- Email: [email protected]
Cation-mediated conformational variants of surfactant protein A
Biochimica et Biophysica Acta 1453 (1999) 23^34
Cation-mediated conformational variants of surfactant protein A Ross A. Ridsdale a , Nades Palaniyar a , Chet E. Holterman a , Kevin Inchley b , Fred Possmayer b , George Harauz a; * a b
Department of Molecular Biology and Genetics, The University of Guelph, 50 Stone Road East, Guelph, Ont. N1G 2W1, Canada Departments of Biochemistry, and Obstetrics and Gynaecology, The University of Western Ontario, London, Ont. N6A 5A5, Canada Received 15 June 1998; accepted 10 August 1998
Abstract Surfactant protein A (SP-A) is the major protein of pulmonary surfactant. This protein is implicated in regulating surfactant secretion, alveolar processing, recycling, and in non-serum-induced immune response. An increasing body of work indicates the importance of cations, particularly calcium, on SP-A function. However, little information exists on the effects of cations on SP-A quaternary structure. Here, the quaternary organisation of bovine surfactant protein A in the presence of cations has been quantitatively and systematically studied by transmission electron microscopy. The conformation of SP-A is altered by the presence of cations, especially calcium, then sodium, and to a small extent, magnesium. There is a transition concentration, unique for each cation, at which a conformational switch occurs. These transition concentrations are: 5 mM for CaCl2 , 100 mM for NaCl and 1 mM for MgCl2 . Below these concentrations, SP-A exists primarily in an opened form with a large head diameter of 20 nm; above it, SP-A is mostly in a closed form due to a compaction of the headgroups resulting in a head diameter of 11 nm. There is a corresponding increase in particle length from 17 nm for opened SP-A to 20 nm for closed SP-A. The fact that the transition concentrations are within physiological range suggests that cation-mediated conformational changes of SP-A could be operative in vivo. ß 1999 Elsevier Science B.V. All rights reserved. Keywords: Pulmonary surfactant; Surfactant protein A; Surfactant protein A structure; Calcium; Cation
1. Introduction Pulmonary surfactant is a complex mixture of phospholipids and proteins that forms a tightly packed lipid monolayer at the air/liquid interface in the lungs. This surfactant works to reduce surface tension which is essential to proper alveolar gas exAbbreviations: CRD, carbohydrate recognition domain; SPA, -B, -C, -D, surfactant protein A, B, C, D; TEM, transmission electron microscope/microscopy; DPPC, dipalmitoylphosphatidylcholine * Corresponding author. Fax: +1-519-837-2075; E-mail: [email protected]
change. Without surfactant the alveoli would tend to collapse during exhalation as a result of intermolecular forces [1]. Surfactant lipids and proteins alike are produced by alveolar type II cells. Some of these components are released as lamellar bodies that are then used to generate the highly structured proteolipid complex known as tubular myelin [2]. Tubular myelin as well as other surfactant aggregates are thought to be storage intermediates that contribute surfactant material to the monolayer [3]. Through a poorly understood process, surfactant is recycled through the type II cells or degraded in alveolar macrophages [4,5]. The most common pulmonary surfactant protein
0925-4439 / 99 / $ ^ see front matter ß 1999 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 5 - 4 4 3 9 ( 9 8 ) 0 0 0 5 7 - X
BBADIS 61761 23-12-98
24
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
is SP-A, a member of the collectin family, which itself is part of the C-type lectin family. Other collectins are mannose-binding protein, complement factor C1q, and another surfactant protein, SP-D. In reduced form, SP-A is 26 to 38 kDa in molecular mass. It has been proposed that there are four structural domains in SP-A (Fig. 1) [6^9]. These domains are: (i) a 7-amino-acid amino-terminus, (ii) a tandemly repeating Gly^X^Y collagen-like region, (iii) an amphipathic `neck' region, and ¢nally (iv) a carboxy-terminal region that is globular and highly similar to other carbohydrate recognition domains. Posttranslationally, SP-A is modi¢ed by cleavage of its signal peptide, hydroxylation of prolines within the collagen-like domain, and asparagine-linked glycosylation [10,11]. Trimeric assembly in SP-A is the result of association through the globular carboxy-terminal domains and the collagen-like regions. Six trimers associate further, among their collagen-like regions, to generate a 700 kDa octadecamer. This gives the protein a characteristic `bouquet of £owers' appearance [12,13] (Fig. 1). Both trimers and octadecamers are postulated to maintain quaternary structure through disulphide bonding [6]. The overall assembly is similar to that of C1q [13,14]. SP-A has a wide variety of potential ligands. It has the capacity to associate lipids, with particular af¢nity for the most abundant and signi¢cant surfactant phospholipid, dipalmitoylphosphatidylcholine (DPPC), in a calcium-dependent manner [15^18].
Earlier studies suggested that this association occurs via the hydrophobic neck [19]; however, there is an accumulation of evidence that identi¢es the carbohydrate recognition domain (CRD) of the headgroup to be the primary region for lipid interaction [20^23]. SP-A can promote aggregation of lipid vesicles in a calcium-dependent manner [24,25], and also self-aggregates in a calcium-in£uenced manner [26]. SP-A
C
Fig. 1. (a) Structural features of the SP-A monomer, trimer and octadecamer. The monomer has a short amino-terminal region, a large collagen-like region (here indicated with a wavy line) which has a kink in it, an amphipathic neck region, and a carbohydrate recognition domain (CRD). In the trimeric form of SP-A (i.e., the basic trimer unit), three CRDs are in close proximity and form a single globular moiety referred to as a headgroup. Six trimers constitute a single SP-A octadecamer. The region from the amino-terminus to the kink forms the octadecamer stem. The distance from the kink to the CRD is referred to as the octadecamer arm length. The region containing the six headgroups is referred to simply as the head. (b^f) When imaged by TEM, SP-A lies in two orientations on the carbon support ¢lm. (b) Pictorial representation and (c) TEM image of SP-A side view. (d) Pictorial representation and (e) TEM image of SP-A top-down view. (f) Pictorial representation of SP-A side view with nomenclature of measured regions.
BBADIS 61761 23-12-98
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
displays xenophilic characteristics by binding to pulmonary microorganisms, lipid A of endotoxin, and pollen grains [27^35]. Many of these associations are attributed to the CRDs and/or to the oligosaccharides that were post-translationally added to SP-A [36]. The collagen-like region of SP-A acts as a ligand to at least one high-a¤nity receptor on macrophages, and can cause chemotaxis and release of anti-microbial agents [37^39]. SP-A is localised both in the distal and conductive respiratory tracts, and may be present in the large and small gastric intestinal tracts [40]. SP-A has been implicated in diverse physiological functions, such as tubular myelin formation, regulating extracellular surfactant concentration by a¡ecting secretion and readsorption, and functioning with SP-B in surfactant adsorption to and spreading at the liquid/air interface [6]. There is an increasing body of evidence concerning SP-A, along with SP-D, acting in host defence because of their xenophilic characteristics and e¡ects on macrophages. For a protein with such a variety of roles, it is surprising that the SPA knock-out mouse shows a viable phenotype [41]. These SP-A knock-out mice lack tubular myelin, are more susceptible to group B Streptococcus [42], but have essentially normal saturated phosphatidylcholine pool sizes [43]. The study of SP-A structural characteristics will improve our understanding of how it performs such a variety of tasks in such di¡erent environments. Our examination here by transmission electron microscopy (TEM) of SP-A in the presence of di¡ering concentrations of mono- and divalent salts indicates that there are two, previously unidenti¢ed, major conformational forms of SP-A. We shall refer to these as closed or compacted SP-A, and opened SP-A. We have characterised the structural anatomy of these in a variety of cationic environments. 2. Materials and methods 2.1. Protein puri¢cation Bovine surfactant was obtained by lung lavage and SP-A puri¢ed as previously described [44]. This protocol involved extracting the surfactant lipids with butanol, and washing with ether/ethanol and ether.
