Pulmonary surfactant protein SP-B nanorings induce the multilamellar organization of surfactant complexes

Journal Pre-proof Pulmonary surfactant protein SP-B nanorings induce the multilamellar organization of surfactant complexes

Marta Martinez-Calle, Manuel Prieto, Bárbara Aleksander Fedorov, Luís M.S. Loura, Jesús Pérez-Gil

Olmeda,

PII:

S0005-2736(20)30042-0

DOI:

https://doi.org/10.1016/j.bbamem.2020.183216

Reference:

BBAMEM 183216

To appear in:

BBA - Biomembranes

Received date:

13 December 2019

Revised date:

17 January 2020

Accepted date:

2 February 2020

Please cite this article as: M. Martinez-Calle, M. Prieto, B. Olmeda, et al., Pulmonary surfactant protein SP-B nanorings induce the multilamellar organization of surfactant complexes, BBA - Biomembranes(2020), https://doi.org/10.1016/j.bbamem.2020.183216

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2020 Published by Elsevier.

Journal Pre-proof Pulmonary Surfactant Protein SP-B Nanorings Induce the Multilamellar Organization of Surfactant Complexes Marta Martinez-Callea,1, Manuel Prietob, Bárbara Olmedaa,*, Aleksander Fedorovb, Luís L.M.S Lourac,* and Jesús Pérez-Gila. a

Department of Biochemistry and Molecular Biology, Faculty of Biology, and Research Institute “Hospital 12

de Octubre”, Complutense University of Madrid, Madrid, Spain. b

iBB-Institute for Bioengineering and Biosciences, Instituto Superior Técnico, Universidade de Lisboa, Lisboa,

Portugal.

f

Faculty of Pharmacy, Coimbra Chemistry Centre, University of Coimbra, Coimbra, Portugal.

oo

c

pr

Marta Martinez-Calle: [email protected]; Bárbara Olmeda: [email protected]; Jesús Pérez-Gil:

e-

[email protected]

Pr

Manuel Prieto: [email protected]; Aleksander Fedorov: [email protected] Luís M. S. Loura: [email protected]

al

*Corresponding authors:

rn

Luís M. S. Loura, Faculdade de Farmácia, Universidade de Coimbra, Pólo das Ciências da Saúde, Azinhaga de Santa Comba, 3000-548 Coimbra, Portugal, e-mail: [email protected]. Phone number: +351-239488485

Jo u

Bárbara Olmeda, Dept. Bioquimica, Fac. de Biologia, Universidad Complutense, Jose Antonio Novais 12,28040 Madrid, Spain, e-mail: [email protected]. Phone number: +34 913944156

1

Marta Martinez-Calle’s present address: Division of Nephrology and Hypertension, Department of Medicine

and Center for Translational Metabolism and Health, Institute for Public Health and Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois, USA. Electronic address: [email protected].

ABBREVIATIONS. A: FRET acceptor. BODIPY/SP-B: SP-B derivatized with a BODIPY FL probe. D: FRET donor. FRET: Förster resonance energy transfer. LUVs: large unilamellar vesicles. ORB: octadecylrhodamine B. POPC: 1palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine. POPG: 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoglycerol. QCM-D: quartz crystal microbalance with dissipation

Journal Pre-proof Abstract Surfactant protein SP-B is absolutely required for the generation of functional pulmonary surfactant, a unique network of multilayered membranes, which stabilizes the respiratory air-liquid interface. It has been proposed that SP-B assembles into hydrophobic rings and tubes that facilitate the rapid transfer of phospholipids from membrane stores into the interface and the formation of multilayered films, ensuring the stability of the alveoli against physical forces leading to their collapse. To elucidate the molecular organization of SP-B-promoted multilamellar membrane structures, time-resolved Förster Resonance Energy Transfer (FRET) experiments between

BODIPY-PC

or

BODIPY-derivatized

SP-B

(BODIPY/SP-B),

as

donor

probes,

and

octadecylrhodamine B, as acceptor probe, were performed in liposomes containing SP-B or BODIPY/SP-B.

oo

f

Our results show that both SP-B and fluorescently labeled SP-B oligomers mediate the connection of adjacent bilayers. Furthermore, by applying rational models to the FRET data, we have been able to provide quantitative

pr

details of the structure of SP-B-induced multilayered membrane arrays at the nanometer scale, defining interactions between SP-B rings as key elements for connecting surfactant membranes. The data sustain the

Pr

e-

structural model and the mechanism of action of SP-B assemblies to sustain the crucial surfactant function.

Keywords

Jo u

rn

al

Lung surfactant, membrane protein, protein complex, phospholipid bilayer, membrane aggregates, FRET.

2

Journal Pre-proof 1. Introduction Pulmonary surfactant is a lipid-protein system, which covers the respiratory surface in all pulmonated animals. It accomplishes an essential role, reducing the surface tension at the air-liquid interface of alveoli and stabilizing the respiratory epithelium during breathing. Alveolar type II cells synthesize and assemble surfactant components inside acidic organelles called lamellar bodies. Upon stimulation, lamellar bodies are secreted towards the alveolar lining fluid where they disassemble giving rise to the formation of the surfactant film, a pool of interconnected membranes attached to the interfacial monolayer. Regarding surfactant composition,

f

lipids account approximately 90% of its mass, whereas 8-10% are specific proteins: SP-A, SP-B, SP-C and SP-

oo

D [1]. The most abundant phospholipid species in surfactant is dipalmitoylphosphatidylcholine (DPPC) with around 40% of the total mass of surfactant, which is also its main surface-active agent. Other lipid components unsaturated

phosphatidylcholines

(about

25%

by

mass)

pr

include

and

the

anionic

phospholipid

e-

phosphatidylglycerol (PG). The presence of two small hydrophobic proteins, SP-B and SP-C, is strictly required for the optimal interfacial activity of pulmonary surfactant, allowing an efficient adsorption of lipids to the

Pr

interface and a proper respiratory mechanics during inhalation-expiration cycles. Specifically, SP-B exerts these interfacial activities in a highly efficient manner, resulting essential for the maintenance of the respiratory

al

surface [2]. The absence of its expression leads to lethal respiratory failure at birth in humans [3] and in SP-Bdeficient mice [4]. Lack of SP-B not only alters the interfacial dynamics of surfactant membranes but also

rn

causes an incomplete processing of SP-C precursor [5]. SP-B also participates in organizing the surfactant lipid

Jo u

packing to generate lamellar bodies, the secreted surfactant assemblies [6,7]. SP-B is a 79-residue polypeptide with a molecular weight of 8.7 kDa. It belongs to the saposin-like family (SAPLIP), consisting of small proteins with four or five amphipathic α-helices, stabilized by 3 intramolecular disulfide bridges that involve 6 highly conserved cysteines. SP-B has also a seventh cysteine responsible for the formation of the intermolecular disulfide bridge of the covalent homodimer of 18 kDa [8]. Due to its highly hydrophobic content (approximately 40%), SP-B is permanently associated with surfactant membranes. Hydrophobic faces of their helices apparently establish interactions with the acyl chains of phospholipids, allowing an orientation of SP-B parallel to the membrane surface [9–11]. Moreover, SP-B, whose net charge is +7, establishes preferential interactions with polar head groups of anionic phospholipids, such as PG [11,12]. Although its tridimensional structure is not yet fully determined, purification of SP-B from porcine surfactant solubilized by detergent revealed its native oligomeric assembly, consisting of a ring-shaped structure of 10 nm of diameter. According to a 3-D structural model developed from the crystallographic structure of Saposin B 3

Journal Pre-proof protein, the native supramolecular structure of SP-B could consist of 6 covalent dimers assembled into a ringshaped organization, which would interact peripherally with the membranes [13]. Unlike purification of SP-B by detergent solubilization of native surfactant, the use of organic solvents during SP-B classical isolation implies disruption of protein-protein interactions, leading to the extraction of covalent dimers. However, upon interaction with lipids under proper conditions, these SP-B dimers are also able to re-assemble to form the native ring-shaped supradimeric structure in bilayers and interfacial monolayers. Regarding its function, SP-B promotes perturbations in membranes that yield to lipid exchange and further fusion of liposomes and leakage of their content [14–17]. These membrane-perturbing activities provide SP-B with the ability to mediate reorganization of monolayers and bilayers, which allows lipid transfer from

oo

f

membranes to an air-liquid interface [2,18] or into a preformed interfacial monolayer [19,20], as well as the formation and stabilization of multilayered stacks attached to an interfacial monolayer [21–24]. Thus,

pr

destabilization of lipid packing induced by SP-B is key to prompt transitions between different surfactant structures, which ultimately entails the reduction of surface tension at the alveolar air-liquid interface and the

e-

stabilization of the respiratory surface during respiratory cycles.

