Domains of surfactant protein A that affect protein oligomerization, lipid structure and surface tension

Ž .Comparative Biochemistry and Physiology Part A 129 2001 109�127

Review

Domains of surfactant protein A that affect proteinoligomerization, lipid structure and surface tension

Nades Palaniyar a,b,�, Machiko Ikegamic, Thomas Korfhagenc,Jeffrey Whitsett c, Francis X. McCormackb

aMRC Immunochemistry Unit, Department of Biochemistry, Uni�ersity of Oxford, South Parks Road, Oxford OX1 3QU, UKbDi�ision of Pulmonary�Critical Care Medicine, Department of Internal Medicine, Uni�ersity of Cincinnati, 231 Albert Sabin

Way, Cincinnati, OH 45267-0564, USAcDi�ision of Pulmonary Biology, Children’s Hospital Medical Center, Cincinnati, OH 43229-3039, USA

Received in revised form 7 December 2000; accepted 12 February 2001

Abstract

Ž .Surfactant protein A SP-A is an abundant protein found in pulmonary surfactant which has been reported to havemultiple functions. In this review, we focus on the structural importance of each domain of SP-A in the functions ofprotein oligomerization, the structural organization of lipids and the surface-active properties of surfactant, with anemphasis on ultrastructural analyses. The N-terminal domain of SP-A is required for disulfide-dependent proteinoligomerization, and for binding and aggregation of phospholipids, but there is no evidence that this domain directlyinteracts with lipid membranes. The collagen-like domain is important for the stability and oligomerization of SP-A. Italso contributes shape and dimension to the molecule, and appears to determine membrane spacing in lipid aggregatessuch as common myelin and tubular myelin. The neck domain of SP-A is primarily involved in protein trimerization,which is critical for many protein functions, but it does not appear to be directly involved in lipid interactions. Theglobular C-terminal domain of SP-A clearly plays a central role in lipid binding, and in more complex functions such asthe formation and�or stabilization of curved membranes. In recent work, we have determined that the maintenance oflow surface tension of surfactant in the presence of serum protein inhibitors requires cooperative interactions betweenthe C-terminal and N-terminal domains of the molecule. This effect of SP-A requires a high degree of oligomericassembly of the protein, and may be mediated by the activity of the protein to alter the form or physical state ofsurfactant lipid aggregates. � 2001 Elsevier Science Inc. All rights reserved.

Keywords: Collectins; Innate immunity; Myelin figure; SP-A; SP-D; Surfactant inhibition; Surface activity; Tubular myelin

Abbre�iations: ARDS, acute respiratory distress syndrome; AP, alveolar proteinosis; CF, cystic fibrosis; CM, common myelin figure;CRD, carbohydrate recognition domain; DPPC, dipalmitoylphospatidylcholine; LB, lamellar body; MBL, mannose binding lectin; PBD,phospholipid binding domain; PI, phospatidylinositol; SP-, surfactant protein; TEM, transmission electron microscopy; TM, tubularmyelin

� Corresponding author: Tel.: �1-865-275795�275357; fax: �1-865-275729.Ž .E-mail address: [email protected] N. Palaniyar .

1095-6433�01�$ - see front matter � 2001 Elsevier Science Inc. All rights reserved.Ž .PII: S 1 0 9 5 - 6 4 3 3 0 1 0 0 3 0 9 - 9

( )N. Palaniyar et al. � Comparati�e Biochemistry and Physiology Part A 129 2001 109�127110

1. Introduction

Ž .Although surfactant protein A SP-A has beenunder investigation for a decade longer than anyof the other protein components of surfactant, itsfundamental role in pulmonary function has yetto be established. Previous studies based primar-ily on in vitro experiments have assigned severalroles for SP-A including the maintenance of sur-factant homeostasis, lipid sorting, tubular myelinŽ .TM formation, protection of surface activeproperties of surfactant in the presence of proteininhibitors, and in innate immune defense in the

Žlungs see reviews by Korfhagen et al., 1998a;.McCormack, 1998 . Transgenic mouse studies

confirm that SP-A is important for the formationŽ .of TM Korfhagen et al., 1996 , in the clearance

of bacterial and viral pathogens from the lungŽ .LeVine et al., 1997, 1999 and possibly in protec-tion of surfactant from protein-mediated inactiva-

Ž .tion Ikegami et al., 1998b; Elhalwagi et al., 1999 .These transgenic mouse studies do not support a

Žrole for SP-A in surfactant homeostasis Ikegami. Ž .and Jobe, 1998a although redundant pathway s

may have obscured this function in the modelsŽ .employed. Surfactant protein D SP-D , a struc-

turally related molecule in the alveolar liningfluid, appears to play a direct role in lipid

Žhomeostasis Botas et al., 1998; Korfhagen et al.,.1998b . In this review, we focus on the role of

SP-A in lipid-related functions, and compare itsdomains with corresponding segments of SP-Dand other related lectins.

2. SP-A domain organization and oligomericstructure

The primary structure of SP-A is conservedamong many organisms with some critical differ-

Ž . Ž .ences McCormack, 1998 Fig. 1 . Rat SP-A con-Ž . Ž .sist of 1 an N-terminal segment; 2 a collagen-

like region composed of Gly�X�Y repeats, whereŽ .Y is often hydroxyproline; 3 a hydrophobic am-

Ž .phipathic �-helical neck region; and 4 a globu-lar domain referred to as carbohydrate recogni-

Ž .tion domain CRD . The group of proteins thatcontain collagen-like domains and calcium-dependent lectin activity is known as the collectinfamily, which includes SP-A, SP-D, mannose

Ž .binding lectin MBL , conglutinin and CL-43.SP-A assembles into basic trimeric subunits by

helix formation in the collagen-like region andbundling of alpha helical coiled coil domains inthe neck, which then further assemble into disul-fide-dependent oligomers consisting of up to 18

Ž .monomers 6-trimers . Electron microscopy re-veals that, like C1q, the fully assembled moleculeis arranged in a fashion similar to a ‘bouquet of

Žtulip flowers’ Voss et al., 1988; Palaniyar et al.,.1998a; Ridsdale et al., 1999 . Bovine SP-A also

forms calcium-dependent supraquaternary struc-tures composed of protein filaments that are visi-

Ž .ble by transmission electron microscopy TEM ,but which as yet have not been visualized in

Žnative surfactant preparations Palaniyar et al.,. Ž .1998b, 1999a Fig. 2 .

3. N-terminal segment

3.1. Structure

The N-terminal segment of SP-A is short, con-sisting only of 7�10 amino acids, whereas thecorresponding segment of SP-D consists of 25amino acids. There are three interesting struc-tural motifs in this region including two inter-chain disulfide-forming cysteines, a conserved hy-drophobic cluster surrounding the cysteine closestto the collagen-like region, and a site for N-linkedglycosylation. SP-A from many organisms includ-ing rat, mouse, and cow contains a complex as-paragine-linked oligosaccharide moiety at the N-

Ž .terminus McCormack, 1998 . TEM images offixed surfactant preparations contain SP-A-likemolecules that exhibit intense staining at the ends

Ž .of the molecule Williams et al., 1991 , whichpresumably represents the sugars decorating theN-terminal fragment. Deletion of the N-terminaloligosaccharide moiety does not affect SP-A-

Žmediated lipid vesicle aggregation McCormack.et al., 1994a , and its function is not known.

By TEM, the N-terminal segment of SP-D andŽconglutinin are visualized as a central hub Lu et

.al., 1993; Hoppe and Reid, 1994 . The hub main-tains the trimeric SP-D subunits in an ‘anti-paral-lel’ organization and confers a cruciform shapeon the fully assembled dodecamer. In contrast,the N-terminal segment of SP-A promotes a‘parallel’ orientation of trimeric subunits, so thatthe fully assembled octadecamer forms a bouquetof flowers-like arrangement.

Rat SP-A is composed of two isoforms which

( )N. Palaniyar et al. � Comparati�e Biochemistry and Physiology Part A 129 2001 109�127 111

Ž . Ž .Fig. 1. Structural domains of SP-A A�C . A Basic trimer assembly. The N-terminal segment contains one or two disulfidebond-forming Cys. The collagen-like domain contains a kink, and is flexible. The hydrophobic neck region mediates protein�proteininteractions between intratrimeric polypeptide chains, and the junctions may be flexible. The C-terminal globular domain contains both

Ž . Ž . Ž .CRD and PBD. B and C Octadecamer assembly showing multivalent ligand binding domains in a top view B , and a side view C .Ž .Structural domains of SP-D. D Dodecameric assembly of SP-D showing ‘X-like’ organization. Overall dimensions of SP-A

octadecamers and SP-D dodecamers are �20 nm and �90 nm, respectively. Diagrams B�D are drawn approximately to scale to showrelative dimensions of SP-A and SP-D molecules.

are distinguished by the presence or absence of acysteine containing tripeptide extension at the

Ž .N-terminus Elhalwagi et al., 1997 . Most of theSP-A polypeptide chains isolated from rat lavagehave only a single Cys at the sixth amino acid

Ž 6.position of the mature peptide Cys but SP-Dconsistently contains two Cys that are separated

Ž . 6by four amino acids Crouch, 1998 . The Cys ofSP-A has been shown to participate in the inter-

Žchain disulfide bond formation McCormack et.al., 1997c, 1999 , and is required for the stability

of quaternary structure of SP-A. The minor iso-form of rat SP-A, which contains the additionaldisulfide bond forming cysteine at location-1,makes up approximately 20% of the molecule. Invitro mutagenesis experiments show that the ad-ditional Cys�1-containing isoform arises due to acombination of alternative translation initiation

and alternative cleavage by signal peptidaseŽ .Damodarasamy et al., 2000 . Human SP-Aisolated from normal volunteers and alveolar pro-teinosis patients also shows N-terminal hetero-

�1 Žgeneity with additional Cys Elhalwagi et al.,.1997 . Full disulfide dependent oligomeric assem-

bly of rat SP-A requires the presence of Cys�1,but it is not essential for the ability of SP-A toform higher order oligomers through non-cova-lent interactions between trimeric subunits andinterchain disulfide bond formation at Cys6

Ž .Elhalwagi et al., 1997 .Additional evidence that interchain disulfides

are not required for association of SP-A subunitsis that rat SP-A which is reduced and alkylatedretains the ability to assemble into highly orga-

Ž .nized oligomers Kuroki et al., 1988 , and dogSP-A, which does not contain Cys�1, forms higher


Fig. 2. Calcium-dependent conformational changes andŽ .oligomeric assembly of SP-A. A Electron micrographs of

ŽSP-A octadecamers in open and compact bouquet forms top. Ž .row , and supraquaternary forms bottom row . The conforma-

tional changes in octadecamers and the formation ofŽ .supraquaternary structures are mediated by calcium. B Bind-

ing of quaternary and supraquaternary forms of SP-A to thelipid film. Gray surface represents lipid film. Each SP-Amolecule is �20 nm.