25
The hydrophilic precipitate was dried and resuspended in 5 mM Hepes^NaOH (pH 7.4), 0.1 mM EDTA, and 1.0 mM CaCl2 . This solution was then passed through an immobilised D-mannose column (Pierce Chemical, Rockford, IL, USA). After the column was washed, the SP-A was eluted using 2.0 mM EDTA in 5 mM Hepes^NaOH (pH 7.4). Sample was stored at 4³C until use in elution bu¡er at a concentration between 100 and 250 Wg/ml. 2.2. Protein preparation for TEM A 5-mM Hepes solution was prepared with no salt, 1 M NaCl, 1 M MgCl2 , or 1 M CaCl2 . The pH of each solution was then adjusted to 7.4 using 1 M NaOH. Each bu¡er required approximately the same concentration of NaOH, and so the ¢nal concentration of sodium and hydroxyl ions in each ¢nal reaction bu¡er was negligible (V2.5 mM). Combinations of these basic bu¡ers were made to produce individual dilution series of NaCl (10, 40, 80, 100, 250, 500 and 1000 mM), MgCl2 and CaCl2 (each of 0.01, 0.1, 1, 5, 10 and 20 mM). In a te£on microwell, 12.5 Wl of bu¡er was combined with 0.5 Wl of 250 Wg/ml SP-A to give a ¢nal protein concentration of approximately 10 Wg/ml. Samples were kept humid in a chamber containing the same bu¡er in which the protein was incubated, and left for 24 h at 32³C to achieve conformational equilibrium. 2.3. TEM analysis SP-A samples from the microwells were spread onto 400-mesh Cu grids that were coated by carbon-stabilised holey plastic (cellulose acetate butyrate), and again covered with thin carbon ¢lm of approximately 5 nm thickness. The samples were negatively stained using 2% uranyl acetate and examined using a JEOL JEM-100CX TEM. Images were obtained at a variety of magni¢cations from 66 000U to 130 000U on 80U100-mm plate ¢lm (Kodak Estar 4489). Measurements of dimensions of SP-A octadecamers were made using a digital capture CCD camera (Princeton MicroMax 1400, EMPIX Imaging, Mississauga, Canada) and analysis software (Alpha Innotech, San Leandro, CA, USA). All dimensions henceforth reported will be rounded o¡ to the î ). Simple statistics (mean, standnearest 0.1 nm (1 A
BBADIS 61761 23-12-98
26
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
ard error of the mean, and signi¢cance values for both one- and two-tailed t-tests) were calculated using Excel (Microsoft O¤ce) software. Distribution statistics shall be presented as mean þ standard error of the mean (n = number of measurements). A con¢dence level of 95% (i.e., P 6 0.05) was considered to represent statistical signi¢cance. 3. Results 3.1. TEM of bovine SP-A It was observed in this and in other studies that SP-A portrayed two orientations when examined by TEM [13,45]. The ¢rst view was of a top-down perspective (Figs. 1 and 2). Here, the trimers of globular CRDs, referred to as trimeric headgroups, appeared as individual globules with six such entities being arranged in a wheel formation. The second orientation was a side view in which a stem was visible from the pro¢le and the headgroups appeared at one end of that stem. Side views were more distinctive than top-down views because more of the particle mass was directly apparent in pro¢le. Top-down views appeared to be the major SP-A orientation (approximately 70%) under lower salt conditions; however, they were readily identi¢able only in locations where the stain concentration was relatively dense. For this reason, we focused in this study on the side views, and on the relative changes in dimensions (Fig. 1) and frequencies of occurrence under di¡ering salt conditions. 3.2. Identi¢cation of conformational variants A histogram of head diameter measurements for SP-A under all of these conditions displayed distinct bimodality (Fig. 3). The peak for the ¢rst population (`narrow-head', `closed-head', or `compacted SP-A') was at 11 nm and the peak for the second population (`wide-head', `opened-head', or `opened SP-A') was at 21 nm. The ¢rst (narrow-head) population had a tight distribution with 50% of its population between 9 nm and 11 nm (i.e., a 2-nm range). The second (wide-head) population had 50% of its population spread from 19 to 23 nm (i.e., a 4-nm range), indicating that there were intermediates that covered a
Fig. 2. Appearance of SP-A via TEM of negatively stained preparations. (a) A gallery of SP-A side view particles from various conditions. (b) Image of a typical medium-sized SP-A self-aggregate in 10 mM CaCl2 . (c) A ¢eld of view of a sample of SP-A in 0.5 mM CaCl2 . Very few stems are visible, one indicated by a circle. This image is typical for SP-A samples below the transition concentration. (d) A ¢eld of view of a sample of SP-A in 5 mM CaCl2 , showing a large number of visible stems, some indicated by circles. This view is typical for SP-A samples above the transition concentration. Scale bar = 20 nm.
broad range of head diameter sizes, yet with a peak value at 21 nm which agreed with independent studies [13,45]. Each population type was interpreted to represent a conformational variant of SP-A. The variation in head diameter sizes represented the result of both the range of head movements relative to one another,
BBADIS 61761 23-12-98
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
27
Fig. 3. Histogram of head diameters measured for SP-A under all salt conditions. Two distinct populations are observed. The curves are an interpretation of the trends for each population, and are not intended to represent ¢ts of Gaussian functions, for example.
and the fundamental error in measurement itself. The narrow-head population showed a sharp increase in the histogram from 7 to 8 nm, with very few smaller measurements. This lower limit was attributable to the ¢nite compactibility of the heads. That is to say, the heads can only come to within 8 nm from one another, and lower values (e.g., 7 nm or less) must have been the result of intrinsic errors in measurement. Therefore, these were estimated to be less than 1 nm in 8 nm, or roughly þ 15%. Both narrow- and wide-head diameter conformations appeared to be visible under di¡ering salt conditions. However, there was a shift towards narrowhead SP-A as the corresponding salt concentration was increased. There was a signi¢cant increase in the number of visible stems at 100 mM NaCl, 5 mM CaCl2 and 1 mM MgCl2 (Fig. 4). Clearly, the side view became more favoured after a critical salt condition was achieved, which we henceforth refer to as the transition concentration. (Strictly speaking, the transition salt concentrations lie within ranges of 80^100 mM NaCl, 1.0^5.0 mM CaCl2 , and 0.1^1.0 mM MgCl2 . Here, we shall use the upper value as a conservative and convenient convention.) Furthermore, the total particle length shifted from 16.6 þ 0.4 nm (n = 44) to 20.2 þ 0.6 nm (n = 83)
when the salt concentration rose above the transition concentration (Fig. 5). Such a rise in overall particle length would be expected if the head compacted; then the arm length would have contributed more to particle length. The shift towards narrow-head diameters under higher salt conditions was particularly pronounced with NaCl and CaCl2 , compared with MgCl2 . Since in these experiments there was 20-times the concentration of sodium relative to calcium, cation-mediated conformational change of SP-A was more sensitive to calcium. To determine the rate of these cationic e¡ects, measurements of numbers of visible stems over a 24 h period were performed at either 5 mM CaCl2 , or 100 mM NaCl, i.e., concentrations above the transition ones. There was a steady increase in the number of visible stems over 2 h, when a maximum was achieved (data not shown). Afterwards, there was a slight and slow decrease in the number of stems, reaching 60% of the peak value at 24 h incubation time. 3.3. Synergistic e¡ects of combinations of cations Combinations of CaCl2 , NaCl, and MgCl2 , each below the transition concentration (1 mM, 80 mM,
BBADIS 61761 23-12-98
28
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
BBADIS 61761 23-12-98
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
29
Fig. 4. Histogram of number of visible stems per 100 Wm2 with respect to cation concentration. TEM was performed on randomly selected regions of the grid, and areas of size 10U10 Wm were selected randomly from the image plates. At a particular concentration indicated by arrows, the number of visible stems rises sharply. This transition concentration is di¡erent for each cation. Error bars represent standard deviations, and the number of grid areas examined is between 5 and 7 for each class.
6
Fig. 5. Histograms of particle lengths (a) above and (b) below the transition concentration. The data have been pooled for all three salts studied. There is a distinct shift in the population towards longer particles above the transition concentration. The hypothesis that the two distributions are di¡erent is statistically signi¢cant at a con¢dence level of 95%.
BBADIS 61761 23-12-98
30
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
Fig. 6. In£uence of combinations of cations on the number of SP-A particles visible in pro¢le. A combination of two or more cations below the transition concentration results in an increase in the number of side views. However, the e¡ect is not as distinct above the transition concentration. Error bars represent one standard deviation, and the number of grid areas examined is between 5 and 7 for each class.