Pr

The aim of the present work is to characterize the geometry of multilayered stacks formed by POPC/POPG liposomes in the presence of SP-B, using time-resolved FRET methodologies, which are particularly useful to

al

study complex samples such as light-scattering lipid and lipid/protein suspensions. Time-resolved FRET experiments between protein-bound donors and lipid-linked acceptors or between lipid-linked donors and

rn

acceptors detected the presence of membrane aggregates induced by SP-B, with particular sensitivity to

Jo u

distances in the 10-nm scale. Previous studies had demonstrated the great potential of combining time-resolved FRET experiments with further analysis based on meaningful models, which provide quantitative topological information regarding the structural effects of peptides or proteins on the organization of lipid bilayers at the nanoscopic level [25,26]. In this work, this tool allowed to determine that SP-B promotes the formation of multilayered structures with a constant interbilayer distance compatible with the stacking of two SP-B oligomers. Thus, SP-B induces the connection of bilayers through SP-B/SP-B interactions.

4

Journal Pre-proof

2. Materials and methods 2.1. Materials Synthetic lipids 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) and 1-palmitoyl-2-oleoyl-snglycero-3-phospho-(1'-rac-glycerol) (POPG) were purchased from Avanti Polar Lipids (Alabaster, Alabama, U.S.A). BODIPY FL NHS-ester, BODIPY 500/510 C12-HPC and octadecyl rhodamine B chloride (R18), also

f

called ORB, were purchased from Thermo Fisher Scientific (Waltham, Massachusetts, U.S.A.). The

oo

concentrations of stock probe solutions were determined spectrophotometrically using the molar absorption coefficient values ε(BODIPY FL, 504 nm, in ethanol) = 82103 M-1cm-1, ε(BODIPY500/510-PC, 509 nm, in

pr

ethanol) = 86 x 103 M-1 cm-1 [27] and ε(ORB, 555 nm, in ethanol) = 85 x 103 M-1 cm-1 [28].

e-

SP-B was purified from minced lungs as previously described [29]. Briefly, saline lavages from minced

Pr

lungs were subjected to a Bligh and Dyer organic extraction, and SP-B was isolated from the organic extract by two sequential exclusion-size chromatographies using SEPHADEX LH-20 and LH-60 resins (GE Healthcare, Chicago, Illinois U.S.A.). Mobile phases used for the chromatographies were chloroform/methanol (2:1, v/v)

al

for LH-20 and chloroform/methanol (1:1, v/v) + 0.1 N HCl (0.5% v/v) for LH-60. Protein was quantified by

rn

amino acid analysis.

Jo u

SP-B was derivatized with BODIPY FL NHS-ester as described in [30]. Briefly, the apparent pH of the organic solution containing SP-B was adjusted to 6.6 by adding an appropriate volume of 50 mM Tris in methanol. BODIPY FL dissolved in methanol (10 mg/ml) was added to the protein solution in a molar probe/protein proportion of 3:1. Conjugation reaction took place overnight at 4 °C. To stop the reaction, the protein/probe solution was reacidified up to an apparent pH of 2. Then, the non-conjugated probe was removed by a Sephadex LH-20 chromatography (GE Healthcare, Chicago, Illinois U.S.A) in chloroform/methanol (1:1, v/v) + 0.1 N HCl (0.5% v/v). BODIPY FL/SP-B fractions were detected by measuring absorbances at 280 nm and 504 nm. The degree of SP-B labeling was estimated from amino acid analysis and spectroscopic determination of the probe. Probe/protein molar ratio determined for BODIPY FL/SP-B was 0.91. 2.2 Liposome preparation All liposomes were prepared by mixing adequate amounts of synthetic lipids and probes in chloroform/methanol solutions and drying them under a nitrogen stream. Dried films were rehydrated in 5 mM 5

Journal Pre-proof Tris pH 7, 150 mM NaCl at 45 °C. Multilamellar vesicle suspensions were subjected to 5 freeze-thaw cycles and then extruded to form LUVs with a diameter of ̴ 100 nm. Although DPPC is considered as the main surface-active component in native pulmonary surfactant, the structure and membrane perturbing activities of SP-B is preserved in lipid systems with different content of saturated/unsaturated phospholipids [15,30]. It has been also determined that SP-B partitions selectively in liquid-disordered regions of surfactant membranes and films [24,31], presumably enriched in unsaturated phospholipid species. We therefore assume that the reconstitution of SP-B in membranes made of unsaturated phospholipids may reflect better the particular local environment of SP-B in surfactant contexts. On the other hand, we expected that the use of the POPC/POPG lipid mixture in the vesicles might favor both the insertion of

oo

f

the protein once added from methanolic solutions and the eventual formation and equilibration of SP-B assemblies.

pr

For water/membrane partitioning assay of ORB, POPC/POPG (7:3, mol/mol) at variable phospholipid

e-

concentrations (0.05 µM – 1 mM) were labeled with a fixed concentration of ORB (0.75 µM). For FRET experiments and liposome aggregation assay, LUVs consisting of POPC/POPG (7:3, molar ratio)

Pr

were prepared at a lipid concentration of 0.5 mM. Then, 5 µl of native SP-B or BODIPY/SP-B in methanolic solution at different protein concentrations were added to the vesicle suspension. Due to the high miscibility of

al

methanol in water and the hydrophobic character of these proteins, both native SP-B and labeled SP-B are

rn

instantaneously incorporated into membranes. To prepare the protein in methanolic solution, the original protein stock in chloroform/methanol solution was transferred to methanol by chloroform evaporation under nitrogen

Jo u

stream, keeping the protein solution at the desired concentration for the assay. Two different vesicle suspensions were prepared following this procedure: 1) SP-B in methanolic solution was added onto POPC/POPG LUVs labeled with BODIPY-PC or BODIPY-PC and ORB. Experiments were performed at final protein concentrations ranging from 0 to 5 µM SP-B. 2) BODIPY/SP-B in methanol was added onto POPC/POPG LUVs labeled or not with ORB. Final protein concentrations in the suspensions varied between 0.4 and 4.44 µM BODIPY/SP-B. 2.3. Membrane/water partition coefficient measurements of ORB Suspensions of POPC/POPG (7:3, mol/mol) LUVs at different phospholipid concentrations (from 0.05 to 1 mM) with a fixed concentration of ORB (0.75 µM) were prepared and their fluorescence intensity was measured (excitation wavelength of 505 nm). Intensity measurements showed that ORB soluble in water was scarcely fluorescent. Fluorescence intensity of ORB increased linearly with the phospholipid concentration, 6

Journal Pre-proof until mostly reaching a plateau at 0.2 mM of POPC/POPG (Supplementary Fig. S1). The partition coefficient, 𝐾𝑝 , of ORB into POPC/POPG vesicles was calculated from the intensity measurements by using the following equation [32]: 𝐼 − 𝐼𝑊 =

(𝐼𝐿 − 𝐼𝑊 )𝐾𝑝 𝛾𝐿 [𝐿] 1 + 𝐾𝑝 𝛾𝐿 [𝐿] (eq. 1)

where 𝐼𝐿 and 𝐼𝑊 are the limit fluorescence intensity with all the probe in lipid phase or in water, respectively, 𝛾𝐿 is the molar volume of lipid phase, and [𝐿] is the lipid concentration. By plotting the values 𝐼 − 𝐼𝑤 measured at 505 nm versus [𝐿] of the different suspensions of ORB-labeled vesicles and fitting to a hyperbolic regression,

oo

f

a value of 𝐾𝑝 = 9.9 (±5.2)·103 was obtained, considering 𝛾𝑃𝑂𝑃𝐶 = 0.754 M-1 [33] and 𝛾𝑃𝑂𝑃𝐺 = 0.723 M-1 [34].

e-

𝐾𝑝,𝑥 [𝐿] [𝑊] + 𝐾𝑝,𝑥 [𝐿]

(eq. 2)

Pr

𝑥𝐿 =

pr

The membrane-bound solute mole fraction of ORB, 𝑥𝐿 , for a defined lipid concentration is described by:

where [𝑊] is the molar concentration of water, approximately 55.5 M at 25ºC, and 𝐾𝑝,𝑥 is the mole-fraction

al

partition constant, which is related to 𝐾𝑝 via the following equation:

rn

𝐾𝑝 = 𝐾𝑝,𝑥 (

𝛾𝑊 ⁄𝛾𝐿 )