Ž .order oligomers Voss et al., 1988 . All cDNAs forSP-A that have been isolated to date, except dog,have a Cys�1 in the propeptide chain but N-terminal amino acid sequence analyses of SP-Asfrom additional organisms will be necessary todetermine if the existence of minor cysteinyl iso-form is a consistent theme. MBL-C, a serumcollectin which has only one Cys in the N-termi-

Ž .nal segment Wallis and Drickamer, 1999 fails toform higher order oligomers. In contrast to SP-A,in vitro mutagenesis studies of the N-terminal Cysin SP-D shows that these Cys are essential for

Žoligomer formation Brown-Augsburger et al.,.1996 . SP-D collapses to trimeric subunits when

Cys are substituted by Ser, implying thatcollagen-like triple helices of SP-D do not havethe same attraction for each other as the colla-gen-like regions of SP-A. Collectively, these stud-ies illustrate differences in the functional impor-tance of the Cys found in the N-terminal segmentof collectins.

3.2. Lipid binding and aggregation

Although mechanisms of lipid binding and vesi-cle aggregation mediated by SP-A are not com-pletely understood, the oligomeric assembly ofSP-A is clearly important for these functions.Disulfide-independent trimeric assembly is suffi-cient for carbohydrate binding because a recom-binant SP-A that lacks N-terminal and collagen-

Ž � N1�P80.like domains rat SP-A trimerizes andŽbinds to mannose affinity columns McCormack

.et al., 1997c, 1999 . This protein, however, fails toaggregate phospholipid vesicles. Deletion of only

Ž �G8�P80.the collagen-like domain rat SP-A , leav-ing all other domains intact including N-terminal

Ž .segment N1-A7 , does not block the ability of thetruncated protein to bind either mannose or

Ž .phospholipids McCormack et al., 1999 . Unlikethe SP-A� N1�P80, the SP-A�G8�P80 retains the

Žability to aggregate phospholipid vesicles Mc-.Cormack et al., 1997c . Biochemical, chromato-

Ž .graphic McCormack et al., 1999 , and TEMŽ .Palaniyar et al., 1999c analyses of rat SP-A�G8�P80 suggest that this recombinant proteinforms hexamers and trimers but fails to formhigher order oligomers. These data indicate thatN-terminal segment is required for proteinoligomerization and phospholipid vesicle aggrega-tion.

Early studies using fixed surfactant materialsand low resolution TEM suggested that N-termi-nal segment of SP-A directly interacts with lipid

Žmembrane surfaces Beckmann and Dierichs,.1984; Williams et al., 1991 . Extensive in vitro

mutagenesis experiments conducted in the lastdecade, however, show that the primary phospho-

Ž .lipid binding domain PBD of SP-A is located inŽthe C-terminal globular domain see review, Mc-

.Cormack, 1998 . Recent studies using high resolu-tion TEM with ‘gentle’ staining procedures on

Žunfixed proteolipid complexes Palaniyar et al.,.1998a , support binding via globular C-terminal

domain with the N-terminal fragment orientedŽ .away from the membrane Fig. 3 . Hence, the

N-terminal segment is required for both lipidaggregation and quaternary structural assemblyof collectins but not directly interact with lipidmembranes.

3.3. Lipid structures

Although the N-terminal segment does not ap-


Fig. 3. Calcium-dependent changes in SP-A structure andŽ .effect on lipid interaction. A Interaction of SP-A octade-

camers with lipid bilayers. Shortly after addition of Ca2�,SP-A octadecamers exist in an open bouquet form and inter-act with lipids via a few globular domains. After a moreextended incubation period, SP-As exist in a closed bouquetform and interact with all the globular domains. SP-As thatinteract with curved membrane surfaces bind with multipleglobular domains and stabilize the membrane curvature. Dia-gram represents the multivalent interactions of globular do-

Ž .mains of an octadecamer with a membrane curvature. BInteraction of supraquaternary forms of SP-As with regularlyorganized membrane curvatures. Scale bar�20 nm.

pear to interact with lipid membranes, it clearlyaffects the formation of complex lipid aggregate

Ž .structures such as common myelin figures CM , asurfactant structure containing multiple concen-tric membranes with clearly identifiable in-

Žtermembranous spacing, and TM Palaniyar et al.,.2000b, unpublished data . In reconstitution

experiments using surfactant isolated from

Ž .SP-A �� mouse and mutant recombinantproteins produced in insect cells, we found thatthe N-terminal segment and collagen-like do-mains were required for TM, but not CM forma-tion.

3.4. Surface acti�ity

Disruption of the Cys6 interchain disulfide bondblocks the ability of SP-A to enhance the surfaceactivity of bovine surfactant in the bubble surfac-

Ž .tometer McCormack et al., 1997c . Recent,bubble surfactometry studies indicate that fea-tures of the N-terminal segment other than disul-fide dependent assembly are also important formaintaining lower surface activity of surfactant in

Žthe presence of protein inhibitors Ikegami et al.,.unpublished data .

4. Collagen-like domain

4.1. Structure

Although collagen-like domain consists ofGly�X�Y repeats, the Y amino acids of SP-Aand SP-D are often different. SP-A contains pro-

Žline or hydroxyproline in this position Mc-.Cormack, 1998 whereas SP-D contain hydroxyly-

Ž .sine or hydroxylysyl glycosides Crouch, 1998 . Ina typical collagen triple helix, the small aminoacid, Gly, occupies the central axis, and the pro-line ring structure provides the curvature of the

Ž .helix Kramer et al., 1999 . Hydrogen bonding via�OH group of the hydroxyproline or hydroxyly-sine holds the individual peptide chains withintriple helices. Intertriple helices interactions ap-pear to be mediated primarily by salt bridging,Van der Waal forces and electrostatic interac-tions. Hence, pH, the presence of various ionsand ionic strength can affect the structure of

Žcollagen Palaniyar et al., 1998a; Ridsdale et al.,.1999 . Since SP-A contains a collagen-like region

Ž .and is acidic isoelectric point of �4.5 , its struc-ture would be predicted to be sensitive to pHchanges of the alveolar microenvironment or ex-perimental buffers.

4.2. Protein oligomerization and flexibility of domains

Interruption of the Gly�X�Y repeat by inser-tion of a single amino acid is sufficient to cause


disruption of the triple helical organization ofpeptide chains. Such an alteration occurs in hu-man SP-A at Gly 44, which corresponds to the

Ž .bend or ‘kink’ detected in the collagen-like do-main visualized by TEM of dog or recombinant

Ž .human SP-A Voss et al., 1988 . The presence ofŽthe kink is also detected in bovine SP-A Palaniyar

.et al., 1998a; Ridsdale et al., 1999 . These resultsshow that triple helices in the collagen-like regionof SP-A can self-associate up to the kink similar

Ž .to C1q Lu et al., 1993; Hoppe and Reid, 1994 .The kink in the MBL is often not visible by TEM,perhaps due to its close proximity to the N-termi-

Žnal domain Lu et al., 1993; Hoppe and Reid,.1994 . The angle of the bend in SP-A is changed

in the presence of various cations and underŽconditions of varying ionic strength Palaniyar et

.al., 1998a; Ridsdale et al., 1999 . These structuralchanges may play an important role in the flex-ibility and orientation of the globular headgroupsof the collectins in relation to lipid or carbohy-drate ligands, in a manner that contributes tocooperative binding of multiple lectin domains to

Ž .multivalent ligands Figs. 1�3 .The four- to fivefold differences in the longest

Ž . Ždimensions of SP-A 20�25 nm and SP-D 90�100.nm determined by TEM are primarily at-

tributable to differences in the lengths of theircollagen-like domains, a structural feature thatmay play an important role in collectin function.Since the collagen-like domain of SP-A is perpen-dicular to the plane of the membrane when SP-A

Ž .binds to lipid monolayers Palaniyar et al., 1998aits length could determine the intermembranousspacing. Intermembranous spacing of lipid mix-tures generated by the covalently cross-linkeddouble length SP-A molecules, isolated from alve-olar proteinosis patients, correspond to two unitlengths of SP-A whereas normal SP-As generatespacing equivalent to one unit length of SP-AŽ . Ž .�20 nm Hattori et al., 1996 . A similar effecton the structure of lipid aggregates has also beenshown for SP-D. In vitro surfactant reconstitutionexperiments with mixtures of SP-D and phos-

Ž .pholipids, dipalmitoylphospatidylcholine DPPC �Ž .phospatidylinositol PI , lipid membranes were

separated by distances that were roughly equiva-Ž .lent to the longest dimension of SP-D �90 nm

Ž .Poulain et al., 1999 . Our recent experimentswith mutant recombinant SP-A that contain tele-scoping deletions in collagen-like domains revealthat the length of the collagen-like domain de-

Žtermines the intermembranous spacing Palaniyar.et al., 2000b . These data suggest to us that this

domain plays an important role in the structureof complex lipid aggregates such as CM and TM.

The presence of covalently linked oligosaccha-ride moieties in collagen-like region of lung lectinsis not clearly understood. SP-D contains an N-lin-

Ž 70 .ked oligosaccharide moiety Asn at approxi-mately 2�5 of the distance from the beginning ofthe collagen-like domain. The presence of shortO-linked oligosaccharides in many other colla-gens has been described and they do not interfere

Ž .with trimerization Bann et al., 2000 . This moietycan provide efficient compaction of collagen triplehelices particularly when proline or hydroxypro-lines are not abundant in X or Y locations. TheO-linked glycosylated hydroxylysines present inSP-D likely provide additional stability for thecollagen-like triple helices.