0.1 mM, respectively), were tested with SP-A (Fig. 6). The combination of any two or all three cations in solution with SP-A produced only a minor shift in equilibrium towards compacted SP-A, compared to any cation alone. The strong cationic e¡ect noted after the transition concentration (i.e., large number of visible stems) was not induced by any combination of salts below the transition concentration. That is to say, under these experimental conditions (cf. Ref. [45]), the e¡ect did not appear to be fully additive or su¤cient to cause the degree of change as observed after transition. The number of side views observed when exposed to a combination of ions remained consistently larger than for any ion on its own. The combination of all three salts resulted in the highest number of compacted SP-As, compared to any combination of two cations. The relatively small increase in side views with combined ions could be attributed to the relatively small e¡ect of increased ionic strength. 3.4. Observation of individual basic trimer units Although the SP-A complexes were found primarily as octadecamers, a small proportion was seen to be basic trimer units (Fig. 7). In the presence of EDTA or cations after short incubation periods (up to about 4 h), these trimers showed a kink in the
stem region. However, in the presence of 5 mM calcium after extended incubation periods, most of the basic trimer units did not contain the kink. This observation suggested that the cations could alter the SP-A structure in the collagen-like region, albeit slowly under the experimental conditions tested. 4. Discussion 4.1. Cationic e¡ects on SP-A conformation The ¢rst e¡ect of cations was to cause the population of SP-A complexes to shift to a primarily compacted conformational variant. There was a critical salt transition concentration for each cation of 5 mM CaCl2 , 100 mM NaCl, and 1 mM MgCl2 . Below this concentration, SP-A conformation was prevalently wide-head diameter and short-particle length. Above this concentration, the narrow-head and long-particle conformation was predominant, and large numbers of side-view particles appeared (Figs. 2^5). Two sets of distributions were formed and compared, the ¢rst being the particle lengths (Fig. 5), and the second being the ratio of particle head diameters to their lengths (data not shown). Both sets of distributions were di¡erent above and below the transition salt concentrations, statistically signi¢cant at a con¢-
BBADIS 61761 23-12-98
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
31
Fig. 7. Nature of basic trimer units of SP-A found under various conditions. Electron micrographs showing clearly identi¢able SP-A trimers were found in various experiments, although they were rarely seen compared to the octadecamers. (a) Long incubation period (24 h) in the presence of 5 mM CaCl2 . (b) Medium incubation period (1^4 h) in the presence of 5 mM CaCl2 . (c) Short incubation period (0^30 min) in the presence of 5 mM CaCl2 . (d) Long incubation period (24 h) in the presence of 5 mM EDTA. Scale bar = 20 nm.
dence level of 95%. The transition concentration occurred at di¡ering concentrations of Cl3 (i.e., 100 mM in NaCl, 2U5 = 10 mM in CaCl2 , and 2U1 = 2 mM in MgCl2 ) and in solutions of di¡ering ionic strength (i.e., the CaCl2 solution was 15% the ionic strength of the NaCl one, and MgCl2 was 3% the ionic strength of NaCl). Therefore, anions were not the cause of the conformational change and ionic strength could have played only a minor role in compacted SP-A octadecamer formation. The head diameter of SP-A could narrow considerably because headgroups were brought closer in
proximity (Fig. 3). This gross conformational change in the SP-A arm region caused a corresponding increase in particle length (Figs. 4 and 5). The close packing was attributed to a straightening at the junction between the SP-A stem and arm, in the region known as the kink (Figs. 1 and 7). In comparison, class A macrophage scavenger receptor is structurally similar to the SP-A basic trimer (i.e., two ¢brous domains with a hinge, and a trimeric globular headgroup with multiple binding a¤nities), and has recently been shown by TEM to pivot at the hinged region [46]. This structural alteration of the class A
BBADIS 61761 23-12-98
32
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
macrophage scavenger receptor occurred without multiple trimer association, suggesting that SP-A basic trimers also may not need to be within an octadecamer for the kink angle to be altered. The measurements of the number of visible stems at various cation concentrations (Figs. 4 and 6) suggested that the change in SP-A quaternary structure led in turn to a change in preferred orientation on the underlying support ¢lm. An opened-head SP-A would be more frequently observed in top-down orientation, whereas a closed-head SP-A would prefer to lie on its side. Thus, we submit that the most probable explanation for the increase in number of visible stems with increased cation concentration is a physical one: there is an increased number of compacted-head SP-A complexes in solution, and these in turn `fall down' upon the carbon support more naturally on their sides and hence are more easily visualised. 4.2. Cationic e¡ects on SP-A self-association As the concentration of each ion increased beyond the transition concentration, the number of isolated particles per unit area decreased. We propose that the cause for the decrease was the result of SP-A self-aggregation (Fig. 2b) [26], even though we tried to minimise this phenomenon in these experiments by using a low initial protein concentration. Indeed, below the transition cation concentration, aggregates were rarely seen. This post-transitional decrease was most pronounced with CaCl2 (Fig. 5b). With MgCl2 the e¡ect on SP-A structure was much less, both in terms of generating compacted SP-A and of the posttransitional decrease (Fig. 4c). There was a considerable increase in compacted SP-A octadecamers between 100 mM to 250 mM NaCl (Fig. 4a). With the low initial protein concentration used here, the self-aggregation e¡ect did occur from 500 mM NaCl upwards, but was not as large as for calcium. Thus, the second e¡ect of cations was to enhance SP-A self-aggregation. The phenomenon of SP-A self-aggregation is important to consider when examining the literature. Previously, Ruano et al. [26] concluded that the half-maximal lipid aggregation mediated by calcium was in the micromolar range, whereas the SP-A selfaggregation was half-maximal at millimolar concen-
trations. In the latter case, the compacted SP-A conformation could functionally contribute to self-aggregation. Other studies involving calcium-dependent SP-A association with lipid have been conducted using 100 to 150 mM NaCl solutions; therefore, they appear to characterise the e¡ects of calcium on compacted SP-A rather than on opened SP-A (e.g., Refs. [47,48]). Future studies on the relative functions of compacted SP-A should be done at the higher concentrations of 250 mM NaCl, as these conditions generate the most compacted SP-A structure with the least self-aggregation, as well as at low concentrations (e.g., 10 mM NaCl) to compare with opened SP-A. The changes in the numbers of visible stems occurred over relatively long time periods. The 2-h period required to achieve maximal conformational change was much greater than the minutes that produce protein/protein and protein/lipid aggregation in vitro [7,21,23]. However, these aggregation experiments were performed in 150 mM NaCl, a concentration above the transition concentration. Under such conditions, compacted SP-A should have been present in large amounts. Therefore, the rapid aggregation that was observed under those conditions was a distinctly di¡erent phenomenon than the slow conformational changes observed here. According to the above studies and the present one, sodium concentrations of 100 to 150 mM sodium are not su¤cient to induce SP-A self-aggregation or vesicle aggregation, but are capable of inducing conformational change. What remains unknown is whether a particular conformation of SP-A contributes to the generation of these di¡erent forms of aggregation. 5. Conclusions The quaternary conformation of the SP-A octadecamer is signi¢cantly and speci¢cally a¡ected by cations. At a transition concentration that is unique for each cation, the population shifts, at a slow rate, from a primarily opened form to a primarily compacted conformation. The nature of the conformational change is suggested to be due to the straightening of the basic trimer units in the region of the kink. This conformational change in SP-A might have physiological signi¢cance, e.g., by altering SP-
BBADIS 61761 23-12-98
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
A interaction with its receptors on type II cells, or by altering the ability of SP-A to interact with microbial invaders and with itself. In all cases, the transition salt concentration is close to that under physiological conditions. Interestingly, the NaCl concentration in the bronchial airway is below the transition concentration (ranging from 40 mM [49] to 80 mM [50]), whereas the alveolar concentration is above it [51,52]. The variation of salt conditions in the lung may thus serve as a regulatory element de¢ning SP-A conformation in vivo. Acknowledgements This work was supported by the Ontario Thoracic Society (G.H.), the Natural Sciences and Engineering Research Council of Canada (G.H.), and the Medical Research Council of Canada (F.P.). References [1] R.E. Pattle, Nature 175 (1955) 1125^1126. [2] M.C. Williams, J. Cell Biol. 72 (1977) 260^277. [3] S. Hawgood, B.J. Benson, J. Schilling, D. Damm, J.A. Clements, R.T. White, Proc. Natl. Acad. Sci. U.S.A. 84 (1987) 66^70. [4] J.R. Wright, J.A. Clements, Am. Rev. Respir. Dis. 135 (1987) 426^444. [5] S.L. Young, J.R. Wright, J.A. Clements, J. Appl. Physiol. 66 (1989) 1336^1342. [6] F. McCormack, Chest 111 (1997) 114S^119S. [7] F.X. McCormack, H.M. Calvert, P.A. Watson, D.L. Smith, R.J. Mason, D.R. Voelker, J. Biol. Chem. 269 (1994) 5833^ 5841. [8] K. Sano, J. Fisher, R.J. Mason, Y. Kuroki, J. Schilling, B. Benson, D. Voelker, Biochem. Biophys. Res. Commun. 144 (1987) 367^374. [9] R.T. White, D. Damm, J. Miller, K. Spratt, J. Schilling, S. Hawgood, B. Benson, B. Cordell, Nature 317 (1985) 361^ 363. [10] J. Johansson, T. Curstedt, Eur. J. Biochem. 244 (1997) 675^ 693. [11] T.E. Weaver, J.A. Whitsett, Biochem. J. 273 (1991) 249^264. [12] R.J. King, D. Simon, P.M. Horowitz, Biochim. Biophys. Acta 1001 (1989) 294^301. [13] T. Voss, J. Eistetter, K.P. Schafer, J. Engel, J. Mol. Biol. 201 (1988) 219^227. [14] V.N. Schumaker, P.H. Poon, G.W. Seegan, C.A. Smith, J. Mol. Biol. 148 (1981) 191^197. [15] R.J. King, Exp. Lung Res. 6 (1984) 237^253.
