lntradermal infection with herpes
`simplex virus 1 (HSV-ll induces inflammation, promoting edema and
`leakage of serum proteins into dermis. Complement is activated and binds
`to viral antigens. Infiltrating leukocytes are also a potential source of C3
`for bind ing to viral antigens. Steps 2 and 3: DCs take up viral antigen and
`migrate into local (peripheral) lymph nodes for delivery to T cells. Viral
`antigen passively drains into the peripheral lymph nodes. Steps 4 and 5:
`Activated HSV-1-specific T cel ls release cytokines such as interferon-y
`(IFN-y) , which stimulate macrophages to release early classica l pathway
`complement proteins Step 6: Recognition of viral antigens by natura l or
`immune lgM activates comp lement and C3 attaches to antigen . Steps 7
`and 8: Cognate B cells respond to C3-coated antigen via coligation of the
`B cel l receptor and coreceptor, expanding into antibody-forming ce lls and
`the formation of a germina l center.
`
`by modulating cytokine release. Stimulation of the chemoattractant
`receptors C3aR and CSaR by products of complement activation also
`alters T cell responses by mechanisms that remain unclear. Finally, the
`CD46 receptor that is widely expressed on human cells appears to be
`involved in regulation of the activity ofT cells via both cytokine mod(cid:173)
`ulation and differentiation of regulatory T cells.
`
`Sources of complement
`Consistent with its involvement in systemic immunity, complement
`proteins for the most part are synthesized in the liver, where they are
`taken into circulation. Although synthesis is principally constitutive,
`during infections the release ofIL-6 induces upregulation of the acute
`phase proteins, which includes certain components such as C3.
`Notably, the early components of the classical pathway, Cl, C2, C4
`and C3, are also expressed by macrophages74- 76. In particular, proin(cid:173)
`flammatory cytokines such as IL-6, tumor necrosis factor and inter(cid:173)
`feron-y stimulate macrophages to express the individual components.
`An explanation for regulated expression is that within peripheral
`lymph nodes and local tissues such as the skin, macrophages could
`provide a complete source of components for assembly of activated
`C3 and coating of foreign ( or self-) antigens.
`
`Systemic versus local complement synthesis
`To determine the relative importance of C3 and C4 production by bone
`marrow-derived cells, Verschoor and colleagues prepared radiation
`chimeras. In these chimeras, stromal and liver cells were C3 deficient
`(or C4 deficient), but bone marrow- derived cells were complement
`sufficient (normal chimeras) or vice versa; that is, systemic C3 but bone
`
`984
`
`VOLUME 5 NUMBER 10 OCTOBER 2004 NATURE IMMUNOLOGY
`
`a.
`:I e
`
`C)
`Cl
`
`C :c
`.!!!
`jj
`:I
`CL
`
`Q) ...
`iv z ..,.
`
`:I
`
`0
`0
`C\I
`@
`
`t:t
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1074 - Page 4
`
`

`

`REVIEW
`
`advances in real- time imaging and the construction of conditional
`knock-in mice, many of these types of questions can be answered.
`Finally, the observations that complement in concert with natural
`antibody participates in regulating T cell immunity, both in activa(cid:173)
`tion and in control, provides a rich area for further investigation. It is
`likely that the complement receptor pathways intersect with other
`innate pathways such as the TLRs, as might occur in response to
`highly conserved microbial antigens or PAMPs. This topic provides
`additional opportunities for exploration.
`In summ ary, the complement system represents a complex com(cid:173)
`ponent of innate immunity that has many functions, including the
`regulation of adaptive immun ity. Further study will no doubt lead
`to the identification of additional pathways that link innate and
`adaptive immunity.
`
`ACKNOWLEDGMENTS
`I thank R. Barrington for review of the manuscript, and current and past members
`of the laboratory for their contributions to the understanding of how complement
`influences adaptive immu nity. The related work from my laboratory is supported
`by grants from the National Institutes of Health (A l 39246-09, A l 36389-08,
`Al52343-03 and Al 53570-02).
`
`COMPETING INTERESTS STATEMENT
`The author declares that he has no competing fi nancial interests.
`
`Published on line at http://www.nature.com/natureimmunology/
`
`1. Bordet, J. & Gengou, 0 . Su r !'existence de substances sensibi lisatrices dans la plu(cid:173)
`part des serum antimicrobiens. Ann. Inst. Pasteur 15, 289-302 (1901).
`2. Hosea, S.W., Brown, E.J. & Fra nk, M.M. The critica l role of complement in experi(cid:173)
`menta l pneumococcal sepsis. J. Infect. Dis. 142 , 903-909 (1980).
`3. Reid, K.B.M. & Porter, R.R. The proteolytic activation systems of complement. Annu .
`Rev. Biochem. 50, 433-464 (1981 ).
`4. Fearon, D.T. & Carrol l, M.C. Regulation of B lymphocyte responses to foreign and self(cid:173)
`antigens by the CD19/CD21 complex. Annu. Rev. lmmunol. 18, 393-422 (2000).
`5. Barton, G.M. & Medzhitov, R. Toll-like receptors and their ligands. Curr. Top.
`Microbiol. lmmunol. 270 , 81-92 (2002).
`6. Janeway, C.A. Jr. & Medzhitov, R. Li poproteins take their toll on the host. Curr. Biol.
`9 , 879-882 (1999.
`7. Epstein, J., Eichba um, Q.E., Sheriff, S. & Ezekowitz, R.A.B. The col lectins in innate
`immunity. Curr. Opin. lmmunol. 8, 29-35 (1996).
`8. Matsushita, M. & Fujita, T. The role of ficolins in innate immunity. lmmunobiol. 205,
`490- 497 (2002) .
`9. Szalai, A.J., Agrawal, A., Greenhough, T.J . & Volanakis, J.E. C-reactive protein: struc(cid:173)
`tural biology, gene expression, and host defense. lmmunol. Res. 16, 127-136
`(1997) .
`10. Korb , L.C. & Ahearn, J.M. Clq binds directly and specifically to surface blebs of
`apoptotic kerati nocytes. J. lmmunol. 158 , 4525-4528 (1997).
`11. Taylor, P.R. et al. A hierarchical role for classical pathway complement proteins in the
`clearance of apoptotic cells in vivo. J. Exp. Med. 192, 359-366 (2000).
`12. Boes, M. , Prodeus, A. , Schmidt, T., Carroll, M.C. & Chen , J. A critica l role of natura l
`lgM in immediate response aga inst system ic bacterial infection. J. Exp. Med. 188,
`238 1-23 86 (1998).
`13. Baumgarth, N. et al. Bl and B2 cel l-derived immunoglobulin M antibodies are non(cid:173)
`redu ndant components of the protective response to influenza virus infection.
`J. Exp. Med. 192, 271-280 (2000).
`14. Reid , R.R. et al. Endotoxin shock in antibody-deficient mice: unraveling t he role of
`in
`the clearance of
`lipopolysaccharide.
`natural antibody and complement
`J. lmmunol. 159, 970-975 (1997).
`15. Reid, K. & Tu rner, M. Mammalian lectins in activat ion and clearance mechanisms
`involving the complement system. Semin. lmmunopathol. 15, 307-326 (1994).
`16. Fujita, T., Matsushita, M. & Endo, Y. The lectin-complement pat hway-
`its role in
`innate immunity and evolution. lmmunol. Rev. 198 , 185-202 (2004).
`17. Stah l, P.D. & Ezekowitz, R.A. The mannose receptor is a pattern recognition receptor
`involved in hos