(eq. 3)

Jo u

Using equations 2 and 3, the molar fraction of ORB inserted into 0.5 mM POPC/POPG (7:3, molar) LUVs was estimated to be 𝑥𝐿 = 0.79 (with a range of uncertainty of 0.64-0.85). It should be noted that complete partition of acceptor to the lipid phase is not required, as long as the acceptor surface concentration is accurately calculated. Water-located ORB molecules, which have very weak fluorescence and are incapable of accepting energy from membrane-bound donors, do not affect the FRET measurements. 2.4. Liposome aggregation assay Protein-induced membrane aggregation was determined from the variation in the apparent absorption at 630 nm, which reflects the increase in turbidity of the LUVs suspension containing SP-B or BODIPY/SP-B. After protein addition to 0.5 mM POPC/POPG LUVs (7:3, molar ratio), samples containing different protein concentrations were incubated for 1 hour at room temperature and then, absorbance at 630 nm was measured in a spectrophotometer. Triplicate absorbance measurements versus protein concentration were plotted as mean ± SD. 7

Journal Pre-proof 2.5. FRET experiments Energy transfer experiments between the membrane probes BODIPY-PC (donor) and ORB (acceptor) or BODIPY FL/SP-B (donor) and ORB were performed in POPC/POPG (7:3, mol/mol) at a final lipid concentration of 0.5 mM. For the FRET pair BODIPY-PC/ORB, the donor/lipid ratio was kept at 1:1000, whereas for the FRET pair formed by BODIPY/SP-B and ORB, the donor/lipid ratio was given by the protein content. In all experiments, the ORB/lipid ratio was kept at 1:400. Förster radius (𝑅0 ), or the critical distance at which the transfer efficiency is 50% for an isolated donoracceptor pair, was calculated for both FRET pairs, BODIPY-PC/ORB and BODIPY-SP-B/ORB, using the

oo

f

relationship [35]: ∞

1

𝑅0 = 0.2108 ∙ [𝜅 2 ∙ 𝜙𝐷 ∙ 𝑛−4 ∙ ∫0 𝐼𝐷 (𝜆) ∙ 𝜀𝐴 (𝜆) ∙ 𝜆4 ∙ 𝑑𝜆]6

(eq. 4)

pr

In eq. 4, 𝜅 2 is the orientation factor (assumed in this study as 2/3, the dynamical isotropic value, whose

e-

validity for fluid bilayers is extensively discussed in [36,37]), 𝜙𝐷 is the donor quantum yield in the absence of acceptor, 𝑛 is the refractive index, 𝜆 is the wavelength, 𝐼𝐷 (𝜆) is the normalized donor emission spectrum and

Pr

𝜀𝐴 (𝜆) is the molar absorption spectrum of the acceptor. The quantum yield of the donor, 𝜙𝐷 , was estimated for BODIPY-FL/SP-B inserted into LUVs to be 0.78, assuming proportionality between the quantum yield and the

al

area under the intensity decay. For BODIPY(500-510)-PC, the published value of quantum yield of the probe in ethanol was used (𝜙𝐷 = 0.90) [27]. From these values and the measured spectra, 𝑅0 of 5.5 and 5.7 nm were

rn

calculated for BODIPY-PC/ORB and BODIPY-SP-B/ORB FRET pairs, respectively.

Jo u

All energy transfer measurements were carried out at room temperature after addition of SP-B or BODIPY FL/SP-B in methanolic solution to vesicle suspensions and subsequent sample stabilization for 1 hour at room temperature. Time-resolved FRET measurements were obtained by fluorescence intensity donor decays in the absence and in the presence of the acceptor probe (ORB). Data analysis was carried out using a nonlinear leastsquares iterative convolution based on the Marquardt algorithm [38]. The goodness of the fit was determined by global XG2 values and a random distribution of weighted residuals and autocorrelation plots. For each experimental condition, FRET efficiency was obtained. From the analysis of time resolved donor decays in the absence (𝑖𝐷 (𝑡)) and in the presence of acceptors (𝑖𝐷𝐴 (𝑡)), FRET efficiency, 𝐸𝑡 , can be calculated by ∞

∫ 𝑖𝐷𝐴 (𝑡)𝑑𝑡 𝐸𝑡 = 1 − 0 ⁄∞ ∫0 𝑖𝐷 (𝑡)𝑑𝑡

(eq. 5)

Moreover, donor intensity decays were faced to different models of protein-membrane interactions, using a set of formalisms (see “FRET modelling” subsection). 8

Journal Pre-proof 2.6. FRET modelling Time-resolved FRET experiments can offer quantitative topological information regarding the interaction of the protein SP-B with POPC/POPG model membranes. To get this aim, two different experimental configurations were considered: 1) FRET between membrane probes (figures 1A-B) and 2) FRET between BODIPY/SP-B and an acceptor membrane probe (figures 1C-E). For each FRET pair, different formalisms

Jo u

rn

al

Pr

e-

pr

oo

f

were developed based on [26], describing different donor/acceptor distributions.

Figure 1. Schematic diagrams of the FRET models used for the study of bilayer stacking induced by SP-B nanooligomers. FRET between two membrane probes (A, B) or between SP-B labeled with BODIPY and acceptor membrane probe (C–E). A and C depict topological models for the interaction of SP-B rings with only one bilayer, whereas B, D and E illustrate arrangements for SP-B-mediated aggregation of bilayers. These models only considered the interaction between two bilayers, since the contribution to FRET of a third bilayer would result negligible given the Förster radii of the FRET pairs. h1, h2 and h3 indicate the donor-acceptor distances for each FRET situation. In the case of BODIPY/SP-B, h1 and h2 would be identical only if single SP-B oligomers would be located between both membranes (panel D), but not if membrane apposition would be mediated by the interaction of two oligomers (panel E), as h1 is the distance between the plane of protein donors to the lipid acceptors on the bilayer that is in direct interaction with this same protein, whereas h2 is the distance to the bilayer in direct interaction with the opposing protein oligomer.

Figure 1A illustrates a FRET scenario between membrane probes in the absence of vesicle aggregation. Each donor fluorophore transfers energy to acceptor molecules located in the same leaflet at transverse distance h1, or 9

Journal Pre-proof located in the opposite leaflet of the bilayer at distance h2. So, the donor decay in the presence of uniformly distributed acceptor is defined by: 𝑖𝐷𝐴 (𝑡) = 𝑖𝐷 (𝑡)𝜌(𝑡, ℎ1 )𝜌(𝑡, ℎ2 )

(eq. 6)

𝑖𝐷 = ∑𝑖 𝐴𝑖 exp⁡(−𝑡⁄𝜏𝑖 )

(eq. 7)

where

is the donor decay in the absence of acceptor, given by a sum of exponentials. The FRET contributions were calculated from [39]: 𝜌(𝑡, ℎ) = exp⁡(−𝑡𝑘𝐶ℎ𝐹(𝑡, ℎ))

(eq. 8)

2

f

where

oo

𝐶 = Γ (3) 𝑛𝜋𝑅0 2 𝑡

𝑅

0 6 ∞ 1−exp[(−𝜏)( ℎ ) 𝛼 ] 𝑑𝛼 ∫0 𝛼3

(eq. 10)

e-

𝐹(𝑡, ℎ) =

6

pr

and

(eq. 9)

[25], and Γ is the complete gamma function.