33
[16] R.J. King, M.C. MacBeth, Biochim. Biophys. Acta 647 (1981) 159^168. [17] R.J. King, M.C. Carmichael, P.M. Horowitz, J. Biol. Chem. 168 (1983) 255^262. [18] Y. Kuroki, T. Akino, J. Biol. Chem. 266 (1991) 3068^3073. [19] G.F. Ross, R.H. Notter, J. Meuth, J.A. Whitsett, J. Biol. Chem. 261 (1986) 14283^14291. [20] Y. Kuroki, R.J. Mason, D.R. Voelker, J. Biol. Chem. 263 (1988) 17596^17602. [21] F.X. McCormack, Y. Kuroki, J.J. Stewart, R.J. Mason, D.R. Voelker, J. Biol. Chem. 269 (1994) 29801^29807. [22] F.X. McCormack, J. Stewart, D.R. Voelker, M. Damodarasamy, Biochem. 36 (1997) 13963^13971. [23] Y. Ogasawara, F.X. McCormack, R.J. Mason, D.R. Voelker, J. Biol. Chem. 269 (1994) 29785^29792. [24] F.X. McCormack, P.A. Watson, H.M. Calvert, D.R. Voelker, Am. Rev. Respir. Dis. 145 (1992) A138. [25] F.X. McCormack, S. Pattanajitvilai, J. Stewart, F. Possmayer, K. Inchley, D.R. Voelker, J. Biol. Chem. 272 (1997) 27971^27979. [26] M.L. Ruano, E. Miguel, J. Perez-Gil, C. Casals, Biochem. J. 313 (1996) 683^689. [27] C.A. Benne, C.A. Kraaijeveld, J.A. van Strijp, E. Brouwer, M. Harmsen, J. Verhoef, L.M. van Golde, J.F. van Iwaarden, J. Infect. Dis. 2 (1995) 335^341. [28] K. Drickamer, M.S. Dordal, L. Reynolds, J. Biol. Chem. 261 (1986) 6878^6887. [29] C.D. Gaynor, F.X. McCormack, D.R. Voelker, S.E. McGowan, L.S. Schlesinger, J. Immunol. 155 (1994) 5343^5351. [30] M. Kalina, H. Blau, S. Riklis, V. Kravtsov, Am. J. Physiol. 268 (1995) L144^L151. [31] R. Malhotra, J. Haurum, S. Thiel, J.C. Jensenius, R.B. Sim, Biosci. Rep. 2 (1993) 79^90. [32] H. Manz-Keinke, C. Egenhofer, H. Plattner, J. SchlepperSchafer, Exp. Cell Res. 192 (1991) 597^603. [33] T.B. McNeely, J.D. Coonrod, Am. J. Respir. Cell Mol. Biol. 11 (1994) 114^122. [34] J.F. van Iwaarden, J.C. Pikaar, J. Storm, E. Brouwer, J. Verhoef, R.S. Oosting, L.M. van Golde, J.A. van Strijp, Biochem. J. 303 (1994) 407^411. [35] P.E. Zimmerman, D.R. Voelker, F.X. McCormack, J.R. Paulsrud, W.J. Martin, J. Clin. Invest. 1 (1992) 143^149. [36] J.F. van Iwaarden, J.A. van Strijp, H. Visser, H.P. Haagsman, J. Verhoef, L.M. van Golde, J. Biol. Chem. 267 (1992) 25039^25043. [37] Z.C. Chroneos, R. Abdolrasulnia, J.A. Whitsett, W.R. Rice, V.L. Shepherd, J. Biol. Chem. 271 (1996) 16375^16383. [38] S. Weissbach, A. Neuendank, M. Pettersson, T. Schaberg, U. Pison, Am. J. Physiol. 267 (1994) L660^L666. [39] J.R. Wright, D.C. Youmans, Am. J. Physiol. 264 (1993) L338^L344. [40] S. Rubio, T. Lacaze-Masmonteil, B. Chailley-Heu, A. Kahn, J.R. Bourbon, R. Ducroc, J. Biol. Chem. 270 (1995) 12162^ 12169. [41] T.R. Korfhagen, M.D. Bruno, G.F. Ross, K.M. Huelsman, M. Ikegami, A.H. Jobe, S.E. Wert, B.R. Stripp, R.E. Morris,
BBADIS 61761 23-12-98
34
[42]
[43] [44] [45] [46]
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34 S.W. Glasser, C.J. Bachurski, H.S. Iwamoto, J.A. Whitsett, Proc. Natl. Acad. Sci. U.S.A. 93 (1996) 9594^9599. A.M. LeVine, M.D. Bruno, K.M. Huelsman, G.F. Ross, J.A. Whitsett, T.R. Korfhagen, J. Immunol. 158 (1997) 4336^4340. M. Ikegami, T.R. Korfhagen, M.D. Bruno, J.A. Whitsett, A.H. Jobe, Am. J. Physiol. 272 (1997) L479^L485. A.M. Cockshutt, J. Weitz, F. Possmayer, Biochemistry 29 (1990) 8424^8429. N. Palaniyar, R.A. Ridsdale, C.E. Holterman, K. Inchley, F. Possmayer, G. Harauz, J. Struct. Biol. 122 (1998) 297^310. D. Resnick, J.E. Chatterton, K. Schwartz, H. Slayter, M. Krieger, J. Biol. Chem. 271 (1996) 26924^26930.
[47] H. Efrati, S. Hawgood, M.C. Williams, K. Hong, B.J. Benson, Biochemistry 26 (1987) 7986^7993. [48] A. Meyboom, D. Maretzki, P.A. Stevens, K.P. Hofmann, J. Biol. Chem. 272 (1997) 14600^14605. [49] E. Cowley, K. Govindaraju, K.D. Lloyd, D.H. Eidelman, Am. J. Physiol. 273 (1997) L895^L899. [50] L. Joris, I. Dab, P.M. Quinton, Am. Rev. Respir. Dis. 148 (1993) 1633^1637. [51] L.W. Matthews, S. Spector, J. Lemm, J.L. Potter, Am. Rev. Respir. Dis. 88 (1963) 83^87. [52] J.L. Potter, L.W. Matthews, S. Spector, J. Lemm, Am. Rev. Respir. Dis. 96 (1967) 83^87.