4.3. Self-aggregation and self-association of SP-A

Intact SP-A from many species including dog,cow, pig and human self-aggregates at low pH

Žand�or at high ionic strength Efrati et al., 1987;McCormack, 1998; Palaniyar et al., 1998b; Ruano

.et al., 1998a . Some of the aggregates are irre-Ž .versible Ruano et al., 2000 , probably due to the

denaturation of the collagen-like domain. Thefailure to refold properly may be due to thedisequlibrium of the hydrogen bonds and�or toincomplete reformation of the salt or ion bridgesbetween helix-stabilizing amino acids in the colla-gen-like domains. Deletion of the collagen-likedomain results in the loss of self-aggregation abil-

Ž .ity of SP-A Haagsman et al., 1990 , suggestingthat this domain may be directly involved in self-aggregation or self-association. Differential hy-droxylation of the collagen-like domain has been

Ž .detected in native Phelps and Floros, 1988 andŽrecombinant SP-As Floros et al., 1986; McCor-

.mack et al., 1994a , which may yield collagen-likedomains of variable stability, and may influencethe reversibility of aggregate formation.

In addition to self-aggregation, SP-A also self-associates in a more ordered configuration ofreversible, calcium-dependent supraquaternary

Ž .structures Palaniyar et al., 1998b . The supraqua-ternary structural form of bovine SP-A, in turn,forms extensive protein networks when it inter-

Žacts with phospholipid monolayers Palaniyar et. Ž .al., 1998b Fig. 2 . The SP-A:SP-A interactions in


these protein networks appears to occur in�ornear the globular domain and not via the distalpart of the collagen-like domain. Our presentunderstanding is that calcium binding causes con-formational changes in the globular domainŽ .Sohma et al., 1993; McCormack et al., 1994a ,and also cause the octadecamer to assume a more

Žcompact configuration Palaniyar et al., 1998a;.Ridsdale et al., 1999 . Palaniyar et al. proposed

that calcium binding results in alterations incharge distribution, which induces ionic and hy-drophobic attraction between the C-terminal do-mains or nearby regions of adjacent SP-A

Ž .molecules Palaniyar et al., 1998b , and causesself-association�protein network formation. Re-moval of calcium relaxes the compacted confor-mation to the open bouquet form, separates theglobular domains in space, and disaggregates SP-Anetworks.

The collagen-like region of human SP-As de-Ž .rived from the first gene SP-A1 contains an

additional Cys at the 85th amino acid positionŽthat could form interchain disulfide bonds Rishi

.et al., 1992; Berg et al., 2000; Ramet et al., 2000 .This feature, which does not occur in SP-A2 geneproducts, may lend additional complexity to higherorder structural assembly of SP-A, in vivo. Inparticular, the structure of human SP-A could bedifferent depending on the genotype of the indi-vidual and the region of the lung where the SP-Ais expressed. A disulfide cross-link in the mid-point of the collagen-like region may enhance thestability of SP-A composed of both SP-A1 andSP-A2 gene products, compared to SP-A com-posed only of SP-A2 gene products. The formergene product combination thought to occur in thealveolus whereas the latter is thought to occur in

Ž .the large airways Saitoh et al., 1998 . It is alsopossible that the collagen-like region of a moreextensively cross-linked and rigid molecule wouldinteract differently with protein receptors andligands.

Although SP-A isolated from pulmonary alve-olar proteinosis patients is used in many experi-ments because of the availability of the lavagesamples, whether this SP-A is representative ofthe normal human SP-A is uncertain. SP-Aisolated from alveolar proteinosis patients has adifferent structure and contains non-reduciblebonds that interconnect polypeptide chains withinthe oligomer, but retains the lipid binding activity

Žand the lectin activity of the native protein Hat-

.tori et al., 1996 . This protein modification, how-ever, causes SP-A to form parachute-like macro-molecules and atypical lipid structures in tubular

Žmyelin reconstitution experiments Hattori et al.,. Ž .1996 . The location s of these non-reducible

bonds in SP-A are not known.

4.4. Lipid binding and aggregation, formation of CMand TM

Fibrillar collagen molecules can bind to DPPCŽ .lipid membranes Ghannam et al., 1999 . Whether

the collagen-like domain of SP-A directly inter-acts with the surfaces of phospholipid membranesto a significant extent is not known, however.SP-A containing a deletion of the collagen-likedomain retains lipid aggregation properties, andlipid binding is diminished by sedimentation anal-

Žyses but detectable by TEM Palaniyar et al.,.2000a . These data indicates that the presence of

the collagen-like domain in SP-A facilitates itsinteraction with phospholipid membranes by pro-moting the association of multiple PBD with sub-strates. In vitro reconstitution of lipid structures

Ž .using surfactant isolated from SP-A �� miceand rat recombinant proteins was employed todetermine the domains of SP-A that are requiredfor the formation of CM, and more complexaggregates such as TM. These studies demon-strated that the length of the collagen-like do-main determines intermembranous spacingŽ .Palaniyar et al., 2000b, unpublished data , andthat the collagen-like domain was required forthe correction of TM formation in the SP-Ž . Ž .A �� mouse McCormack et al., 2000 .


Recombinant rat SP-A containing a deletion ofcollagen-like domain retains the ability to im-prove the surface active properties of surfactant

Ž .lipids, in vitro McCormack et al., 1997c . Thissuggests that collagen-like domain is not essentialfor the surface tension lowering properties ofSP-A. Circular dichroism analysis of SP-A or itsfragments suggests that the collagen-like domainis important for the stability of the proteinŽ .Haagsman et al., 1987; Ruano et al., 2000 . Sincelipid aggregation by this truncated recombinantprotein was blocked at 37�C, either temperaturedependent protein�lipid interactions or the con-formational stability of the protein trimers ap-


peared to be important for this function. Recentdata indicates that the collagen-like domain isrequired for the property to block the activity ofprotein inhibitors on the surface activity of sur-

Ž .factant in vitro Ikegami, unpublished data andŽ .in vivo McCormack et al., 2000 .

5. Neck region

5.1. Structure

The region that connects the collagen-like do-main and the globular C-terminal domain of SP-Acontains segments that are predicted to from a

Ž35-amino acid long amphipathic �-helix McLean.et al., 1993 . The X-ray crystallographic data forŽ .the SP-D Hakansson et al., 1999 and MBL

Ž .Sheriff et al., 1994; Weis and Drickamer, 1994neck shows the existence of a coiled-coil organi-zation. Molecular modeling of SP-A trimers sug-gests that a similar structure likely to exist in

Ž .SP-A as well Palaniyar et al., 2000a . A tyrosinelocated at the end of the hydrophobic region ofthe neck is detected in human SP-D. Comparison

Ž .with similar structures Shu et al., 2000 suggeststhat this tyrosine acts as a hydrophobic cap at theend of the coiled-coil domain that stabilizes thecoiled-coil bundle.

5.2. Protein trimerization

Although other functions have previously beenŽenvisaged for the neck domain Ross et al., 1986;

.McLean et al., 1993 , recent evidence suggestits primarily role is to mediate intertrimeric pro-

Ž .tein�protein interactions Palaniyar et al., 2000a .ŽRecombinant forms of both SP-A McCormack et

. Žal., 1999 and SP-D Hoppe et al., 1994;.Hakansson et al., 1999 that contain only the neck

and the globular domain trimerize in solution.Nucleation of trimerization within the neck do-main is critical for the processing and secretion of

Ž .SP-A Spissinger et al., 1991 . Oligomerizationcan occur in both directions in synthetic

Ž .collagen-like peptides Liu et al., 1998 but dataŽ .from MBL-C Wallis and Drickamer, 1997 and

Ž .SP-D Hoppe et al., 1994 suggests that folding inthe collagen-like domains of SP-A probablyproceeds in a zipper like fashion along the C-terminal to N-terminal axis. Triple helix forma-tion in the collagen-like domain aligns the pep-

tide chains for N-terminal disulfide bond forma-tion. Further experiments will be required to de-termine the exact mechanism of oligomerizationin the collagen-like domain of lung collectins.

5.3. Flexibility of the globular domain

The globular domains of SP-D and MBL areoriented in space by the hydrophobic interactionsof three neck domains that are assembled in a

Žcoiled-coil pattern Weis and Drickamer, 1994;.Drickamer, 1999; Hakansson et al., 1999 . This

neck domain is folded such that the three peptidechains are aligned in perfect register. In the adja-cent collagen triple helix, in contrast, stable bun-dle formation required that the three peptidechains are shifted out of register with respect to

Ž .each other Kramer et al., 1999 . Therefore, theremust be a ‘swivel’ region between the neck and

Ž .the collagen-like domain Fig. 1A which func-tions as an adapter between the aligned anddisaligned segments. The presence of adapters

Žhas been suggested for MBL-C Wallis and.Drickamer, 1997 and for the scavenger receptor,

which has a C-terminal domain with a range ofmotion around the axis of the neck of 0�180�Ž .Resnick et al., 1996 . This additional flexibilityfor the globular domain may enable the ligandbinding domains of collectins to accommodateuneven surface structures. Simultaneous bindingof multiple globular domain units within eacholigomer enhances binding affinity for polyvalentligands such as membrane surfaces and oligosac-charides decorating microorganisms. Another po-tential region for flexibility in the collectins is thejunction between the coiled-coil neck and theglobular domain. Further studies are needed todefine the physiological relevance of flexible re-gions in collectins.

5.4. Lipid binding

Ž .Ross et al. 1986 suggested that the neck re-gion of SP-A is important for binding to lipid

Ž .vesicles, and McClean et al. 1993 demonstratedthat short synthetic peptides corresponding to theneck region bound to lipid membranes and im-proved surface active properties of surfactant.Data from mutagenesis studies, however, suggestthat CRD and not neck domain contains the

Ž .primary lipid binding site McCormack, 1998 .Recently, mutant recombinant SP-A containing a


deletion of the collagen-like domain was imagedcomplexed with lipid monolayers using TEMŽ .Palaniyar et al., 2000a . The three-dimensionalstructure of this protein was reconstructed andfitted with a homology-based molecular modelbased of MBL. The data obtained indicate thatthe neck region of SP-A trimers does not interactwith lipid membranes.