Pr

being 𝑘 = 2/𝑅02 ⁡, 𝑛 is the surface density of acceptors, 𝜏 is the average donor lifetime in the absence of acceptor If ℎ1 ≪ 𝑅0 , where donor and acceptor probes show identical transverse location, the corresponding FRET

al

contribution is simplified

(eq. 11)

Jo u

rn

𝜌(𝑡, ℎ1 ) ≅ ⁡𝜌(𝑡) = exp [−𝐶

1

𝑡 3 (𝜏) ]

Figure 1B illustrates the situation of SP-B-induced bilayer aggregation. Here, a new acceptor plane becomes available for each donor at distance h3. Thus, an additional term is inserted in the donor decay law, 𝑖𝐷𝐴 (𝑡) = 𝑖𝐷 (𝑡)𝜌(𝑡, ℎ1 )𝜌(𝑡, ℎ2 )𝜌(𝑡, ℎ3 )

(eq. 12)

Figure 1C depicts the FRET situation between donors bound to SP-B and acceptor membrane probe without vesicle aggregation. Each donor fluorophore can only sense acceptors that share the same transverse location 𝑖𝐷𝐴 (𝑡) = 𝑖𝐷 (𝑡)𝜌(𝑡, ℎ1 )

(eq. 13)

Figures 1D-E refers to FRET from BODIPY/SP-B to two planes of acceptor of different bilayers. Both cartoons depict the situation of vesicle aggregation induced by the protein, although they differ on the mechanism by which the protein mediates contacts between bilayers, through a single SP-B oligomer (D) or 10

Journal Pre-proof through the vertical apposition of two SP-B oligomers (E). In the case depicted in figure 1E, the donor decay is given by 𝑖𝐷𝐴 (𝑡) = 𝑖𝐷 (𝑡)𝜌(𝑡, ℎ1 )𝜌(𝑡, ℎ2 )

(eq. 14)

If the membrane aggregation is induced by only one single SP-B oligomer (figure 1D), then ℎ1 ≈ ℎ2 (similar distance from donors to both acceptor planes). 2.7. Instrumentation Absorption spectroscopy was performed with a Shimadzu UV-3101PC spectrophotometer (Shimadzu

f

Scientific Instruments, Kyoto, Kyoto, Japan). When required, absorption spectra were corrected for turbidity

oo

using the method described by [40].

pr

Time-resolved fluorescence intensity decays with picosecond resolution were obtained by the time-correlated single-photon timing system described elsewhere [36,41]. The samples were excited at 335 nm by pulses from a

e-

frequency doubled, mode-locked Tsunami Ti: sapphire laser (Spectra Physics, Chicago, Illinois U.S.A) pumped by Nd:YVO4 diode laser (Spectra Physics, Chicago, Illinois U.S.A). The fluorescence was detected by a

Pr

Hamamatsu R-2809U microchannel plate photomultiplier (Hamamatsu Photonics K.K., Hamamatsu City, Shizuoka, Japan) at 515 nm that was selected using a Jobin-Yvon HR 320 monochromator (Horiba, Kyoto,

al

Kyoto, Japan) in combination with an adequate cut-off filter to avoid interference from Rayleigh-scattered light.

rn

The instrument response function (IRF) was recorded as excitation light scattered by a Ludox solution (silica, colloidal water solution from Sigma Aldrich, St. Louis, Missouri, USA). Data were collected using a

Jo u

multichannel analyzer with a time window of 1024 channels, at time scale of 39.1 ps/channel, and up to 50000 and 20000 counts in the peak channel of the IRF and decay curves, respectively. Quartz cuvettes of 5 x 5 mm were used for all spectroscopy measurements. In the case of time-resolved measurements, samples were continuously agitated with a magnetic stirring bar. 2.8. Atomic force microscopy SP-B or BODIPY/SP-B dissolved in methanolic solution was added to 0.5 mM POPC/POPG LUVs (7:3, mol/mol) at a final protein concentration of 5 µM. After 1 hour at room temperature, proteolipid samples were deposited onto freshly cleaved mica surface, allowing vesicle adsorption and spreading. Bilayer-coated mica supports were then washed with milli-Q water and dried under nitrogen flow. AFM images were taken using a Multimode Nanoscope IIIA microscope (Veeco Instruments, Plainview, New York, USA) operating in tapping mode. Tips used were super sharp silicon TESPSS probes (Bruker, Billerica, Massachusetts, USA), with 2-5 nm 11

Journal Pre-proof of radius and nominal spring constant and resonance frequency of 42 N/m and 320 kHz, respectively. Image

Jo u

rn

al

Pr

e-

pr

oo

f

analysis was performed using NanoScope Analysis software (Bruker, Billerica, Massachusetts, USA).

12

Journal Pre-proof 3. Results To study the structure of aggregates formed by SP-B and negatively-charged membranes using time-resolved FRET methodologies, two different energy transfer experiments were designed: 1) FRET between membrane probes, BODIPY bound to phosphatidylcholine (BODIPY-PC) and octadecylrhodamine B (ORB); and 2) between SP-B labeled with BODIPY FL (BODIPY/SP-B) and the membrane probe ORB. The FRET pair BODIPY/ORB was selected in all cases as the most appropriate considering their R0 and the range of distances expected to define the organization of SP-B/lipid multilayers. 3.1. FRET between membrane probes

oo

f

The ability of SP-B to induce membrane aggregation was firstly evaluated by turbidity analysis. SP-B solubilized in methanol was added (at different protein concentrations, ranging from 0 to 5 µM) to preformed

pr

POPC/POPG (7:3, mol/mol) LUVs. Absorption measurements at a wavelength of 630 nm showed higher turbidity of vesicle suspensions as the protein concentration increases (figure 2A). Thus, SP-B added in

e-

methanolic solution induced vesicle aggregation, as previously reported at the literature [14,15,17].

Pr

To study the effect of SP-B on the organization of POPC/POPG membrane aggregates, time-resolved FRET experiments were then performed introducing two membrane probes, BODIPY-PC as donor (D) and ORB as

al

acceptor (A). This FRET pair showed a large R0 value of 5.7 nm. Once SP-B dissolved in methanolic solution (at different protein concentrations, 0-5 µM of SP-B) was added to labeled POPC/POPG LUVs, the

rn

fluorescence donor decays were measured in the absence or in the presence of acceptor. Time-resolved experiments showed that FRET efficiency between membrane probes increased upon addition of SP-B to

Jo u

vesicle suspension (figure 2B), confirming that SP-B dissolved in methanolic solution induced vesicle aggregation. The graph depicts that FRET efficiency rose in a protein concentration-dependent manner, although this tendency was not maintained for SP-B concentrations higher than 2.5 µM. This could be due to differences in membrane probes concentration among samples. The analysis of absorption ratios of donor to acceptor molecules (εD/εA) in the vesicle suspensions showed the existence of slight differences in the concentration of probes in the different samples (figure 2C). Particularly, D/A absorption ratios were diminished for samples containing 0.44 and 2.5 µM of SP-B. Moreover, the maximal absorption values for ORB in these two samples were slightly higher (data not shown), showing that the acceptor concentration in these samples was above the expected value, which likely caused a slight overestimation of FRET efficiency values for 0.44 and 2.5 µM of SP-B. In any case, it should be noted that, even if turbidity increases continuously as the aggregation process evolves, the FRET efficiency is not expected to exceed a high limiting value (see below, in FRET analysis between BODIPY/SP-B and membrane probe). 13

Pr

e-

pr

oo

f

Journal Pre-proof

Jo u

rn

al

Figure 2. Liposome aggregation induced by SP-B. A fixed volume of 5 µl of a methanolic solution containing different SP-B amounts was added to 0.5 mM POPC/POPG (7:3, molar ratio) LUVs. Final SP-B concentrations in the vesicle suspension were: 0, 0.4, 1, 2.5 and 5 µM SP-B. A) Turbidity of LUVs induced by SP-B, measured by the increase in apparent absorbance at 630 nm. Turbidity was measured in three different batches of liposome suspensions: two batches of unlabeled liposomes and one of labeled liposomes (BODIPY-PC and ORB; probe/phospholipid molar ratios, 1:1000 and 1:400, respectively). B) FRET efficiency between BODIPY-PC (donor) and ORB (acceptor) in labeled vesicles (probe/phospholipid molar ratios, 1:1000 and 1:400, respectively). Data are plotted as mean ± SD. C) Donor to acceptor absorption ratios for the different samples. Maximal BODIPY-PC and ORB absorption values were measured at 510 and 562 nm, respectively.