BBADIS 61761 23-12-98
Cation-mediated conformational variants of surfactant protein A Ross A. Ridsdale a , Nades Palaniyar a , Chet E. Holterman a , Kevin Inchley b , Fred Possmayer b , George Harauz a; * a b
Department of Molecular Biology and Genetics, The University of Guelph, 50 Stone Road East, Guelph, Ont. N1G 2W1, Canada Departments of Biochemistry, and Obstetrics and Gynaecology, The University of Western Ontario, London, Ont. N6A 5A5, Canada Received 15 June 1998; accepted 10 August 1998
Abstract Surfactant protein A (SP-A) is the major protein of pulmonary surfactant. This protein is implicated in regulating surfactant secretion, alveolar processing, recycling, and in non-serum-induced immune response. An increasing body of work indicates the importance of cations, particularly calcium, on SP-A function. However, little information exists on the effects of cations on SP-A quaternary structure. Here, the quaternary organisation of bovine surfactant protein A in the presence of cations has been quantitatively and systematically studied by transmission electron microscopy. The conformation of SP-A is altered by the presence of cations, especially calcium, then sodium, and to a small extent, magnesium. There is a transition concentration, unique for each cation, at which a conformational switch occurs. These transition concentrations are: 5 mM for CaCl2 , 100 mM for NaCl and 1 mM for MgCl2 . Below these concentrations, SP-A exists primarily in an opened form with a large head diameter of 20 nm; above it, SP-A is mostly in a closed form due to a compaction of the headgroups resulting in a head diameter of 11 nm. There is a corresponding increase in particle length from 17 nm for opened SP-A to 20 nm for closed SP-A. The fact that the transition concentrations are within physiological range suggests that cation-mediated conformational changes of SP-A could be operative in vivo. ß 1999 Elsevier Science B.V. All rights reserved. Keywords: Pulmonary surfactant; Surfactant protein A; Surfactant protein A structure; Calcium; Cation
1. Introduction Pulmonary surfactant is a complex mixture of phospholipids and proteins that forms a tightly packed lipid monolayer at the air/liquid interface in the lungs. This surfactant works to reduce surface tension which is essential to proper alveolar gas exAbbreviations: CRD, carbohydrate recognition domain; SPA, -B, -C, -D, surfactant protein A, B, C, D; TEM, transmission electron microscope/microscopy; DPPC, dipalmitoylphosphatidylcholine * Corresponding author. Fax: +1-519-837-2075; E-mail: [email protected]
change. Without surfactant the alveoli would tend to collapse during exhalation as a result of intermolecular forces [1]. Surfactant lipids and proteins alike are produced by alveolar type II cells. Some of these components are released as lamellar bodies that are then used to generate the highly structured proteolipid complex known as tubular myelin [2]. Tubular myelin as well as other surfactant aggregates are thought to be storage intermediates that contribute surfactant material to the monolayer [3]. Through a poorly understood process, surfactant is recycled through the type II cells or degraded in alveolar macrophages [4,5]. The most common pulmonary surfactant protein
0925-4439 / 99 / $ ^ see front matter ß 1999 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 5 - 4 4 3 9 ( 9 8 ) 0 0 0 5 7 - X
BBADIS 61761 23-12-98
24
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
is SP-A, a member of the collectin family, which itself is part of the C-type lectin family. Other collectins are mannose-binding protein, complement factor C1q, and another surfactant protein, SP-D. In reduced form, SP-A is 26 to 38 kDa in molecular mass. It has been proposed that there are four structural domains in SP-A (Fig. 1) [6^9]. These domains are: (i) a 7-amino-acid amino-terminus, (ii) a tandemly repeating Gly^X^Y collagen-like region, (iii) an amphipathic `neck' region, and ¢nally (iv) a carboxy-terminal region that is globular and highly similar to other carbohydrate recognition domains. Posttranslationally, SP-A is modi¢ed by cleavage of its signal peptide, hydroxylation of prolines within the collagen-like domain, and asparagine-linked glycosylation [10,11]. Trimeric assembly in SP-A is the result of association through the globular carboxy-terminal domains and the collagen-like regions. Six trimers associate further, among their collagen-like regions, to generate a 700 kDa octadecamer. This gives the protein a characteristic `bouquet of £owers' appearance [12,13] (Fig. 1). Both trimers and octadecamers are postulated to maintain quaternary structure through disulphide bonding [6]. The overall assembly is similar to that of C1q [13,14]. SP-A has a wide variety of potential ligands. It has the capacity to associate lipids, with particular af¢nity for the most abundant and signi¢cant surfactant phospholipid, dipalmitoylphosphatidylcholine (DPPC), in a calcium-dependent manner [15^18].
Earlier studies suggested that this association occurs via the hydrophobic neck [19]; however, there is an accumulation of evidence that identi¢es the carbohydrate recognition domain (CRD) of the headgroup to be the primary region for lipid interaction [20^23]. SP-A can promote aggregation of lipid vesicles in a calcium-dependent manner [24,25], and also self-aggregates in a calcium-in£uenced manner [26]. SP-A
C
Fig. 1. (a) Structural features of the SP-A monomer, trimer and octadecamer. The monomer has a short amino-terminal region, a large collagen-like region (here indicated with a wavy line) which has a kink in it, an amphipathic neck region, and a carbohydrate recognition domain (CRD). In the trimeric form of SP-A (i.e., the basic trimer unit), three CRDs are in close proximity and form a single globular moiety referred to as a headgroup. Six trimers constitute a single SP-A octadecamer. The region from the amino-terminus to the kink forms the octadecamer stem. The distance from the kink to the CRD is referred to as the octadecamer arm length. The region containing the six headgroups is referred to simply as the head. (b^f) When imaged by TEM, SP-A lies in two orientations on the carbon support ¢lm. (b) Pictorial representation and (c) TEM image of SP-A side view. (d) Pictorial representation and (e) TEM image of SP-A top-down view. (f) Pictorial representation of SP-A side view with nomenclature of measured regions.
BBADIS 61761 23-12-98
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
displays xenophilic characteristics by binding to pulmonary microorganisms, lipid A of endotoxin, and pollen grains [27^35]. Many of these associations are attributed to the CRDs and/or to the oligosaccharides that were post-translationally added to SP-A [36]. The collagen-like region of SP-A acts as a ligand to at least one high-a¤nity receptor on macrophages, and can cause chemotaxis and release of anti-microbial agents [37^39]. SP-A is localised both in the distal and conductive respiratory tracts, and may be present in the large and small gastric intestinal tracts [40]. SP-A has been implicated in diverse physiological functions, such as tubular myelin formation, regulating extracellular surfactant concentration by a¡ecting secretion and readsorption, and functioning with SP-B in surfactant adsorption to and spreading at the liquid/air interface [6]. There is an increasing body of evidence concerning SP-A, along with SP-D, acting in host defence because of their xenophilic characteristics and e¡ects on macrophages. For a protein with such a variety of roles, it is surprising that the SPA knock-out mouse shows a viable phenotype [41]. These SP-A knock-out mice lack tubular myelin, are more susceptible to group B Streptococcus [42], but have essentially normal saturated phosphatidylcholine pool sizes [43]. The study of SP-A structural characteristics will improve our understanding of how it performs such a variety of tasks in such di¡erent environments. Our examination here by transmission electron microscopy (TEM) of SP-A in the presence of di¡ering concentrations of mono- and divalent salts indicates that there are two, previously unidenti¢ed, major conformational forms of SP-A. We shall refer to these as closed or compacted SP-A, and opened SP-A. We have characterised the structural anatomy of these in a variety of cationic environments. 2. Materials and methods 2.1. Protein puri¢cation Bovine surfactant was obtained by lung lavage and SP-A puri¢ed as previously described [44]. This protocol involved extracting the surfactant lipids with butanol, and washing with ether/ethanol and ether.