6. Globular domain

6.1. Structure

The ability of the globular domain of lungcollectins to interact with multiple ligands includ-ing carbohydrates, calcium, phospholipids and

Žlipopolysaccharides McCormack, 1998; Clark et.al., 2000 is critical for performance of their anti-

microbial and surfactant functions. The CRD ofrat SP-A contains 115 amino acids that fold intovariety of secondary structures including �-sheet,�-helices and random coils. The domain is

Žstabilized by two pairs of disulfide bonds Berg et.al., 2000; Palaniyar et al., 2000a . A covalently-

linked oligosaccharide moiety is found in all SP-Asstudied but not in SP-D from most of species

Ž .except pig van Eijk et al., 2000 . Haagsman et al.Ž .1990 reported that SP-A binds two to threeCa2� ions per monomer, with at least one highaffinity site in the CRD. The X-ray crystallo-graphic structures of SP-D and MBL suggest thatthere may be two Ca2� binding sites in the globu-

Žlar domain of SP-A Weis and Drickamer, 1994;. 2�Hakansson et al., 1999 . Ca binding at one of

these sites may also coordinate the binding ofcarbohydrate ligands, as it has been shown to do

Ž .for MBL Weis et al., 1992 . Calcium causesconformational changes in the globular domain ofSP-A, and appears to increase the �-sheet con-

Ž .tent Sohma et al., 1993 and resistance to pro-Žteases Haagsman et al., 1990; McCormack et al.,.1994a . Binding of calcium ions also changes the

charge distribution on the surface of the globularŽdomain of many collectins Hakansson and Reid,

.2000 .

6.2. Lipid binding

Although the amino acid sequence of globulardomain is conserved among C-type lectins, dif-ferent lectins bind different lipids. For example,

ŽSP-A preferentially binds to DPPC Kuroki and.Akino, 1991; Yu et al., 1999 whereas SP-D and

ŽMBL bind to PI Ogasawara et al., 1994; Chiba et.al., 1999 . Association of SP-A with lipids has

Žbeen reported in the absence of calcium King et. Žal., 1983 , but in some other studies Kuroki and

. ŽAkino, 1991 and in our hands McCormack et.al., 1997a,b specific and high affinity binding to

phospholipids requires at least �M concentra-Žtions of calcium Ruano et al., 1996; Meyboom et

.al., 1997, 1999 . Depending on the experimentalconditions employed, SP-A appears to interact

Ž .with both acyl chains Kuroki and Akino, 1991Ž .and polar head groups Yu et al., 1999 of phos-

pholipids. In monolayers, SP-A appears to inter-act with the polar headgroup of the lipid

Ž .molecules Yu et al., 1999 . The TEM�molecularmodel of recombinant rat SP-A trimers bound tophospholipid monolayers predicts the amino acidswhich are located at the protein membrane inter-face, and which may interact with lipid mem-

Ž .branes Fig. 4 . These data are consistent with invitro mutagenesis and monoclonal antibody ex-periments which have identified the region of theshort disulfide loop of the CRD of rat SP-A asthe region required for phospholipid bindingŽKuroki et al., 1994; Ogasawara et al., 1994;

.Tsunezawa et al., 1998 . Since specific phospho-Žlipid binding is calcium dependent Kuroki and.Akino, 1991; McCormack et al., 1997c , and point

mutations of putative calcium binding residuesŽinterfere with lipid binding McCormack et al.,

.1997b,c , these residues are either directly in-volved in the lipid binding or responsible forconformational shifts which exposes remote lipidbinding sites. The phospholipid binding domain ofSP-A reveals important amino acid sequence dif-ferences when compared with the corresponding

Ž .regions of SP-D and MBL. Kuroki et al. 1994have demonstrated that the binding preference ofSP-D can be conferred upon SP-A by introducingcritical sequences from the globular domain ofSP-D into the corresponding region of SP-A. Amore complete understanding of molecular mech-anism of lipid binding of lung lectins will certainlybe enhanced by the solution of the crystal struc-ture of SP-A complexed with a lipid ligand.

6.3. Lipid aggregation

SP-A mediates phospholipid vesicle aggrega-Ž .tion in vitro McCormack, 1998 ; a function that


Fig. 4. Binding of individual SP-A trimers to lipid membranesand a homology-based molecular model that is fitted inside

Ž .the reconstituted 3D-electron density map. A The TEMimages of rat recombinant SP-A, which contains a deletion of

Ž �G8�P80 .the collagen-like domain SP-A , in association with aŽ .lipid film. Some individual trimers are circled. B View of

reconstructed 3D structure of SP-A�G8�P80 at CRD�lipid filminterface. Glu195, blue; Arg197, green; N-linked glycosylationsite, i.e. Asn187, white.

could assist with lipid sorting and recycling in thealveolar environment. Since the SP-A gene tar-geted mouse did not show significant alterations

Žin lipid homeostasis Korfhagen et al., 1996;.Ikegami et al., 1997 or impaired surface tension

under resting or stressed situations compared withŽ .wild type animals Ikegami et al., 1998b , SP-A

may not be essential for these functions in vivo. Arequirement for SP-A in surfactant function orhomeostasis under different types of stress condi-tions has not been completely excluded, however.

Changes in the optical density of the surfactantprotein and lipid mixtures suggests that SP-A canaggregate phospholipid vesicles in both a calcium

Ždependent and a pH-dependent manner Efrati etal., 1987; Haagsman et al., 1990; Ruano et al.,

.1998a, 2000 . Differences in the mechanisms ofthese types of vesicle aggregation are not known.In vitro mutagenesis experiments have identifiedthe amino acid residues that influence the SP-A-mediated phospholipid vesicle aggregation in the

Žglobular domain Sano et al., 1998; Saitoh et al.,.2000 . Alanine mutagenesis of putative calcium

Ž E195Acoordinating residues of SP-A SP-A , SP-E202A D215A .A , SP-A altered the calcium-dependent

lipid binding and vesicle aggregation ability of theŽ .protein McCormack et al., 1997b . These studies

showed that although rat SPAE195Q,R197D was ca-pable of binding to phospholipid, it failed to ag-gregate the vesicles, and hence, suggested that

Žthese two activities are separable McCormack et.al., 1994b . Similar mutagenesis studies suggested

that different amino acids in the same region ofthe CRD are required for the calcium-indepen-dent, the pH-dependent vesicle aggregation

Ž .property of SP-A McCormack et al., 1997b .These results suggest that the domains of SP-Athat mediate calcium and pH-dependent vesicleaggregation overlap but they are not identical.Since calcium ion concentration and pH are dif-ferent in various compartments of the alveolus orinside phagocytic or secretory cells, these twomodes of vesicle aggregation could be operativein different microenvironments.

6.4. Lipid monolayers

Whether lung surfactant forms a uniform lipidmonolayer at the alveolar air�liquid interface is

Žcontroversial Schurch et al., 1998; Grunder et al.,.1999 and the exact organization of the surfactant

in the alveolus is uncertain. TEM analysis ofsurfactant fixed in situ suggests that stacks of

Žmembranes are present at the interface see re-.view, Schurch et al., 1998 . This type of organiza-

tion could explain the interconversion of variouslipid forms such as vesicles, bilayer discs or lamel-

Ž .lar bodies LBs during the surface film formationin the lung. Several in vitro experiments con-ducted using surfactant or surfactant-like lipidmixtures suggest that SP-A interacts with thesurface films and assists in the reorganization of

Žlipid domains Nag et al., 1998; Ruano et al.,.1998b . A role for SP-A in the refinement of the

lipid film to cause enrichment in DPPC couldfacilitate the maintenance of surface active

Žproperties of surfactant Veldhuizen et al., 1998;.Yu and Possmayer, 1998 . In the presence of

serum proteins, the effect of SP-A in the mainte-nance of surface active properties of surfactant is

Žmore pronounced Cockshutt et al., 1990; Ueda et.al., 1994; Seidner et al., 1995 . Surfactant isolated

Ž .from SP-A �� mouse is more sensitive toserum protein inhibitors compared with the sur-

Žfactant isolated from wild type mouse Ikegami et.al., 1998b , and over expression of rat SP-A en-

Žhances resistance to protein inhibition Elhalwagi.et al., 1999 . Analysis of surface activity of surfac-

Ž .tant isolated from SP-A �� mice in the pres-ence of various mutant or wild type rat recombi-nant proteins indicates that the function requiresa cooperative interaction between the N-terminal


Žand C-terminal domains of the protein Ikegami.et al., unpublished data .

The mechanism of this SP-A-mediated protec-tion of surfactant is not known. Palaniyar et al.Ž .1998b have suggested that SP-A protein net-

Ž .works Fig. 2 , visible by TEM and fluorescenceŽ .microscopy Ruano et al., 1998b , could stabilize

the lipid film. In this scheme, protein inhibitorsmay interfere with the formation of protein net-works, defeating the stabilizing effect of SP-A.Preliminary data from Palaniyar et al. supportsthis notion. When incubated in the presence of acommon acute phase response protein, C-reactiveprotein, formation of protein networks by rat

ŽSP-A was partially inhibited Palaniyar et al., un-.published data . Recombinant SP-A containing a

Ždeletion of the collagen-like domain rat SP-�G8�P80.A , which failed to form protein networks

and interacted with lipid monolayers in an unor-Ž . Ž .ganized manner Palaniyar et al., 2000a Fig. 4 ,

also failed to reduce the surface activity of surfac-Žtant in the presence of protein inhibitors Ikeg-

.ami et al., unpublished data . Our working hy-pothesis is that the supraquaternary organizationof SP-A may prevent the collapse of the lipid filmin the presence of protein inhibitors. Alterna-tively, protein inhibitors may interact with otherforms of surfactant lipid aggregates and alter the

Ž .physical characteristics Nag et al., 2000 . Thesealterations may interfere with lipid adsorptionand the formation of the surface film, which isnecessary for the maintenance of minimum sur-face tension.