For an extensive study of membrane organization in the presence of SP-B, the experimental fluorescence donor decays in the presence and absence of acceptor were analyzed using the formalisms described in “FRET modelling” at the Materials and Methods section (eq. 6 and 12). For FRET between membrane probes, the distribution model predicts that donors and acceptors in the same leaflet of the bilayer locate at the same plane, at a fixed distance h1 ≈ 0 nm (figure 1A and table 1). The fluorophore of ORB locates at the lipid interface, whereas for BODIPY-PC a previous work determined that this probe has a main population in which the fluorophore loops to the interface and another one with the fluorophore embedded within the bilayer [42]. Considering the coexistence of these two populations, the repeat distance recovered from this model geometry would probably correspond to the major one. Then, the distance between donor and acceptor fluorophores in opposite leaflets of the bilayer, h2, was determined by fitting fluorescence decays in the absence of SP-B with 14

Journal Pre-proof diverse h2 values to the equation 6. The best fit was achieved with h2 = 3.9 nm for POPC/POPG (7:3, mol/mol) bilayers. In this same line, a previous work determined by time-resolved FRET methodology that the thickness of POPC/POPS (4:1, molar ratio) was 3.7 nm [26]. Fixing h1 ≈ 0 nm and h2 ≈ 3.9 nm, a third plane of acceptors in an additional bilayer was attempted at variable distance h3 (figure 1B and table 1). The inclusion of a third plane of acceptors for samples with [SP-B] ≤ 1 µm did not led to an improvement in the fitness compared to the two acceptor planes model. However, for higher SP-B concentrations (≥ 2.5 µM), the incorporation of a third plane of acceptors resulted in a refinement of the goodness of the fit (XG2). In addition, the acceptor surface density for the different protein concentrations settled around a constant value, as physically expected, apart from the suspension with 2.5 µM SP-B. The imbalance for this sample might be due to an excessive acceptor

f

concentration in the lipid mixture, as previously observed (figure 2C). Recovered distances h3 varied inversely

oo

to [SP-B]. For instance, the sample with 2.5 µM of SP-B showed a good fitting of the aggregated bilayers model with an interbilayer distance (h3) of 9.3 nm, whereas for 5 µM of SP-B the recovered distance was 8.1

pr

nm. Despite the range of protein concentrations assayed does not allow confirmation of whether this distance h3

e-

would be stabilized to around 8.1 nm for higher SP-B concentrations, it is reasonable to assume this is the case. Thus, results obtained from the FRET experiment including both probes in the membrane revealed that SP-B is

Pr

able to promote membrane aggregation at high protein concentrations.

rn

al

Table 1. Comparison of the fitting parameters recovered for single bilayer and stacked bilayer models applied to FRET experiments between the membrane probes BODIPY-PC and ORB, in the presence of increasing concentrations of SPB. Single bilayer Aggregated bilayers

h1 ≈ 0 nm; h2 = 3.9 nm

h1 ≈ 0 nm; h2 = 3.9 nm; h3, optimized

n

XG2

h3 (nm)

n

XG2

0

0.00295

1.10

>10

0.00286

1.11

0.00316

1.08

>10

0.00308

1.12

0.00329

1.03

>10

0.00319

1.04

2.5

0.00407

1.08

9.3

0.00385

1.05

5

0.00331

1.11

8.1

0.00301

1.06

0.4 1

Jo u

[SP-B] (µM)

Parameters obtained from pairwise global analysis of time-resolved decays of BODIPY-PC (donor) in the absence and in the presence of ORB (acceptor) for POPC/POPG (7:3, molar ratio) LUVs with different SP-B concentrations (0.5 mM of total phospholipid; molar ratios of probe/phospholipid were 1:1000 for BODIPY-PC and 1:400 for ORB). n is the number of acceptors/nm2. The time-resolved FRET data were analyzed assuming BODIPY-PC looping, so h1 ≈ 0 was assumed; and h2 = 3.9 nm was recovered from fitting the data obtained from the protein free sample to the single bilayer model.

3.2. FRET between BODIPY/SP-B and membrane probe

15

Journal Pre-proof The ability of BODIPY/SP-B to also induce vesicle aggregation was confirmed by absorption measurements at 630 nm (figure 3A). Upon addition of increasing concentrations of BODIPY/SP-B dissolved in methanol (at final protein concentrations from 0 to 4.40 µM), the turbidity of POPC/POPG vesicle suspensions increased linearly. At the highest protein concentrations assessed, both SP-B and BODIPY/SP-B induced vesicle

e-

pr

oo

f

aggregation to a similar extent (figures 2A and 3A).

rn

al

Pr

Figure 3. Liposome aggregation induced by BODIPY/SP-B. A fixed volume of 5 µl of a methanolic solution containing different BODIPY/SP-B amounts were added to 0.5 mM POPC/POPG (7:3, molar ratio) LUVs. Increasing final SP-B concentrations in the vesicle suspension were: 0, 0.44, 0.88, 2.20 and 4.40 µM BODIPY/SP-B. A) Turbidity of LUVs induced by SP-B. The extent of light scattering was determined by absorbance at 630 nm. Turbidity was measured in three different batches of liposome suspensions: two batches of unlabeled liposomes and one of labeled liposomes (ORB/phospholipid molar ratio, 1:400). Data are plotted as mean ± SD. B) FRET efficiency between BODIPY/SP-B (donor) and ORB (acceptor) in labeled vesicles (ORB/phospholipid molar ratio, 1:400).

Jo u

Time-resolved FRET experiments were then performed using BODIPY/SP-B and ORB as FRET pair, with a value of R0 = 5.5 nm. Addition of increasing BODIPY/SP-B concentrations led to a progressive increase in FRET efficiency (figure 3B). For protein concentrations between 0.44 and 2.2 µM of BODIPY/SP-B the efficiency underwent an apparently linear increase and then remained constant for higher protein concentrations. Therefore, FRET efficiencies suggested the existence of multibilayer structures induced by [BODIPY/SP-B] ≥ 2.2 µM. The apparent stabilization of FRET efficiency values for BODIPY/SP-B concentrations higher than 2.2 µM probably stems from the fact that, for higher concentration values, even if more bilayers become stacked each one on top of another, most donors will already report a saturation efficiency of FRET. This corresponds to complete coverage by adjacent bilayers, immediately on top or below their own locations in the stack. Because FRET only reports interactions between adjacent bilayers in the stack, it is not being affected by increasing the size of the aggregate upon further protein addition (unlike turbidity).

16

Journal Pre-proof For obtaining quantitative information of protein and membrane geometry, time-resolved donor fluorescence decay curves were analyzed using the formalisms described in “FRET modelling” in the Materials and methods section (eq. 13-14). The distance (h1 = 2.5 nm) between the protein donor to one acceptor plane was fixed from the best fit to the donor decay of vesicle suspension with the lowest protein concentration (0.44 µM BODIPY/SP-B) (figure 1C). This single acceptor plane turned out to be a well-fitting model for [BODIPY/SPB] ≤ 0.88 µM (table 2). On the contrary, in the case of higher protein concentrations, data did not fit to such a model of single bilayer. Comparing the two FRET models under consideration, XG2 values underwent a significant improvement (> 24%) in samples with [BODIPY/SP-B] ≥ 2.2 µM when an additional second plane was allowed, at a distance (h2) of around 5-5.5 nm (table 2 and figure 1D-E). Figure 4 shows the decay of donor

f

emission in the absence and in the presence of acceptor for samples with [BODIPY/SP-B] = 4.4 µM, and

oo

depicts how the experimental data fitted better to the aggregated bilayers model, as it is drawn from residual weights and autocorrelation plots. Our results showed that concentrations of BODIPY/SP-B from 2.2 µM

pr

induced the formation of aggregated membrane structures, where bilayers connected by SP-B were located at a

e-

distance of approximately 7.5 nm (h1+h2). Moreover, given that FRET data analysis for aggregated bilayers model showed a geometrical arrangement of BODIPY/SP-B and ORB at distances h1 ≠ h2, it entails that

Pr

connection between bilayers is mediated by the vertical apposition of SP-B oligomers (figure 1E and table 2).

h1 = 2.5 nm; h2, optimized

rn

h1 = 2.5 nm

al

Table 2. Comparison of the fitting parameters recovered for single bilayer and stacked bilayers models applied to FRET experiments between BODIPY/SP-B and ORB. Single bilayer Aggregated bilayers

XG2

h2 (nm)

n

XG 2

1.20

>10

0.00292

1.21

0.00420

1.22

>10

0.00413

1.22

2.2

0.00649

1.83

5.5

0.00416

1.12

4.4

0.00575

1.49

5.0

0.00370

1.13

n

0.44

0.00297

0.88

Jo u

[BODIPY/SP-B] (µM)

Parameters obtained from pairwise global analysis of time-resolved decays of BODIPY/SP-B (donor) in the absence and in the presence of ORB (acceptor) for POPC/POPG (7:3, molar ratio) LUVs with different BODIPY/SP-B concentrations (0.5 mM of total phospholipid; ORB/phospholipid molar ratio, 1:400). n is the number of acceptors/nm2. The distance h1 = 2.5 nm was recovered from fitting the data obtained from the sample with the lowest [BODIPY/SP-B] (0.44 µM) to the single bilayer model.