25
The hydrophilic precipitate was dried and resuspended in 5 mM Hepes^NaOH (pH 7.4), 0.1 mM EDTA, and 1.0 mM CaCl2 . This solution was then passed through an immobilised D-mannose column (Pierce Chemical, Rockford, IL, USA). After the column was washed, the SP-A was eluted using 2.0 mM EDTA in 5 mM Hepes^NaOH (pH 7.4). Sample was stored at 4³C until use in elution bu¡er at a concentration between 100 and 250 Wg/ml. 2.2. Protein preparation for TEM A 5-mM Hepes solution was prepared with no salt, 1 M NaCl, 1 M MgCl2 , or 1 M CaCl2 . The pH of each solution was then adjusted to 7.4 using 1 M NaOH. Each bu¡er required approximately the same concentration of NaOH, and so the ¢nal concentration of sodium and hydroxyl ions in each ¢nal reaction bu¡er was negligible (V2.5 mM). Combinations of these basic bu¡ers were made to produce individual dilution series of NaCl (10, 40, 80, 100, 250, 500 and 1000 mM), MgCl2 and CaCl2 (each of 0.01, 0.1, 1, 5, 10 and 20 mM). In a te£on microwell, 12.5 Wl of bu¡er was combined with 0.5 Wl of 250 Wg/ml SP-A to give a ¢nal protein concentration of approximately 10 Wg/ml. Samples were kept humid in a chamber containing the same bu¡er in which the protein was incubated, and left for 24 h at 32³C to achieve conformational equilibrium. 2.3. TEM analysis SP-A samples from the microwells were spread onto 400-mesh Cu grids that were coated by carbon-stabilised holey plastic (cellulose acetate butyrate), and again covered with thin carbon ¢lm of approximately 5 nm thickness. The samples were negatively stained using 2% uranyl acetate and examined using a JEOL JEM-100CX TEM. Images were obtained at a variety of magni¢cations from 66 000U to 130 000U on 80U100-mm plate ¢lm (Kodak Estar 4489). Measurements of dimensions of SP-A octadecamers were made using a digital capture CCD camera (Princeton MicroMax 1400, EMPIX Imaging, Mississauga, Canada) and analysis software (Alpha Innotech, San Leandro, CA, USA). All dimensions henceforth reported will be rounded o¡ to the î ). Simple statistics (mean, standnearest 0.1 nm (1 A
BBADIS 61761 23-12-98
26
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
ard error of the mean, and signi¢cance values for both one- and two-tailed t-tests) were calculated using Excel (Microsoft O¤ce) software. Distribution statistics shall be presented as mean þ standard error of the mean (n = number of measurements). A con¢dence level of 95% (i.e., P 6 0.05) was considered to represent statistical signi¢cance. 3. Results 3.1. TEM of bovine SP-A It was observed in this and in other studies that SP-A portrayed two orientations when examined by TEM [13,45]. The ¢rst view was of a top-down perspective (Figs. 1 and 2). Here, the trimers of globular CRDs, referred to as trimeric headgroups, appeared as individual globules with six such entities being arranged in a wheel formation. The second orientation was a side view in which a stem was visible from the pro¢le and the headgroups appeared at one end of that stem. Side views were more distinctive than top-down views because more of the particle mass was directly apparent in pro¢le. Top-down views appeared to be the major SP-A orientation (approximately 70%) under lower salt conditions; however, they were readily identi¢able only in locations where the stain concentration was relatively dense. For this reason, we focused in this study on the side views, and on the relative changes in dimensions (Fig. 1) and frequencies of occurrence under di¡ering salt conditions. 3.2. Identi¢cation of conformational variants A histogram of head diameter measurements for SP-A under all of these conditions displayed distinct bimodality (Fig. 3). The peak for the ¢rst population (`narrow-head', `closed-head', or `compacted SP-A') was at 11 nm and the peak for the second population (`wide-head', `opened-head', or `opened SP-A') was at 21 nm. The ¢rst (narrow-head) population had a tight distribution with 50% of its population between 9 nm and 11 nm (i.e., a 2-nm range). The second (wide-head) population had 50% of its population spread from 19 to 23 nm (i.e., a 4-nm range), indicating that there were intermediates that covered a
Fig. 2. Appearance of SP-A via TEM of negatively stained preparations. (a) A gallery of SP-A side view particles from various conditions. (b) Image of a typical medium-sized SP-A self-aggregate in 10 mM CaCl2 . (c) A ¢eld of view of a sample of SP-A in 0.5 mM CaCl2 . Very few stems are visible, one indicated by a circle. This image is typical for SP-A samples below the transition concentration. (d) A ¢eld of view of a sample of SP-A in 5 mM CaCl2 , showing a large number of visible stems, some indicated by circles. This view is typical for SP-A samples above the transition concentration. Scale bar = 20 nm.
broad range of head diameter sizes, yet with a peak value at 21 nm which agreed with independent studies [13,45]. Each population type was interpreted to represent a conformational variant of SP-A. The variation in head diameter sizes represented the result of both the range of head movements relative to one another,
BBADIS 61761 23-12-98
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
27
Fig. 3. Histogram of head diameters measured for SP-A under all salt conditions. Two distinct populations are observed. The curves are an interpretation of the trends for each population, and are not intended to represent ¢ts of Gaussian functions, for example.
and the fundamental error in measurement itself. The narrow-head population showed a sharp increase in the histogram from 7 to 8 nm, with very few smaller measurements. This lower limit was attributable to the ¢nite compactibility of the heads. That is to say, the heads can only come to within 8 nm from one another, and lower values (e.g., 7 nm or less) must have been the result of intrinsic errors in measurement. Therefore, these were estimated to be less than 1 nm in 8 nm, or roughly þ 15%. Both narrow- and wide-head diameter conformations appeared to be visible under di¡ering salt conditions. However, there was a shift towards narrowhead SP-A as the corresponding salt concentration was increased. There was a signi¢cant increase in the number of visible stems at 100 mM NaCl, 5 mM CaCl2 and 1 mM MgCl2 (Fig. 4). Clearly, the side view became more favoured after a critical salt condition was achieved, which we henceforth refer to as the transition concentration. (Strictly speaking, the transition salt concentrations lie within ranges of 80^100 mM NaCl, 1.0^5.0 mM CaCl2 , and 0.1^1.0 mM MgCl2 . Here, we shall use the upper value as a conservative and convenient convention.) Furthermore, the total particle length shifted from 16.6 þ 0.4 nm (n = 44) to 20.2 þ 0.6 nm (n = 83)
when the salt concentration rose above the transition concentration (Fig. 5). Such a rise in overall particle length would be expected if the head compacted; then the arm length would have contributed more to particle length. The shift towards narrow-head diameters under higher salt conditions was particularly pronounced with NaCl and CaCl2 , compared with MgCl2 . Since in these experiments there was 20-times the concentration of sodium relative to calcium, cation-mediated conformational change of SP-A was more sensitive to calcium. To determine the rate of these cationic e¡ects, measurements of numbers of visible stems over a 24 h period were performed at either 5 mM CaCl2 , or 100 mM NaCl, i.e., concentrations above the transition ones. There was a steady increase in the number of visible stems over 2 h, when a maximum was achieved (data not shown). Afterwards, there was a slight and slow decrease in the number of stems, reaching 60% of the peak value at 24 h incubation time. 3.3. Synergistic e¡ects of combinations of cations Combinations of CaCl2 , NaCl, and MgCl2 , each below the transition concentration (1 mM, 80 mM,
BBADIS 61761 23-12-98
28
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
BBADIS 61761 23-12-98
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
29
Fig. 4. Histogram of number of visible stems per 100 Wm2 with respect to cation concentration. TEM was performed on randomly selected regions of the grid, and areas of size 10U10 Wm were selected randomly from the image plates. At a particular concentration indicated by arrows, the number of visible stems rises sharply. This transition concentration is di¡erent for each cation. Error bars represent standard deviations, and the number of grid areas examined is between 5 and 7 for each class.
6
Fig. 5. Histograms of particle lengths (a) above and (b) below the transition concentration. The data have been pooled for all three salts studied. There is a distinct shift in the population towards longer particles above the transition concentration. The hypothesis that the two distributions are di¡erent is statistically signi¢cant at a con¢dence level of 95%.
BBADIS 61761 23-12-98
30
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
Fig. 6. In£uence of combinations of cations on the number of SP-A particles visible in pro¢le. A combination of two or more cations below the transition concentration results in an increase in the number of side views. However, the e¡ect is not as distinct above the transition concentration. Error bars represent one standard deviation, and the number of grid areas examined is between 5 and 7 for each class.