Since many disease conditions including acuteŽ .respiratory distress syndrome ARDS , alveolar

Ž . Ž .proteinosis AP and cystic fibrosis CF are asso-ciated with proteinaecious and lipidic exudates inthe airspace, SP-A may play a critical role inmaintaining a lower surface activity of surfactantin those conditions. For example, surfactantisolated from CF patients contains high concen-trations of lipids and proteins but relatively low

Žconcentrations of SP-A and SP-D Postle et al.,.1999; Meyer et al., 2000 . Surfactant isolated from

CF patients shows poor surface active propertiesŽ .Meyer et al., 2000 , and likely reflects the pres-ence of lower amounts of SP-A and increasedlevels of foreign proteins in the surfactant.Administering SP-A together with natural or syn-thetic surfactants to pups delivered preterm im-proved the lung compliance and surface tensionreduction properties of surfactant in the presence

Ž .of serum protein inhibitors Yukitake et al., 1995 .These results indicate a potential role for SP-A inmaintaining a lower surface tension in the pres-ence of serum protein inhibitors, in vivo. SinceSP-A knockout mouse studies under conditions ofstress and lung injury failed to show any clear

Ždisadvantage to SP-A deficiency Ikegami et al.,.1998b, 2000 , a role for SP-A in the protection of

surfactant is not clearly established in transgenicmouse models.

6.5. Lipid �esicles and common myelin

The fact that SP-A interacts with lipid vesiclesand alters their structure has been established by

Žreconstitution studies Suzuki et al., 1989;.Williams et al., 1991; Poulain et al., 1992 and by

analyzing the surfactant of different organismsŽ .Williams, 1977, 1978, 1982 . Recent studies haveexplored the protein�phospholipid interactionswhich are responsible for these structural effectsŽ .Palaniyar et al., 1998a, 1999a,b . The globulardomains of SP-A, in an octadecamer, form amushroom-like curved surface, and calcium in-duces a conformational change that moves thetrimers closer together. Upon binding to themembrane surface via the globular domain, cal-cium-induced conformational change could leadto a more efficient configuration in which all ofthe globular domains bind to the membrane sur-

Ž . Žface simultaneously Palaniyar et al., 1998a Fig..3 . In addition, the configuration of the mem-

brane surface may be a determinant of SP-Abinding. SP-A preferentially interacts with curvedsurfaces, perhaps because they are better fitted tothe contour of the CRD and the location of thephospholipid binding site within those domains.SP-A can also organize into elongated filamentsthat interact with curved surfaces found on the

Ž . Ž .membranes Palaniyar et al., 1999a Fig. 3 . Theseregions may harbor the highest affinity bindingregion for SP-A in the membrane, the gel-phase

ŽDPPC domains King et al., 1983; Casals et al.,.1993 . Multivalent interaction of SP-A with lipid

membranes in this context likely stabilizes thelipid structures, as well as the SP-A�SP-A inter-actions that occur in, or near, the globular do-

Ž .mains Palaniyar et al., 1998b .

6.6. Tubular myelin

TM has been detected in several organisms, but


its function and the mechanism of formation arenot clearly established. Formation of this lattice-like lipid structure is dependent on both SP-A

Ž .and SP-B in vitro Suzuki et al., 1989 , and in vivoŽ .Korfhagen et al., 1996; Weaver and Beck, 1999 .SP-A has been detected in the corners of thesquare lattice in the cross-section and along edgesin the longitudinal axis of the tubular structureŽWilliams, 1978; Beckmann and Dierichs, 1984;

.Voorhout et al., 1991; Palaniyar et al., 1999b .There is a central core within each box of thelattice that in some EM preparations appears

Želectron dense Beckmann and Dierichs, 1984;.Palaniyar et al., 1999b and in others as an empty

Ž .space Goerke, 1998; Voorhout et al., 1991 . Thesedifferences appear to be due to variations in thefixation and staining procedures and�or the dif-ferences among the animal species. In some ofthe better preserved TM structures, an electrondense ‘X’ connecting the corners of the TM lat-

Žtice could be seen Beckmann and Dierichs, 1984;.Palaniyar et al., 1999b , which likely represent

SP-A. In others, electron dense regions were seenat the corners of the square lattice, which are

Ž .likely to represent collapsed SP-A Fig. 5Ž .Goerke, 1998; Palaniyar et al., 2000c . Structureswhich appeared to be SP-A could also be visual-ized in TEM of transition phases; from lipidmembranes folding to form TM-like structures in

Ž .vitro Palaniyar et al., 1999b , and lamellar bodiesŽ .unfolding to form TM Williams, 1977 .

Ž .Palaniyar et al. 1999b have recently suggesteda model for the formation of TM based on obser-vations made on intermediate lipid structures and

Žresults from previous studies Williams, 1977,.1978, 1982; Poulain et al., 1992 . However, forma-

tion of TM from LBs in vivo is likely to bemore complex. Calcium-induced conformationalchanges of SP-A and possible steps in lipid mem-brane reorganization, based on current under-standing, are presented in Fig. 5. Several lines ofevidence suggest that SP-A is secreted via alamellar body-independent pathway, and subse-quently become enriched in the unraveling lipid

Ž .structures such as TM Savov et al., 2000 . Duringthis process, the membranes that are tightlypacked in LBs become loose, particularly at theouter layers, and at points of connection withTM. Compact forms of SP-A-like molecules couldbe detected on the surfaces of the interconnect-

Ž .ing membranes Williams, 1977 . These observa-tions provide clues to the understanding of the

Fig. 5. Possible steps in the formation of TM from LBs inalveolus and hypothetical role for SP-A in surfactant struc-

Ž .tural transitions. A Most of the SP-A molecules are secretedŽ .into the alveolar hypophase independently of the LBs. B The

free SP-A octadecamers can interact with outer layers of theLB membranes in a compacted form, and provide widerintermembranous spacing. SP-A may effect membrane curva-ture, or stabilize the existing curved surfaces on the lipid

Ž .membrane. C Folding and eventual SP-B-mediated fusion ofthe membranes can lead to the formation of TM with lattice-like tubes, where SP-As predominantly occupy corners of thelattices.

fundamental processes involved in interconver-sion of various lipid forms in the alveolus.


SP-D can also form TM-like structures in vitro,when PI is included as one of the phospholipid

Ž .components Poulain et al., 1999 . The lipid struc-tures under these conditions produced a hexago-nal�circular lattice with �100 nm diameter anda central electron dense target-like core. Themeasurements suggest that the central dot repre-sents the N-terminal hub of SP-D and the ringrepresents the oligosaccharides attached to thecollagen-like domain. This organization supportsthe idea that the N-terminal domain of the pro-tein occupies the central region of the tube whilethe globular domain interacts with surfaces of theinner side of the membranes. A similar structuralform of TM was observed in the surfactant se-

Žcreted by cultured type-II cells Poulain et al.,.1999 and in alveolar proteinosis patients

Ž .Takemura et al., 1987 . Poulain et al. suggestedthat the atypical TM may be formed under abnor-mal conditions, in which SP-D could compensatethe lipid structural alterations in the diseasedlung.

Lipid structures that are observed in SP-A orSP-D gene targeted mice are different from thenormal mice. SP-A knockout mice show virtually

Ž .no TM Korfhagen et al., 1996 , and overexpres-Ž .sion of rat SP-A in the SP-A �� background

Ž .corrects the defect McCormack et al., 2000 .SP-D knockout mice show condensed lipid struc-

Žtures and reduced amounts of TM Botas et al.,.1998; Korfhagen et al., 1998b , but the latter

could be attributable to the reduced levels ofSP-A present in the SP-D knockout mice.

TM has been thought to be an important inter-mediate lipid form that is essential for efficientinterconversion of the secreted LB to surfactantfilm. However, since lipid cycling and surfactant

Ž .homeostasis in the SP-A �� mouse is normalcompared with wild type mouse, TM is not essen-tial for this function under a variety of experi-

Žmental stress conditions Korfhagen et al., 1996;.Ikegami et al., 1998b . Therefore, the true func-

tion of TM remains to be defined, in vivo. It isconceivable that it could help to maintain thesurface activity of surfactant in the presence ofserum protein inhibitors. We have analyzed theeffect of various mutant rat SP-As expressed ininsect cells on the surface tension of surfactant

Ž .isolated from SP-A �� mouse in the pres-ence of serum protein inhibitors. These experi-ments revealed that TM formation is a complexfunction that requires both an intact phospholipid

binding domain in the CRD and intact oligomericŽ .assembly Ikegami, unpublished data .

In vitro experiments show that the lipid compo-sition and aggregate structure of surfactant are

Žimportant for its surface active properties Veld-.huizen et al., 1998 . Differences in the surfactant

composition of various air breathing organismsfrom lizard and amphibians to birds has alsoyielded insight into the structure and function of

Ž .surfactant Daniels et al., 1995, 1998 . Large ag-gregate surfactant isolated from various organ-isms is more active than small aggregate surfac-tant, and most of the SP-A and TM are detected

Žin the former fraction Ikegami and Jobe, 1998a;.Veldhuizen et al., 1998 . SP-A inhibits the conver-

sion of large aggregate surfactant to the inactiveŽ .small aggregate form Veldhuizen et al., 1996 .

Furthermore, lung preservation and maintenanceof surfactant activity after ischemia�reperfusiondepends on the preservation of TM and other

Ž .lipid structures Fehrenbach et al., 2000 . Theseresults suggest that various lipid structures, in-cluding TM, may play important roles in surfac-tant function.