Additionally, these time-resolved donor decays were analyzed using a linear combination of the formalisms for one-acceptor plane and two-acceptor planes, with fixed distances h1 = 2.5 nm and h2 = 5 nm (table 3). The fraction of donor molecules that transfer energy to one acceptor plane or to two acceptor planes was calculated for each protein concentration. So, for the lowest BODIPY/SP-B concentration, all donors transferred energy to 17

Journal Pre-proof only one acceptor plane at a distance h1 = 2.5 nm (f2planes = 0.001). Conversely, for [BODIPY/SP-B] ≥ 2.2 µM, the whole fraction of donors sensed the two acceptor planes (f2planes = 1). Thus, all BODIPY/SP-B molecules were involved in vesicle aggregation at these high protein concentrations. An intermediate situation was reported for samples with [BODIPY/SP-B] = 0.88 µM, where the best fit to this linear combination model indicated that 70% of donors transferred energy only to a single acceptor plane, against 30% of donors that did

rn

al

Pr

e-

pr

oo

f

it to two acceptor planes (f2planes = 0.299).

Jo u

Figure 4. Time-resolved fluorescence intensity decays of BODIPY/SP-B in phospholipid bilayers. Upper panels: fluorescence intensity BODIPY/SP-B decay in the absence (donor only, blue) or presence of ORB (donor + acceptor, red) in POPC/POPG (7:3, mol/mol) LUVs, upon addition of 4.4 µM BODIPY/SP-B (0.5 mM of total phospholipid; ORB/phospholipid molar ratio, 1:400). The fittings of donor fluorescence decays to single bilayer (left) and to stacked bilayer model (right) are compared. Weighted residual plots (middle panels) and autocorrelation plots (lower panels) are included for the fits of donor and donor + acceptor decays to the single bilayer (left) and aggregated bilayers models (right). Table 2. Comparison of the fitting parameters recovered for single bilayer and stacked bilayers models applied to FRET experiments between BODIPY/SP-B and ORB. Single bilayer Aggregated bilayers

h1 = 2.5 nm

h1 = 2.5 nm; h2, optimized h2 (nm)

n

XG 2

1.20

>10

0.00292

1.21

0.00420

1.22

>10

0.00413

1.22

2.2

0.00649

1.83

5.5

0.00416

1.12

4.4

0.00575

1.49

5.0

0.00370

1.13

[BODIPY/SP-B] (µM)

n

XG

0.44

0.00297

0.88

2

18

Journal Pre-proof Parameters obtained from pairwise global analysis of time-resolved decays of BODIPY/SP-B (donor) in the absence and in the presence of ORB (acceptor) for POPC/POPG (7:3, molar ratio) LUVs with different BODIPY/SP-B concentrations (0.5 mM of total phospholipid; ORB/phospholipid molar ratio, 1:400). n is the number of acceptors/nm2. The distance h1 = 2.5 nm was recovered from fitting the data obtained from the sample with the lowest [BODIPY/SP-B] (0.44 µM) to the single bilayer model.

3.3. Oligomeric organization of SP-B and BODIPY/SP-B Previous studies have shown that the native structure of SP-B in pulmonary surfactant membranes consists of an oligomer [13]. Throughout this work we studied the organization of SP-B/negatively-charged membrane structures by addition of SP-B or BODIPY/SP-B dimers dissolved in methanol to preformed POPC/POPG

oo

f

vesicles. According to the performed time-resolved FRET analysis, SP-B promotes the formation of interconnected bilayer assemblies by the interaction between protein oligomers. To evaluate the ability of the

pr

dimeric form of the protein obtained in organic solution to form oligomers, SP-B or BODIPY/SP-B in methanolic solution were added to preformed POPC/POPG (7:3, molar ratio) LUVs and samples were

e-

visualized by atomic force microscopy. Micrographs of both samples showed the presence of ring-shaped

Pr

particles with similar averaged diameters (13.8 ± 2.7 nm for SP-B, and 13.9 ± 2.9 nm for BODIPY/SP-B) (figure 5). These particles were consistent in size and shape with the SP-B oligomers previously described by transmission electron and atomic force microscopies [13]. Thus, both SP-B and BODIPY/SP-B dimers

Jo u

rn

al

solubilized in methanol were able to assemble into ring-shaped oligomers in the presence of liposomes.

Figure 5. AFM micrographs of SP-B and BODIPY/SP-B nanorings in bilayers. SP-B (A) or BODIPY/SP-B (B) in methanolic solution were added to 0.5 mM POPC/POPG (7:3, molar ratio) LUVs, at a final protein concentration of 5 µM. The lipid/protein LUVs were then adsorbed and spread onto mica supports for AFM imaging. Scale bars, 50 nm. Magnified views of ring shaped SP-B nanoparticles are shown in the insets of its corresponding sample (scale bars, 10 nm).

19

Journal Pre-proof

4. Discussion SP-B is a hydrophobic protein widely known for promoting membrane perturbations leading to vesicle

f

aggregation, lipid exchange and membrane fusion [14,15,17,43]. These activities are crucial for SP-B to

oo

promote the reorganization of lung surfactant layers to transfer surface-active phospholipids into multilayered interfacial films at the respiratory surface. As reported in those previous studies, the present work confirmed, by

pr

turbidity measurements, that upon addition of SP-B or BODIPY/SP-B to vesicle suspensions both proteins had the ability to induce vesicle aggregation. Time-resolved FRET experiments performed upon addition of SP-B

e-

(both labeled and unlabeled) also suggested that high protein concentrations induced vesicle aggregation,

Pr

because FRET efficiencies reached maximal values for protein concentrations ≥ 2.2 µM of BODIPY/SP-B and ≥ 2.5 µM SP-B, with higher relative increases in the case of BODIPY/SP-B (compare figures 2B and 3B). This was an expected result considering that donor-acceptor distances between protein-bound donor and the

al

membrane probe ORB are shorter than distances between the FRET pair constituted by membrane probes,

rn

BODIPY-PC/ORB, located in different membranes, hence resulting in a higher energy transfer efficiency.

Jo u

Although FRET efficiency is a useful parameter to inform about the relative proximity between probes, it cannot resolve changes in membrane organization. Therefore, more advanced models to fit FRET experimental data are required [25,26]. In this work, the analysis of the fluorescence intensity donor decays was performed using FRET formalisms describing different probe distributions which correspond to two distinct models: single bilayer or aggregated bilayers. Our results showed that increasing concentrations of SP-B or labeled SP-B led to a progressive approaching between membranes and the generation of membrane aggregates. Using a linear combination of the formalisms describing the models for single bilayer and aggregated bilayers, we determined quantitatively the capability of BODIPY/SP-B to induce membrane aggregation. BODIPY/SP-B was able to promote connection between membranes from a protein concentration of 0.88 µM, where non-aggregated and aggregated vesicles coexisted. However, higher protein concentrations (≥ 2.2 µM) led to a scenario where all BODIPY/SP-B molecules were connecting membranes (figure 6A). Furthermore, analysis of time-resolved FRET experiments between BODIPY/SP-B and ORB revealed that the interaction between two labeled SP-B oligomers induced the formation of membrane aggregated structures, where the bilayers connected by the 20

Journal Pre-proof protein were located at a distance of approximately 7.5 nm (h1+h2) (figure 6B). On the other hand, FRET experiments using membrane probes, BODIPY-PC and ORB, showed an interbilayer distance (h3) of 8.1 nm for unlabeled SP-B-mediated membrane aggregates (figure 6C). It is relevant to note that the interbilayer distance (h1+h2), estimated using the protein-bound donor and ORB as FRET pair, is expected to be determined more accurately because the recovered distances refer to the attachment point between the two bilayers. Beyond this consideration, it is clear that both FRET experimental configurations revealed the effect of SP-B

Jo u

rn

al

Pr

e-

pr

oo

f

oligomerization on mediating membrane aggregation.

Figure 6. Schematic representation of membrane stacking mediated by SP-B. (A) BODIPY/SP-B promotes membrane approaching and extensive multilayer stacking in a concentration-dependent manner (0.44, 0.8 and 2.2 µM BODIPY/SP-B). Lower panels illustrate the distances determined by BODIPY-SP-B/ORB (B) and BODIPY-PC/ORB (C) FRET pairs in POPC/POPG membranes mediated by BODIPY/SP-B or SP-B (protein concentrations 4.4 µM and 5 µM, respectively). Distance values are reported in nanometers.