0.1 mM, respectively), were tested with SP-A (Fig. 6). The combination of any two or all three cations in solution with SP-A produced only a minor shift in equilibrium towards compacted SP-A, compared to any cation alone. The strong cationic e¡ect noted after the transition concentration (i.e., large number of visible stems) was not induced by any combination of salts below the transition concentration. That is to say, under these experimental conditions (cf. Ref. [45]), the e¡ect did not appear to be fully additive or su¤cient to cause the degree of change as observed after transition. The number of side views observed when exposed to a combination of ions remained consistently larger than for any ion on its own. The combination of all three salts resulted in the highest number of compacted SP-As, compared to any combination of two cations. The relatively small increase in side views with combined ions could be attributed to the relatively small e¡ect of increased ionic strength. 3.4. Observation of individual basic trimer units Although the SP-A complexes were found primarily as octadecamers, a small proportion was seen to be basic trimer units (Fig. 7). In the presence of EDTA or cations after short incubation periods (up to about 4 h), these trimers showed a kink in the
stem region. However, in the presence of 5 mM calcium after extended incubation periods, most of the basic trimer units did not contain the kink. This observation suggested that the cations could alter the SP-A structure in the collagen-like region, albeit slowly under the experimental conditions tested. 4. Discussion 4.1. Cationic e¡ects on SP-A conformation The ¢rst e¡ect of cations was to cause the population of SP-A complexes to shift to a primarily compacted conformational variant. There was a critical salt transition concentration for each cation of 5 mM CaCl2 , 100 mM NaCl, and 1 mM MgCl2 . Below this concentration, SP-A conformation was prevalently wide-head diameter and short-particle length. Above this concentration, the narrow-head and long-particle conformation was predominant, and large numbers of side-view particles appeared (Figs. 2^5). Two sets of distributions were formed and compared, the ¢rst being the particle lengths (Fig. 5), and the second being the ratio of particle head diameters to their lengths (data not shown). Both sets of distributions were di¡erent above and below the transition salt concentrations, statistically signi¢cant at a con¢-
BBADIS 61761 23-12-98
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
31
Fig. 7. Nature of basic trimer units of SP-A found under various conditions. Electron micrographs showing clearly identi¢able SP-A trimers were found in various experiments, although they were rarely seen compared to the octadecamers. (a) Long incubation period (24 h) in the presence of 5 mM CaCl2 . (b) Medium incubation period (1^4 h) in the presence of 5 mM CaCl2 . (c) Short incubation period (0^30 min) in the presence of 5 mM CaCl2 . (d) Long incubation period (24 h) in the presence of 5 mM EDTA. Scale bar = 20 nm.
dence level of 95%. The transition concentration occurred at di¡ering concentrations of Cl3 (i.e., 100 mM in NaCl, 2U5 = 10 mM in CaCl2 , and 2U1 = 2 mM in MgCl2 ) and in solutions of di¡ering ionic strength (i.e., the CaCl2 solution was 15% the ionic strength of the NaCl one, and MgCl2 was 3% the ionic strength of NaCl). Therefore, anions were not the cause of the conformational change and ionic strength could have played only a minor role in compacted SP-A octadecamer formation. The head diameter of SP-A could narrow considerably because headgroups were brought closer in
proximity (Fig. 3). This gross conformational change in the SP-A arm region caused a corresponding increase in particle length (Figs. 4 and 5). The close packing was attributed to a straightening at the junction between the SP-A stem and arm, in the region known as the kink (Figs. 1 and 7). In comparison, class A macrophage scavenger receptor is structurally similar to the SP-A basic trimer (i.e., two ¢brous domains with a hinge, and a trimeric globular headgroup with multiple binding a¤nities), and has recently been shown by TEM to pivot at the hinged region [46]. This structural alteration of the class A
BBADIS 61761 23-12-98
32
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
macrophage scavenger receptor occurred without multiple trimer association, suggesting that SP-A basic trimers also may not need to be within an octadecamer for the kink angle to be altered. The measurements of the number of visible stems at various cation concentrations (Figs. 4 and 6) suggested that the change in SP-A quaternary structure led in turn to a change in preferred orientation on the underlying support ¢lm. An opened-head SP-A would be more frequently observed in top-down orientation, whereas a closed-head SP-A would prefer to lie on its side. Thus, we submit that the most probable explanation for the increase in number of visible stems with increased cation concentration is a physical one: there is an increased number of compacted-head SP-A complexes in solution, and these in turn `fall down' upon the carbon support more naturally on their sides and hence are more easily visualised. 4.2. Cationic e¡ects on SP-A self-association As the concentration of each ion increased beyond the transition concentration, the number of isolated particles per unit area decreased. We propose that the cause for the decrease was the result of SP-A self-aggregation (Fig. 2b) [26], even though we tried to minimise this phenomenon in these experiments by using a low initial protein concentration. Indeed, below the transition cation concentration, aggregates were rarely seen. This post-transitional decrease was most pronounced with CaCl2 (Fig. 5b). With MgCl2 the e¡ect on SP-A structure was much less, both in terms of generating compacted SP-A and of the posttransitional decrease (Fig. 4c). There was a considerable increase in compacted SP-A octadecamers between 100 mM to 250 mM NaCl (Fig. 4a). With the low initial protein concentration used here, the self-aggregation e¡ect did occur from 500 mM NaCl upwards, but was not as large as for calcium. Thus, the second e¡ect of cations was to enhance SP-A self-aggregation. The phenomenon of SP-A self-aggregation is important to consider when examining the literature. Previously, Ruano et al. [26] concluded that the half-maximal lipid aggregation mediated by calcium was in the micromolar range, whereas the SP-A selfaggregation was half-maximal at millimolar concen-
trations. In the latter case, the compacted SP-A conformation could functionally contribute to self-aggregation. Other studies involving calcium-dependent SP-A association with lipid have been conducted using 100 to 150 mM NaCl solutions; therefore, they appear to characterise the e¡ects of calcium on compacted SP-A rather than on opened SP-A (e.g., Refs. [47,48]). Future studies on the relative functions of compacted SP-A should be done at the higher concentrations of 250 mM NaCl, as these conditions generate the most compacted SP-A structure with the least self-aggregation, as well as at low concentrations (e.g., 10 mM NaCl) to compare with opened SP-A. The changes in the numbers of visible stems occurred over relatively long time periods. The 2-h period required to achieve maximal conformational change was much greater than the minutes that produce protein/protein and protein/lipid aggregation in vitro [7,21,23]. However, these aggregation experiments were performed in 150 mM NaCl, a concentration above the transition concentration. Under such conditions, compacted SP-A should have been present in large amounts. Therefore, the rapid aggregation that was observed under those conditions was a distinctly di¡erent phenomenon than the slow conformational changes observed here. According to the above studies and the present one, sodium concentrations of 100 to 150 mM sodium are not su¤cient to induce SP-A self-aggregation or vesicle aggregation, but are capable of inducing conformational change. What remains unknown is whether a particular conformation of SP-A contributes to the generation of these di¡erent forms of aggregation. 5. Conclusions The quaternary conformation of the SP-A octadecamer is signi¢cantly and speci¢cally a¡ected by cations. At a transition concentration that is unique for each cation, the population shifts, at a slow rate, from a primarily opened form to a primarily compacted conformation. The nature of the conformational change is suggested to be due to the straightening of the basic trimer units in the region of the kink. This conformational change in SP-A might have physiological signi¢cance, e.g., by altering SP-
BBADIS 61761 23-12-98
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34
A interaction with its receptors on type II cells, or by altering the ability of SP-A to interact with microbial invaders and with itself. In all cases, the transition salt concentration is close to that under physiological conditions. Interestingly, the NaCl concentration in the bronchial airway is below the transition concentration (ranging from 40 mM [49] to 80 mM [50]), whereas the alveolar concentration is above it [51,52]. The variation of salt conditions in the lung may thus serve as a regulatory element de¢ning SP-A conformation in vivo. Acknowledgements This work was supported by the Ontario Thoracic Society (G.H.), the Natural Sciences and Engineering Research Council of Canada (G.H.), and the Medical Research Council of Canada (F.P.). References [1] R.E. Pattle, Nature 175 (1955) 1125^1126. [2] M.C. Williams, J. Cell Biol. 72 (1977) 260^277. [3] S. Hawgood, B.J. Benson, J. Schilling, D. Damm, J.A. Clements, R.T. White, Proc. Natl. Acad. Sci. U.S.A. 84 (1987) 66^70. [4] J.R. Wright, J.A. Clements, Am. Rev. Respir. Dis. 135 (1987) 426^444. [5] S.L. Young, J.R. Wright, J.A. Clements, J. Appl. Physiol. 66 (1989) 1336^1342. [6] F. McCormack, Chest 111 (1997) 114S^119S. [7] F.X. McCormack, H.M. Calvert, P.A. Watson, D.L. Smith, R.J. Mason, D.R. Voelker, J. Biol. Chem. 269 (1994) 5833^ 5841. [8] K. Sano, J. Fisher, R.J. Mason, Y. Kuroki, J. Schilling, B. Benson, D. Voelker, Biochem. Biophys. Res. Commun. 144 (1987) 367^374. [9] R.T. White, D. Damm, J. Miller, K. Spratt, J. Schilling, S. Hawgood, B. Benson, B. Cordell, Nature 317 (1985) 361^ 363. [10] J. Johansson, T. Curstedt, Eur. J. Biochem. 244 (1997) 675^ 693. [11] T.E. Weaver, J.A. Whitsett, Biochem. J. 273 (1991) 249^264. [12] R.J. King, D. Simon, P.M. Horowitz, Biochim. Biophys. Acta 1001 (1989) 294^301. [13] T. Voss, J. Eistetter, K.P. Schafer, J. Engel, J. Mol. Biol. 201 (1988) 219^227. [14] V.N. Schumaker, P.H. Poon, G.W. Seegan, C.A. Smith, J. Mol. Biol. 148 (1981) 191^197. [15] R.J. King, Exp. Lung Res. 6 (1984) 237^253.