Surfactant isolated from SP-A deficient mice ismore susceptible to protein inhibition than the

Žsurfactant from wild type mice Ikegami et al.,.1998b . Resistance to protein inhibition is re-

Ž .stored in the SP-A �� mouse by overexpres-Ž .sion of full length SP-A Elhalwagi et al., 1999 ,

but not by overexpression of SP-A containing aŽdeletion of the collagen-like domain McCormack

.et al., 2000 . This result indicates that the CRDalone is not sufficient for the protection of surfac-tant from inhibitors. We had previously notedthat mutations in the CRD of SP-A block theeffect of the protein to improve the surface activeproperties of bovine lipid extract surfactantmonolayers as they are compressed and expandedin the bubble surfactometer in the absence of

Ž .inhibitors Yu et al., 1999 . To address this issuefurther, we reconstituted surfactant from the SP-Ž .A �� mouse with mutant recombinant SP-As

produced in insect cells. All mutations in theCRD that affect lipid binding also blocked theability of SP-A to protect the surface activity of

Žsurfactant from protein inhibition Ikegami et al.,.unpublished data . Both wild type and SP-

Ž .A �� mice have similar and high lipid con-


Ž .centrations in the order of mg�ml in the alve-olus, and the SP-A gene does not confer anyadvantage during induced lung injuryŽ .Ikegami et al., 1997, 2000 . Hence, the SP-Amediated effect on surface activity may only berelevant when the surfactant level is com-promised.

7. Summary

SP-A contains discrete structural domains thathave distinct properties, but which functioncooperatively to perform specialized functions.Although trimer formation of SP-A is dependenton both the collagen-like region and the neckregion of SP-A, association of trimers to formoctadecamers is dependent on interchain disul-fide bond formation in the N-terminal segmentsand non-covalent interactions between the colla-gen-like domains. Self-association of SP-A andthe formation of supraquaternary filaments ofSP-A may be important for phospholipid vesicleaggregation and the formation of CM and TM, aswell as in protecting the integrity of the surfac-tant film in the presence of protein inhibitors.Although most of the activities of SP-A, includinglipid binding and lectin activities, are located inthe globular domain, the N-terminal domains areimportant for proper assembly and presentationof globular domains in an optimal orientation formultivalent�cooperative binding functions. Thesemultivalent interactions are important for bothhigh affinity recognition of oligosaccharides onthe surface of microorganisms, and in stable bind-ing to phospholipid membranes and surfactantaggregates.

8. Future perspectives-SP-A structure andfunction

Although extensive in vitro mutagenesis experi-ments conducted in the last decade have eluci-dated the domains involved in various propertiesof SP-A, many of the fundamental mechanismsinvolved in oligomer formation and protein func-tion are not completely understood. The geneti-cally engineered mouse models of SP-A de-ficiency and excess, which have becomeindispensable for the study of SP-A biology, alsoprovide useful tools for exploring structure�func-tion relationships in vivo. Molecular modeling of

SP-A, using both low resolution TEM techniquesand high resolution X-ray crystallographic struc-tures, will be used to study the effects of aminoacid substitutions in protein conformation andfunction. Ultimately, the impact of our work willbe judged on how data from the laboratory andanimal models helps us understand the impor-tance of alterations in SP-A structure, levels, andself-association in the pathogenesis of diseasessuch as alveolar proteinosis, pneumonia, idio-pathic pulmonary fibrosis and acute respiratorydistress syndrome.

Acknowledgements

It is a great pleasure to dedicate this review toRichard Pattle, Ph. D. who is a pioneer in the useof electron microscopy to understand the struc-ture-function relationships of the pulmonary sur-factant. We thank Drs Kenneth B.M. Reid andFred Possmayer for providing critical commentson the manuscript. This work was supported byPostdoctoral Fellowships of American Lung Asso-

Ž .ciation N.P , Canadian Lung Association�Ž .Canadian Medical Research Council N.P , and

Ž .MRC-Welcome Trust, UK N.P. , and Career In-vestigator Award from American Lung Associa-

Ž .tion F.X.M , Veterans’ Affairs Merit AwardŽ . Ž .F.X.M , NIH HL61612 F.X.M , and HL 61646Ž .M.I., T.K., J.W. .

References

Bann, J.G., Peyton, D.H., Bachinger, H.P., 2000. Sweetis stable: glycosylation stabilizes collagen. FEBS Lett.473, 237�240.

Beckmann, H.J., Dierichs, R., 1984. Extramembrane-ous particles and structural variations of tubularmyelin figures in rat lung surfactant. J. Ultrastruct.Res. 86, 57�66.

Berg, T., Leth-Larsen, R., Holmskov, U., Hojrup, P.,2000. Structural characterisation of human pro-teinosis surfactant protein A. Biochim. Biophys. Acta1543, 159�173.

Botas, C., Poulain, F., Akiyama, J. et al., 1998. Alteredsurfactant homeostasis and alveolar type II cell mor-phology in mice lacking surfactant protein D. Proc.Natl. Acad. Sci. USA 95, 11869�11874.

Brown-Augsburger, P., Hartshorn, K., Chang, D. et al.,1996. Site-directed mutagenesis of Cys-15 and Cys-20of pulmonary surfactant protein D. Expression of atrimeric protein with altered anti-viral properties. J.Biol. Chem. 271, 13724�13730.


Casals, C., Miguel, E., Perez-Gil, J., 1993. Tryptophanfluorescence study on the interaction of pulmonarysurfactant protein A with phospholipid vesicles.Biochem. J. 296, 585�593.

Chiba, H., Sano, H., Saitoh, M., Sohma, H. et al., 1999.Introduction of mannose binding protein-type phos-phatidylinositol recognition into pulmonary surfac-tant protein A. Biochemistry 38, 7321�7331.

Clark, H.W., Reid, K.B., Sim, R.B., 2000. Collectinsand innate immunity in the lung. Microbes Infect. 2,273�278.

Cockshutt, A.M., Weitz, J., Possmayer, F., 1990. Pulmo-nary surfactant-associated protein A enhances thesurface activity of lipid extract surfactant and re-verses inhibition by blood proteins in vitro. Biochem-istry 29, 8424�8429.

Crouch, E.C., 1998. Structure, biologic properties, andŽ .expression of surfactant protein D SP-D . Biochim.

Biophys. Acta 1408, 278�289.Damodarasamy, M., Zhang, M., Dienger, K., McCor-

mack, F.X., 2000. Two rat surfactant protein A iso-forms arise by a novel mechanism that includesalternative translation initiation. Biochemistry 39,10189�10195.

Daniels, C.B., Orgeig, S., Smits, A.W., 1995. The evolu-tion of the vertebrate pulmonary surfactant system.Physiol. Zool. 68, 536�539.

Daniels, C.B., Orgeig, S., Wood, P.G., Sullivan, L.C.,Lopatko, O.V., Smits, A.W., 1998. The changingstate of surfactant lipids: New insights from ancientanimals. Am. Zool. 38, 305�320.

Drickamer, K., 1999. C-type lectin-like domains. Curr.Opin. Struct. Biol. 9, 585�590.

Efrati, H., Hawgood, S., Williams, M.C., Hong, K.,Benson, B.J., 1987. Divalent cation and hydrogen ioneffects on the structure and surface activity of pul-monary surfactant. Biochemistry 26, 7986�7993.

Elhalwagi, B.M., Damodarasamy, M., McCormack, F.X.,1997. Alternate amino terminal processing of surfac-tant protein A results in cysteinyl isoforms requiredfor multimer formation. Biochemistry 36, 7018�7025.

Elhalwagi, B.M., Zhang, M., Ikegami, M. et al., 1999.Normal surfactant pool sizes and inhibition-resistantsurfactant from mice that overexpress surfactantprotein A. Am. J. Respir. Cell. Mol. Biol. 21,380�387.

Fehrenbach, A., Ochs, M., Warnecke, T. et al., 2000.Beneficial effect of lung preservation is related toultrastructural integrity of tubular myelin after ex-perimental ischemia and reperfusion. Am. J. Respir.Crit. Care Med. 161, 2058�2065.

Floros, J., Steinbrink, R., Jacobs, K. et al., 1986. Isola-tion and characterization of cDNA clones for the35-kDa pulmonary surfactant-associated protein. J.Biol. Chem. 261, 9029�9033.

Ghannam, M.M., Mady, M.M., Khalil, W.A., 1999. In-

teraction of type-I collagen with phospholipidmonolayer. Biophys. Chem. 80, 31�40.

Goerke, J., 1998. Pulmonary surfactant: functions andmolecular composition. Biochim. Biophys. Acta 1408,79�89.

Grunder, R., Gehr, P., Bachofen, H., Schurch, S.,Siegenthaler, H., 1999. Structures of surfactant films:a scanning force microscopy study. Eur. Respir. J.14, 1290�1296.

Haagsman, H.P., Hawgood, S., Sargeant, T. et al., 1987.The major lung surfactant protein, SP 28-36, is acalcium-dependent, carbohydrate-binding protein. J.Biol. Chem. 262, 13877�13880.

Haagsman, H.P., Sargeant, T., Hauschka, P.V., Benson,B.J., Hawgood, S., 1990. Binding of calcium to SP-A,a surfactant-associated protein. Biochemistry 29,8894�8900.

Hakansson, K., Lim, N.K., Hoppe, H.J., Reid, K.B.,1999. Crystal structure of the trimeric alpha-helicalcoiled-coil and the three lectin domains of humanlung surfactant protein D. Structure Fold. Des. 7,255�264.

Hakansson, K., Reid, K.B., 2000. Collectin structure: areview. Protein Sci. 9, 1607�1617.

Hattori, A., Kuroki, Y., Katoh, T. et al., 1996. Surfac-tant protein A accumulating in the alveoli of patientswith pulmonary alveolar proteinosis: oligomericstructure and interaction with lipids. Am. J. Respir.Cell. Mol. Biol. 14, 608�619.

Hoppe, H.J., Reid, K.B., 1994. Collectins � solubleproteins containing collagenous regions and lectindomains � and their roles in innate immunity.Protein Sci. 3, 1143�1158.

Hoppe, H.J., Barlow, P.N., Reid, K.B., 1994. A parallelthree stranded alpha-helical bundle at the nucle-ation site of collagen triple-helix formation. FEBSLett. 344, 191�195.

Ikegami, M., Korfhagen, T.R., Bruno, M.D., Whitsett,J.A., Jobe, A.H., 1997. Surfactant metabolism insurfactant protein A-deficient mice. Am. J. Physiol.272, L479�485.

Ikegami, M., Jobe, A.H., 1998a. Surfactant proteinmetabolism in vivo. Biochim. Biophys. Acta 1408,218�225.

Ikegami, M., Korfhagen, T.R., Whitsett, J.A. et al.,1998b. Characteristics of surfactant from SP-A-defi-cient mice. Am. J. Physiol. 275, L247�254.