SP-B dimerization has been shown to be important for the in vitro and in vivo function of the protein [44,45]. Furthermore, oligomerization of SP-B in supradimeric forms had been previously reported in the literature [46,47], although functional implications of such structures have not been yet properly stablished. In experiments using crystal microbalance with dissipation (QCM-D), a role for SP-B/SP-B interactions had been already proposed as a basis to generate contacts between membranes [24], whereas some authors suggested that SP-B aggregates could promote bending of lipid monolayers, according to molecular dynamic simulations [48]. 21

Journal Pre-proof SP-B shows an oligomeric organization in pulmonary surfactant membranes, as well as in lipid model monolayers and bilayers [13]. In the present work it has been also shown that, upon injection of the protein in small volumes of methanol, SP-B is still able to interact with bilayers and adopt its oligomeric structure, which ensures a proper protein activity [13,49]. Transmission electron and atomic force micrographs confirmed that SP-B is assembled as ring-shaped structures, which stand parallel to the surface of lipid layers. According to the 3D model structure based on Saposin B structure, the SP-B oligomer, formed by the association of six dimers, would be projected around 3.7 nm over the lipid layer [13]. Considering this height value, the apposition of two SP-B rings, one on top of the other, would connect two membranes with an interbilayer distance of 7.4 nm, which is in close agreement with the distance revealed by our time-resolved FRET experiments (around 7.5

f

nm). Therefore, the present work confirms that, as previously suggested, SP-B/SP-B interactions mediate

oo

contacts between membranes. Our work provides substantial information concerning the mechanism of generation of multilamellar structures, in which SP-B was supposed to be a key factor. SP-B participates in

pr

vesicle aggregation processes [15] by allowing membrane interactions that deform vesicles, which evolve to the

e-

formation of multilamellar structures [24,50,51] where adjacent membranes would be connected by two SP-B oligomers. This membrane perturbation process would be governed by the combined action of SP-B/SP-B

Pr

interactions and lipid-protein interactions. Due to the large interbilayer distances induced by the protein ( ̴ 7.5 nm), the present time-resolved FRET study could not detect potential stacking of membranes beyond two

al

bilayers. The development of complementary simulation in silico studies, already in progress, could in this sense aid to understand how the spatial organization and the particular geometry of SP-B rings may shape

Jo u

5. Conclusions

rn

membrane-based surfactant arrays at different length scales.

SP-B has been suggested as essential for ensuring an efficient flow of lipid molecules between different surfactant structures and with the associated-interfacial film. This activity has been connected with the ability of SP-B to promote the organization of multilayered networks of surfactant membranes with defined geometries at the nanoscale. The close apposition of membrane layers to an interfacial monolayer induced by SP-B was reported by AFM and QCM-D analysis [21,22,24]. Our time-resolved FRET approach resolved the structure of multibilayer organization induced by SP-B at the nanoscale, showing that the ring-shaped nanostructure of SPB oligomers is essential to induce multilamellar assemblies. Moreover, we have reported for the first time a quantitative topological analysis that supports the apposition of two SP-B ring-shaped oligomers connecting bilayers, which had been previously suggested as an important feature for the optimal function of surfactant [13]. These SP-B/membrane structures would not only be essential for lipid adsorption into the air-liquid interface, but also would be required for allowing the proper biogenesis of lamellar bodies, their unraveling to 22

Journal Pre-proof form diverse extracellular networks and the generation of the multibilayer reservoir of surfactant membranes necessary to stabilize the respiratory surface during breathing dynamics.

oo

f

Acknowledgements

The authors wish to thank the ICTS Centro Nacional de Microscopía Electrónica (U.C.M.) for technical

pr

assistance in AFM analysis. This work was funded by grants from the Spanish Ministry of Science, Innovation and Universities (RTI2018-094564-B-I00), and the Regional Government of Madrid (P2018/NMT-4389).

e-

M.M.-C was a recipient of a fellowship from the Spanish Ministry of Economy and Competitiveness (BES-

Pr

2013-066972). This work was also supported by an EMBO short-term fellowship (7343). L.M.S.L. acknowledges funding by Portuguese “Fundação para a Ciência e a Tecnologia” (FCT) through projects 007630 UID/QUI/00313/2013

and

PT2020-PTDC/DTP-FTO/2784/2014,

M.P.

acknowledges

projects

rn Jo u

FCT (Portugal).

al

FAPESP/20107/2014 and PTDC/BIA-BFS/30959/2017 and A.F. Research Contact IST - ID/105/2018 from

23

f

Journal Pre-proof

[1]

oo

References

J. Goerke, Pulmonary surfactant: functions and molecular composition, Biochim Biophys Acta. 1408

D. Schurch, O.L. Ospina, A. Cruz, J. Perez-Gil, Combined and independent action of proteins SP-B and

e-

[2]

pr

(1998) 79–89.

Pr

SP-C in the surface behavior and mechanical stability of pulmonary surfactant films, Biophys J. 99 (2010) 3290–3299.

L.M. Nogee, Alterations in SP-B and SP-C expression in neonatal lung disease., Annu. Rev. Physiol. 66

al

[3]

[4]

rn

(2004) 601–623.

J.C. Clark, S.E. Wert, C.J. Bachurski, M.T. Stahlman, B.R. Stripp, T.E. Weaver, J.A. Whitsett, Targeted

Jo u

disruption of the surfactant protein B gene disrupts surfactant homeostasis, causing respiratory failure in newborn mice, PNAS. 92 (1995) 7794–7798. [5]

D.K. Vorbroker, S.A. Profitt, L.M. Nogee, J.A. Whitsett, Aberrant processing of surfactant protein C in hereditary SP-B deficiency., Am. J. Physiol. 268 (1995) L647-56.

[6]

M.T. Stahlman, M.P. Gray, M.W. Falconieri, J.A. Whitsett, T.E. Weaver, Lamellar body formation in normal and surfactant protein B-deficient fetal mice, Lab Invest. 80 (2000) 395–403.

[7]

C.D. Foster, P.X. Zhang, L.W. Gonzales, S.H. Guttentag, In vitro surfactant protein B deficiency inhibits lamellar body formation, Am J Respir Cell Mol Biol. 29 (2003) 259–266.

[8]

J. Johansson, T. Curstedt, Molecular structures and interactions of pulmonary surfactant components, Eur J Biochem. 244 (1997) 675–693.

[9]

J.E. Baatz, B. Elledge, J.A. Whitsett, Surfactant protein SP-B induces ordering at the surface of model 24

Journal Pre-proof membrane bilayers, Biochemistry. 29 (1990) 6714–6720. [10]

M.R. Morrow, J. Perez-Gil, G. Simatos, C. Boland, J. Stewart, D. Absolom, V. Sarin, K.M. Keough, Pulmonary surfactant-associated protein SP-B has little effect on acyl chains in dipalmitoylphosphatidylcholine dispersions, Biochemistry. 32 (1993) 4397–4402.

[11]

E.J. Cabré, L.M.S. Loura, A. Fedorov, J. Perez-Gil, M. Prieto, Topology and lipid selectivity of pulmonary surfactant protein SP-B in membranes: Answers from fluorescence, Biochim Biophys Acta Biomembr. 1818 (2012) 1717–1725.

[12]

J. Perez-Gil, C. Casals, D. Marsh, Interactions of hydrophobic lung surfactant proteins SP-B and SP-C

oo

f

with dipalmitoylphosphatidylcholine and dipalmitoylphosphatidylglycerol bilayers studied by electron spin resonance spectroscopy, Biochemistry. 34 (1995) 3964–3971.

B. Olmeda, B. Garcia-Alvarez, M.J. Gomez, M. Martinez-Calle, A. Cruz, J. Perez-Gil, A model for the

pr

[13]

[14]

e-

structure and mechanism of action of pulmonary surfactant protein B, FASEB J. 29 (2015) 4236–4247. F.R. Poulain, L. Allen, M.C. Williams, R.L. Hamilton, S. Hawgood, Effects of surfactant apolipoproteins

[15]

Pr

on liposome structure: implications for tubular myelin formation., Am. J. Physiol. 262 (1992) L730-9. A. Cruz, C. Casals, K.M. Keough, J. Perez-Gil, Different modes of interaction of pulmonary surfactant

R. Chang, S. Nir, F.R. Poulain, Analysis of binding and membrane destabilization of phospholipid

rn

[16]

al

protein SP-B in phosphatidylcholine bilayers, Biochem J. 327 ( Pt 1 (1997) 133–138.

[17]

Jo u

membranes by surfactant apoprotein B, Biochim Biophys Acta. 1371 (1998) 254–264. M.A. Ryan, X. Qi, A.G. Serrano, M. Ikegami, J. Perez-Gil, J. Johansson, T.E. Weaver, Mapping and analysis of the lytic and fusogenic domains of surfactant protein B, Biochemistry. 44 (2005) 861–872. [18]

A. Cruz, L.A. Worthman, A.G. Serrano, C. Casals, K.M. Keough, J. Perez-Gil, Microstructure and dynamic surface properties of surfactant protein SP-B/dipalmitoylphosphatidylcholine interfacial films spread from lipid-protein bilayers, Eur Biophys J. 29 (2000) 204–213.