33
[16] R.J. King, M.C. MacBeth, Biochim. Biophys. Acta 647 (1981) 159^168. [17] R.J. King, M.C. Carmichael, P.M. Horowitz, J. Biol. Chem. 168 (1983) 255^262. [18] Y. Kuroki, T. Akino, J. Biol. Chem. 266 (1991) 3068^3073. [19] G.F. Ross, R.H. Notter, J. Meuth, J.A. Whitsett, J. Biol. Chem. 261 (1986) 14283^14291. [20] Y. Kuroki, R.J. Mason, D.R. Voelker, J. Biol. Chem. 263 (1988) 17596^17602. [21] F.X. McCormack, Y. Kuroki, J.J. Stewart, R.J. Mason, D.R. Voelker, J. Biol. Chem. 269 (1994) 29801^29807. [22] F.X. McCormack, J. Stewart, D.R. Voelker, M. Damodarasamy, Biochem. 36 (1997) 13963^13971. [23] Y. Ogasawara, F.X. McCormack, R.J. Mason, D.R. Voelker, J. Biol. Chem. 269 (1994) 29785^29792. [24] F.X. McCormack, P.A. Watson, H.M. Calvert, D.R. Voelker, Am. Rev. Respir. Dis. 145 (1992) A138. [25] F.X. McCormack, S. Pattanajitvilai, J. Stewart, F. Possmayer, K. Inchley, D.R. Voelker, J. Biol. Chem. 272 (1997) 27971^27979. [26] M.L. Ruano, E. Miguel, J. Perez-Gil, C. Casals, Biochem. J. 313 (1996) 683^689. [27] C.A. Benne, C.A. Kraaijeveld, J.A. van Strijp, E. Brouwer, M. Harmsen, J. Verhoef, L.M. van Golde, J.F. van Iwaarden, J. Infect. Dis. 2 (1995) 335^341. [28] K. Drickamer, M.S. Dordal, L. Reynolds, J. Biol. Chem. 261 (1986) 6878^6887. [29] C.D. Gaynor, F.X. McCormack, D.R. Voelker, S.E. McGowan, L.S. Schlesinger, J. Immunol. 155 (1994) 5343^5351. [30] M. Kalina, H. Blau, S. Riklis, V. Kravtsov, Am. J. Physiol. 268 (1995) L144^L151. [31] R. Malhotra, J. Haurum, S. Thiel, J.C. Jensenius, R.B. Sim, Biosci. Rep. 2 (1993) 79^90. [32] H. Manz-Keinke, C. Egenhofer, H. Plattner, J. SchlepperSchafer, Exp. Cell Res. 192 (1991) 597^603. [33] T.B. McNeely, J.D. Coonrod, Am. J. Respir. Cell Mol. Biol. 11 (1994) 114^122. [34] J.F. van Iwaarden, J.C. Pikaar, J. Storm, E. Brouwer, J. Verhoef, R.S. Oosting, L.M. van Golde, J.A. van Strijp, Biochem. J. 303 (1994) 407^411. [35] P.E. Zimmerman, D.R. Voelker, F.X. McCormack, J.R. Paulsrud, W.J. Martin, J. Clin. Invest. 1 (1992) 143^149. [36] J.F. van Iwaarden, J.A. van Strijp, H. Visser, H.P. Haagsman, J. Verhoef, L.M. van Golde, J. Biol. Chem. 267 (1992) 25039^25043. [37] Z.C. Chroneos, R. Abdolrasulnia, J.A. Whitsett, W.R. Rice, V.L. Shepherd, J. Biol. Chem. 271 (1996) 16375^16383. [38] S. Weissbach, A. Neuendank, M. Pettersson, T. Schaberg, U. Pison, Am. J. Physiol. 267 (1994) L660^L666. [39] J.R. Wright, D.C. Youmans, Am. J. Physiol. 264 (1993) L338^L344. [40] S. Rubio, T. Lacaze-Masmonteil, B. Chailley-Heu, A. Kahn, J.R. Bourbon, R. Ducroc, J. Biol. Chem. 270 (1995) 12162^ 12169. [41] T.R. Korfhagen, M.D. Bruno, G.F. Ross, K.M. Huelsman, M. Ikegami, A.H. Jobe, S.E. Wert, B.R. Stripp, R.E. Morris,
BBADIS 61761 23-12-98
34
[42]
[43] [44] [45] [46]
R.A. Ridsdale et al. / Biochimica et Biophysica Acta 1453 (1999) 23^34 S.W. Glasser, C.J. Bachurski, H.S. Iwamoto, J.A. Whitsett, Proc. Natl. Acad. Sci. U.S.A. 93 (1996) 9594^9599. A.M. LeVine, M.D. Bruno, K.M. Huelsman, G.F. Ross, J.A. Whitsett, T.R. Korfhagen, J. Immunol. 158 (1997) 4336^4340. M. Ikegami, T.R. Korfhagen, M.D. Bruno, J.A. Whitsett, A.H. Jobe, Am. J. Physiol. 272 (1997) L479^L485. A.M. Cockshutt, J. Weitz, F. Possmayer, Biochemistry 29 (1990) 8424^8429. N. Palaniyar, R.A. Ridsdale, C.E. Holterman, K. Inchley, F. Possmayer, G. Harauz, J. Struct. Biol. 122 (1998) 297^310. D. Resnick, J.E. Chatterton, K. Schwartz, H. Slayter, M. Krieger, J. Biol. Chem. 271 (1996) 26924^26930.
[47] H. Efrati, S. Hawgood, M.C. Williams, K. Hong, B.J. Benson, Biochemistry 26 (1987) 7986^7993. [48] A. Meyboom, D. Maretzki, P.A. Stevens, K.P. Hofmann, J. Biol. Chem. 272 (1997) 14600^14605. [49] E. Cowley, K. Govindaraju, K.D. Lloyd, D.H. Eidelman, Am. J. Physiol. 273 (1997) L895^L899. [50] L. Joris, I. Dab, P.M. Quinton, Am. Rev. Respir. Dis. 148 (1993) 1633^1637. [51] L.W. Matthews, S. Spector, J. Lemm, J.L. Potter, Am. Rev. Respir. Dis. 88 (1963) 83^87. [52] J.L. Potter, L.W. Matthews, S. Spector, J. Lemm, Am. Rev. Respir. Dis. 96 (1967) 83^87.
BBADIS 61761 23-12-98