Ikegami, M., Jobe, A.H., Whitsett, J., Korfhagen, T.,2000. Tolerance of SP-A-deficient mice to hyperoxiaor exercise. J. Appl. Physiol. 89, 644�648.

King, R.J., Carmichael, M.C., Horowitz, P.M., 1983.Reassembly of lipid-protein complexes of pulmonarysurfactant. Proposed mechanism of interaction. J.Biol. Chem. 258, 10672�10680.

Korfhagen, T.R., Bruno, M.D., Ross, G.F. et al., 1996.Altered surfactant function and structure in SP-A


gene targeted mice. Proc. Natl. Acad. Sci USA 93,9594�9599.

Korfhagen, T.R., LeVine, A.M., Whitsett, J.A., 1998a.Ž .Surfactant protein A SP-A gene targeted mice.

Biochim. Biophys. Acta 1408, 296�302.Korfhagen, T.R., Sheftelyevich, V., Burhans, M.S. et

al., 1998b. Surfactant protein-D regulates surfactantphospholipid homeostasis in vivo. J. Biol. Chem. 273,28438�28443.

Kramer, R.Z., Bella, J., Mayville, P., Brodsky, B.,Berman, H.M., 1999. Sequence dependent confor-mational variations of collagen triple-helical struc-ture. Nat. Struct. Biol. 6, 454�457.

Kuroki, Y., Mason, R.J., Voelker, D.R., 1988. Pulmo-nary surfactant apoprotein A structure and modula-tion of surfactant secretion by rat alveolar type IIcells. J. Biol. Chem. 263, 3388�3394.

Kuroki, Y., Akino, T., 1991. Pulmonary surfactant pro-Ž .tein A SP-A specifically binds dipalmitoylphos-

phatidylcholine. J. Biol. Chem. 266, 3068�3073.Kuroki, Y., McCormack, F.X., Ogasawara, Y., Mason,

R.J., Voelker, D.R., 1994. Epitope mapping formonoclonal antibodies identifies functional domainsof pulmonary surfactant protein A that interact withlipids. J. Biol. Chem. 269, 29793�29800.

LeVine, A.M., Bruno, M.D., Huelsman, K.M., Ross,G.F., Whitsett, J.A., Korfhagen, T.R., 1997. Surfac-tant protein A-deficient mice are susceptible to groupB streptococcal infection. J. Immunol. 158,4336�4340.

LeVine, A.M., Gwozdz, J., Stark, J., Bruno, M., Whit-sett, J., Korfhagen, T., 1999. Surfactant protein-Aenhances respiratory syncytial virus clearance in vivo.J. Clin. Invest. 103, 1015�1021.

Liu, X., Kim, S., Dai, Q.H., Brodsky, B., Baum, J., 1998.Nuclear magnetic resonance shows asymmetric lossof triple helix in peptides modeling a collagen muta-tion in brittle bone disease. Biochemistry 37,15528�15533.

Lu, J., Wiedemann, H., Timpl, R., Reid, K.B., 1993.Similarity in structure between C1q and the col-lectins as judged by electron microscopy. Behring.Inst. Mitt. 93, 6�16.

McCormack, F.X., Calvert, H.M., Watson, P.A., Smith,D.L., Mason, R.J., Voelker, D.R., 1994a. The struc-ture and function of surfactant protein A. Hydroxy-proline- and carbohydrate-deficient mutant proteins.J. Biol. Chem. 269, 5833�5841.

McCormack, F.X., Kuroki, Y., Stewart, J.J., Mason,R.J., Voelker, D.R., 1994b. Surfactant protein Aamino acids Glu195 and Arg197 are essential forreceptor binding, phospholipid aggregation, regula-tion of secretion, and the facilitated uptake of phos-pholipid by type II cells. J. Biol. Chem. 269,29801�29807.

McCormack, F.X., Festa, A.L., Andrews, R.P., Linke,M., Walzer, P.D., 1997a. The carbohydrate recogni-tion domain of surfactant protein A mediates bind-ing to the major surface glycoprotein of Pneumocys-tis carinii. Biochemistry 36, 8092�8099.

McCormack, F.X., Stewart, J., Voelker, D.R., Da-modarasamy, M., 1997b. Alanine mutagenesis of sur-factant protein A reveals that lipid binding andpH-dependent liposome aggregation are mediatedby the carbohydrate recognition domain. Biochem-istry 36, 13963�13971.

McCormack, F.X., Pattanajitvilai, S., Stewart, J., Poss-mayer, F., Inchley, K., Voelker, D.R., 1997c. TheCys6 intermolecular disulfide bond and thecollagen-like region of rat SP-A play critical roles ininteractions with alveolar type II cells and surfactantlipids. J. Biol. Chem. 272, 27971�27979.

McCormack, F.X., 1998. Structure, processing andproperties of surfactant protein A. Biochim. Bio-phys. Acta 1408, 109�131.

McCormack, F.X., Damodarasamy, M., Elhalwagi, B.M.,1999. Deletion mapping of N-terminal domains ofsurfactant protein A. The N-terminal segment isrequired for phospholipid aggregation and specificinhibition of surfactant secretion. J. Biol. Chem. 274,3173�3181.

McCormack, F.X., Ikegami, M., LeVine, A.M., Korfha-gen, T.R., Whitsett, J.A., Dienger, K., 2000. In vivostructure�function analyses of surfactant protein AŽ .Abstract . Am. J. Respir. Crit. Care Med. 161, A42.

McLean, L.R., Lewis, J.E., Hagaman, K.A., Owen, T.J.,Jackson, R.L., 1993. Amphipathic alpha-helical pep-tides based on surfactant apoprotein SP-A. Biochim.Biophys. Acta 1166, 31�38.

Meyboom, A., Maretzki, D., Stevens, P.A., Hofmann,K.P., 1997. Reversible calcium-dependent interac-tion of liposomes with pulmonary surfactant proteinA. Analysis by resonant mirror technique and near-infrared light scattering. J. Biol. Chem. 272,14600�14605.

Meyboom, A., Maretzki, D., Stevens, P.A., Hofmann,K.P., 1999. Interaction of pulmonary surfactant pro-tein A with phospholipid liposomes: a kinetic studyon head group and fatty acid specificity. Biochim.Biophys. Acta 1441, 23�35.

Meyer, K.C., Sharma, A., Brown, R. et al., 2000. Func-tion and composition of pulmonary surfactant andsurfactant-derived fatty acid profiles are altered inyoung adults with cystic fibrosis. Chest 118, 164�174.

Nag, K., Perez-Gil, J., Ruano, M.L. et al., 1998. Phasetransitions in films of lung surfactant at the air�waterinterface. Biophys. J. 74, 2983�2995.

Nag, K., Frederick, L., Hearn, S.A., Veldhuizen, R.A.,Petersen, N.O., Possmayer, F., 2000. Role of phos-phorylcholine binding proteins SP-A and CRP on


pulmonary surfactant function studied using captiveŽ .bubble tensiometry Abstract . Am. J. Respir. Crit.

Care Med. 161, A660.Ogasawara, Y., McCormack, F.X., Mason, R.J.,

Voelker, D.R., 1994. Chimeras of surfactant proteinsA and D identify the carbohydrate recognition do-mains as essential for phospholipid interaction. J.Biol. Chem. 269, 29785�29792.

Palaniyar, N., Ridsdale, R.A., Holterman, C.E., Inchley,K., Possmayer, F., Harauz, G., 1998a. Structuralchanges of surfactant protein A induced by cationsreorient the protein on lipid bilayers. J. Struct. Biol.122, 297�310.

Palaniyar, N., Ridsdale, R.A., Possmayer, F., Harauz,Ž .G., 1998b. Surfactant protein A SP-A forms a novel

supraquaternary structure in the form of fibers.Biochem. Biophys. Res. Commun. 250, 131�136.

Palaniyar, N., Ridsdale, R.A., Hearn, S.A. et al., 1999a.Filaments of surfactant protein A specifically inter-act with corrugated surfaces of phospholipid mem-branes. Am. J. Physiol. 276, L631�641.

Palaniyar, N., Ridsdale, R.A., Hearn, S.A., Possmayer,F., Harauz, G., 1999b. Formation of membrane lat-tice structures and their specific interactions withsurfactant protein A. Am. J. Physiol. 276, L642�649.

Palaniyar, N., McCormack, F., Possmayer, F., Harauz,Ž .G., 1999c. Assembly of surfactant protein a SP-A

trimers and their interaction with lipid monolayersŽ .and bilayers Abstract . Am. J. Respir. Crit. Care

Med. 158, A729.Palaniyar, N., McCormack, F.X., Possmayer, F.,

Harauz, G., 2000a. Three-dimensional structure ofrat surfactant protein A trimers in associationwith phospholipid monolayers. Biochemistry 39,6310�6316.

Palaniyar, N., Korfhagen, T.R., Whitsett, J.A., Mcor-mack, F.X., 2000b. The collagen-like domain of sur-factant protein A is essential for the formation of

Ž .tubular myelin in vitro and in vivo Abstract . Am. J.Respir. Crit. Care Med. 161, A657.

Palaniyar, N., Korfhagen, T.R., Whitsett, J.A., Mcor-mack, F.X., 2000c. The N-terminal segment of SP-Dcan functionally replace the N-terminal segment of

Ž .SP-A for tubular myelin formation in vivo Abstract .Am. J. Respir. Crit. Care Med. 161, A42.

Phelps, D.S., Floros, J., 1988. Proline hydroxylationalters the electrophoretic mobility of pulmonary sur-factant-associated protein A. Electrophoresis 9,231�233.

Postle, A.D., Mander, A., Reid, K.B., Wang, J.Y.,Wright, S.M., Moustaki, M. et al., 1999. Deficienthydrophilic lung surfactant proteins A and D withnormal surfactant phospholipid molecular species incystic fibrosis. Am. J. Respir. Cell. Mol. Biol. 20,90�98.

Poulain, F.R., Allen, L., Williams, M.C., Hamilton,R.L., Hawgood, S., 1992. Effects of surfactantapolipoproteins on liposome structure: implicationsfor tubular myelin formation. Am. J. Physiol. 262,L730�739.

Poulain, F.R., Akiyama, J., Allen, L. et al., 1999. Ultra-structure of phospholipid mixtures reconstituted withsurfactant proteins B and D. Am. J. Respir. Cell.Mol. Biol. 20, 1049�1058.

Ramet, M., Haataja, R., Marttila, R., Floros, J., Hall-man, M., 2000. Association between the surfactant

Ž .protein A SP-A gene locus and respiratory-distresssyndrome in the Finnish population. Am. J. Hum.Genet. 66, 1569�1579.

Resnick, D., Chatterton, J.E., Schwartz, K., Slayter, H.,Krieger, M., 1996. Structures of class A macrophagescavenger receptors. Electron microscopic study offlexible, multidomain, fibrous proteins and determi-nation of the disulfide bond pattern of the scavengerreceptor cysteine-rich domain. J. Biol. Chem. 271,26924�26930.

Ridsdale, R.A., Palaniyar, N., Holterman, C.E., Inchley,K., Possmayer, F., Harauz, G., 1999. Cation-media-ted conformational variants of surfactant protein A.Biochim. Biophys. Acta 1453, 23�34.

Rishi, A., Hatzis, D., McAlmon, K., Floros, J., 1992. Anallelic variant of the 6A gene for human surfactantprotein A. Am. J. Physiol. 262, L566�573.

Ross, G.F., Notter, R.H., Meuth, J., Whitsett, J.A.,1986. Phospholipid binding and biophysical activity

Ž .of pulmonary surfactant-associated protein SAP -35and its non-collagenous COOH-terminal domains. J.Biol. Chem. 261, 14283�14291.

Ruano, M.L., Miguel, E., Perez-Gil, J., Casals, C., 1996.Comparison of lipid aggregation and self-aggrega-tion activities of pulmonary surfactant-associatedprotein A. Biochem. J. 313, 683�689.

Ruano, M.L., Perez-Gil, J., Casals, C., 1998a. Effect ofacidic pH on the structure and lipid binding proper-ties of porcine surfactant protein A. Potential role ofacidification along its exocytic pathway. J. Biol.Chem. 273, 15183�15191.

Ruano, M.L., Nag, K., Worthman, L.A., Casals, C.,Perez-Gil, J., Keough, K.M., 1998b. Differential par-titioning of pulmonary surfactant protein SP-A intoregions of monolayers of dipalmitoylphosphatidyl-choline and dipalmitoylphosphatidylcholine�di-palmitoylphosphatidylglycerol. Biophys. J. 74,1101�1109.

Ruano, M.L., Garcia-Verdugo, I., Miguel, E., Perez-Gil,J., Casals, C., 2000. Self-aggregation of surfactantprotein A. Biochemistry 39, 6529�6537.

Saitoh, H., Okayama, H., Shimura, S., Fushimi, T.,Masuda, T., Shirato, K., 1998. Surfactant protein A2gene expression by human airway submucosal glandcells. Am. J. Respir. Cell. Mol. Biol. 19, 202�209.


Saitoh, M., Sano, H., Chiba, H. et al., 2000. Importanceof the carboxy-terminal 25 amino acid residues oflung collectins in interactions with lipids and alve-olar type II cells. Biochemistry 39, 1059�1066.

Sano, H., Kuroki, Y., Honma, T. et al., 1998. Analysisof chimeric proteins identifies the regions in thecarbohydrate recognition domains of rat lung col-lectins that are essential for interactions with phos-pholipids, glycolipids, and alveolar type II cells. J.Biol. Chem. 273, 4783�4789.

Savov, J., Wright, J.R., Young, S.L., 2000. Incorpora-tion of biotinylated SP-A into rat lung surfactantlayer, type II cells, and Clara cells. Am. J. Physiol.Lung Cell. Mol. Physiol. 279, L118�L126.

Schurch, S., Green, F.H., Bachofen, H., 1998. Forma-tion and structure of surface films: captive bubblesurfactometry. Biochim. Biophys. Acta 1408,180�202.

Seidner, S.R., Ikegami, M., Yamada, T., Rider, E.D.,Castro, R., Jobe, A.H., 1995. Decreased surfactantdose-response after delayed administration topreterm rabbits. Am. J. Respir. Crit. Care Med. 152,113�120.

Sheriff, S., Chang, C.Y., Ezekowitz, R.A., 1994. Humanmannose-binding protein carbohydrate recognitiondomain trimerizes through a triple alpha-helical

�coiled-coil published erratum appears in Nat. Struct.Ž . �Biol. 1996 3, 103 . Nat. Struct. Biol. 1, 789�794.

Shu, W., Liu, J., Ji, H., Lu, M., 2000. Core structure ofthe outer membrane lipoprotein from Escherichiacoli at 1.9 A resolution. J. Mol. Biol. 299, 1101�1112.

Sohma, H., Hattori, A., Kuroki, Y., Akino, T., 1993.Calcium and dithiothreitol dependent conformatio-nal changes in beta-sheet structure of collagenaseresistant fragment of human surfactant protein A.Biochem. Mol. Biol. Int. 30, 329�336.

Spissinger, T., Schafer, K.P., Voss, T., 1991. Assemblyof the surfactant protein SP-A. Deletions in theglobular domain interfere with the correct folding ofthe molecule. Eur. J. Biochem. 199, 65�71.

Suzuki, Y., Fujita, Y., Kogishi, K., 1989. Reconstitutionof tubular myelin from synthetic lipids and proteinsassociated with pig pulmonary surfactant. Am. Rev.Respir. Dis. 140, 75�81.

Takemura, T., Fukuda, Y., Harrison, M., Ferrans, V.J.,1987. Ultrastructural, histochemical, and freeze-frac-ture evaluation of multilamellated structures in hu-man pulmonary alveolar proteinosis. Am. J. Anat.179, 258�268.

Tsunezawa, W., Sano, H., Sohma, H., McCormack,F.X., Voelker, D.R., Kuroki, Y., 1998. Site-directedmutagenesis of surfactant protein A reveals dissocia-tion of lipid aggregation and lipid uptake by alveolartype II cells. Biochim. Biophys. Acta 1387, 433�446.

Ueda, T., Ikegami, M., Jobe, A., 1994. Surfactant sub-

types. In vitro conversion, in vivo function, and ef-fects of serum proteins. Am. J. Respir. Crit. CareMed. 149, 1254�1259.

van Eijk, M., Haagsman, H.P., Skinner, T. et al., 2000.Porcine lung surfactant protein D: complementaryDNA cloning, chromosomal localization, and tissue

�distribution published erratum appears in J. Im-Ž . �munol. 2000 164, 3444 . J. Immunol. 164,

1442�1450.Veldhuizen, R.A., Yao, L.J., Hearn, S.A., Possmayer,

F., Lewis, J.F., 1996. Surfactant-associated protein Ais important for maintaining surfactant large-aggre-gate forms during surface-area cycling. Biochem. J.313, 835�840.

Veldhuizen, R., Nag, K., Orgeig, S., Possmayer, F.,1998. The role of lipids in pulmonary surfactant.Biochim. Biophys. Acta 1408, 90�108.

Voorhout, W.F., Veenendaal, T., Haagsman, H.P.,Verkleij, A.J., van Golde, L.M., Geuze, H.J., 1991.Surfactant protein A is localized at the corners ofthe pulmonary tubular myelin lattice. J. Histochem.Cytochem. 39, 1331�1336.

Voss, T., Eistetter, H., Schafer, K.P., Engel, J., 1988.Macromolecular organization of natural and recom-binant lung surfactant protein SP 28-36. Structuralhomology with the complement factor C1q. J. Mol.Biol. 201, 219�227.

Wallis, R., Drickamer, K., 1997. Asymmetry adjacent tothe collagen-like domain in rat liver mannose-bind-ing protein. Biochem. J. 325, 391�400.

Wallis, R., Drickamer, K., 1999. Molecular determi-nants of oligomer formation and complement fixa-tion in mannose-binding proteins. J. Biol. Chem.274, 3580�3589.

Weaver, T.E., Beck, D.C., 1999. Use of knockout miceto study surfactant protein structure and function.

Ž .Biol. Neonate 76 Suppl. 1 , 15�18.Weis, W.I., Drickamer, K., Hendrickson, W.A., 1992.

Structure of a C-type mannose-binding protein com-plexed with an oligosaccharide. Nature 360, 127�134.

Weis, W.I., Drickamer, K., 1994. Trimeric structure ofa C-type mannose-binding protein. Structure 2,1227�1240.

Williams, M.C., 1977. Conversion of lamellar bodymembranes into tubular myelin in alveoli of fetal ratlungs. J. Cell. Biol. 72, 260�277.

Williams, M.C., 1978. Freeze-fracture studies of tubu-lar myelin and lamellar bodies in fetal and adult ratlungs. J. Ultrastruct. Res. 64, 352�361.

Williams, M.C., 1982. Ultrastructure of tubular myelinand lamellar bodies in fast-frozen adult rat lung.Exp. Lung Res. 4, 37�46.

Williams, M.C., Hawgood, S., Hamilton, R.L., 1991.Changes in lipid structure produced by surfactantproteins SP-A, SP-B, and SP-C. Am. J. Respir. Cell.Mol. Biol. 5, 41�50.


Yu, S.H., Possmayer, F., 1998. Interaction of pulmo-nary surfactant protein A with dipalmitoylphos-phatidylcholine and cholesterol at the air�water in-terface. J. Lipid Res. 39, 555�568.

Yu, S.H., McCormack, F.X., Voelker, D.R., Possmayer,F., 1999. Interactions of pulmonary surfactant pro-tein SP-A with monolayers of dipalmitoylphos-

phatidylcholine and cholesterol. Roles of SP-A do-mains. J. Lipid Res. 40, 920�929.

Yukitake, K., Brown, C.L., Schlueter, M.A., Clements,J.A., Hawgood, S., 1995. Surfactant apoprotein Amodifies the inhibitory effect of plasma proteins onsurfactant activity in vivo. Pediatr. Res. 37, 21�25.

Documents

Domains of surfactant protein A that affect protein oligomerization, lipid structure and surface tension