[19]

M.A. Oosterlaken-Dijksterhuis, H.P. Haagsman, L.M. van Golde, R.A. Demel, Interaction of lipid vesicles with monomolecular layers containing lung surfactant proteins SP-B or SP-C., Biochemistry. 30 (1991) 8276–8281.

[20]

M.A. Oosterlaken-Dijksterhuis, H.P. Haagsman, L.M. van Golde, R.A. Demel, Characterization of lipid insertion into monomolecular layers mediated by lung surfactant proteins SP-B and SP-C., Biochemistry. 25

Journal Pre-proof 30 (1991) 10965–10971. [21]

S. Krol, M. Ross, M. Sieber, S. Kunneke, H.J. Galla, A. Janshoff, Formation of three-dimensional protein-lipid aggregates in monolayer films induced by surfactant protein B, Biophys J. 79 (2000) 904– 918.

[22]

R. V Diemel, M.M.E. Snel, A.J. Waring, F.J. Walther, L.M.G. van Golde, G. Putz, H.P. Haagsman, J.J. Batenburg, Multilayer formation upon compression of surfactant monolayers depends on protein concentration as well as lipid composition: an atomic force microscopy study, J. Biol. Chem. 277 (2002) 21179–21188.

f

A. Cruz, L. Vazquez, M. Velez, J. Perez-Gil, Effect of pulmonary surfactant protein SP-B on the micro-

oo

[23]

and nanostructure of phospholipid films, Biophys J. 86 (2004) 308–320. E.J. Cabre, J. Malmstrom, D. Sutherland, J. Perez-Gil, D.E. Otzen, Surfactant protein SP-B strongly

pr

[24]

dissipation, Biophys J. 97 (2009) 768–776.

L.M.S. Loura, A. Coutinho, A. Silva, A. Fedorov, M. Prieto, Structural effects of a basic peptide on the

Pr

[25]

e-

modifies surface collapse of phospholipid vesicles: insights from a quartz crystal microbalance with

organization of dipalmitoylphosphatidylcholine/dipalmitoylphosphatidylserine membranes: a fluorescent

A. Coutinho, L.M.S. Loura, A. Fedorov, M. Prieto, Pinched Multilamellar Structure of Aggregates of

rn

[26]

al

resonance energy transfer study., J. Phys. Chem. B. 110 (2006) 8130–8141.

4726–4736.

Jo u

Lysozyme and Phosphatidylserine-Containing Membranes Revealed by FRET, Biophys. J. 95 (2008)

[27]

R.P. Haugland, Handbook of fluorescent probes and research products, 9th ed., Molecular Probes, 2002.

[28]

L.B.A. Johansson, A. Niemi, Electronic energy transfer in anisotropic systems. 1. Octadecylrhodamine B in vesicles, J. Phys. Chem. 91 (1987) 3020–3023.

[29]

J. Perez-Gil, A. Cruz, C. Casals, Solubility of hydrophobic surfactant proteins in organic solvent/water mixtures. Structural studies on SP-B and SP-C in aqueous organic solvents and lipids, Biochim Biophys Acta. 1168 (1993) 261–270.

[30]

E.J. Cabre, M. Martinez-Calle, M. Prieto, A. Fedorov, B. Olmeda, L.M.S. Loura, J. Perez-Gil, Homoand hetero-oligomerization of hydrophobic pulmonary surfactant proteins SP-B and SP-C in surfactant phospholipid membranes., J. Biol. Chem. 293 (2018) 9399–9411. doi:10.1074/jbc.RA117.000222.

26

Journal Pre-proof [31]

K. Nag, S.G. Taneva, J. Perez-Gil, A. Cruz, K.M. Keough, Combinations of fluorescently labeled pulmonary surfactant proteins SP-B and SP-C in phospholipid films., Biophys. J. 72 (1997) 2638–2650.

[32]

N.C. Santos, M. Prieto, M.A.R.B. Castanho, Interaction of the major epitope region of HIV Protein gp41 with membrane model systems. A fluorescence spectroscopy study, Biochemistry. 37 (1998) 8674–8682.

[33]

B.W. Koenig, K. Gawrisch, Specific volumes of unsaturated phosphatidylcholines in the liquid crystalline lamellar phase., Biochim Biophys Acta. 1715 (2005) 65–70. D. Marsh, CRC Handbook of lipid bilayers, CRC Press, Boca Raton, FL, US, 2013.

[35]

J.R. Lakowicz, Principles of fluorescence spectroscopy, 3rd ed., Springer, New York, NY, US, 2006.

[36]

L.M. Loura, A. Fedorov, M. Prieto, Resonance energy transfer in a model system of membranes:

oo

f

[34]

[37]

pr

application to gel and liquid crystalline phases, Biophys. J. 71 (1996) 1823–1836. B.W. Van Der Meer, G. Coker III, S.Y. Chen, Resonance Energy Transfer: Theory and Data, VCH, New

D.W. Marquardt, An Algorithm for Least-Squares Estimation of Nonlinear Parameters, J. Soc. Ind. Appl.

Pr

[38]

e-

York, NY, US, 1994.

Math. 11 (1963) 431–441.

L. Davenport, R.E. Dale, R.H. Bisby, R.B. Cundall, Transverse location of the fluorescent probe 1, 6-

al

[39]

diphenyl-1, 3, 5-hexatriene in model lipid bilayer membrane systems by resonance excitation energy

M.A.R.B. Castanho, N.C. Santos, L.M.S. Loura, Separating the turbidity spectra of vesicles from the

Jo u

[40]

rn

transfer, Biochemistry. 24 (1985) 4097–4108.

absorption spectra of membrane probes and other chromophores, Eur. Biophys. J. 26 (1997) 253–259. [41]

R.F.M. de Almeida, L.M.S. Loura, A. Fedorov, M. Prieto, Lipid Rafts have Different Sizes Depending on Membrane Composition: A Time-resolved Fluorescence Resonance Energy Transfer Study, J. Mol. Biol. 346 (2005) 1109–1120.

[42]

R.D. Kaiser, E. London, Determination of the depth of BODIPY probes in model membranes by parallax analysis of fluorescence quenching., Biochim Biophys Acta. 1375 (1998) 13–22.

[43]

M.A. Oosterlaken-Dijksterhuis, M. van Eijk, L.M. van Golde, H.P. Haagsman, Lipid mixing is mediated by the hydrophobic surfactant protein SP-B but not by SP-C., Biochim Biophys Acta. 1110 (1992) 45– 50.

[44]

D.C. Beck, M. Ikegami, C.L. Na, S. Zaltash, J. Johansson, J.A. Whitsett, T.E. Weaver, The role of 27

Journal Pre-proof homodimers in surfactant protein B function in vivo, J Biol Chem. 275 (2000) 3365–3370. [45]

S. Zaltash, W.J. Griffiths, D. Beck, C.X. Duan, T.E. Weaver, J. Johansson, Membrane activity of (Cys48Ser) lung surfactant protein B increases with dimerisation, Biol Chem. 382 (2001) 933–939. doi:10.1515/BC.2001.116.

[46]

H. Bunger, R.P. Kruger, S. Pietschmann, N. Wustneck, L. Kaufner, R. Tschiersch, U. Pison, Two hydrophobic protein fractions of ovine pulmonary surfactant: isolation, characterization, and biophysical activity, Protein Expr Purif. 23 (2001) 319–327.

[47]

N. Wustneck, R. Wustneck, J. Perez-Gil, U. Pison, Effects of oligomerization and secondary structure on

[48]

oo

f

the surface behavior of pulmonary surfactant proteins SP-B and SP-C, Biophys J. 84 (2003) 1940–1949. S. Baoukina, D.P. Tieleman, Computer simulations of lung surfactant, Biochim Biophys Acta. 1858

M. Martinez-Calle, B. Olmeda, P. Dietl, M. Frick, J. Perez-Gil, Pulmonary surfactant protein SP-B

e-

[49]

pr

(2016) 2431–2440.

promotes exocytosis of lamellar bodies in alveolar type II cells., FASEB J. 32 (2018) 4600–4611. J. Bernardino de la Serna, R. Vargas, V. Picardi, A. Cruz, R. Arranz, J.M. Valpuesta, L. Mateu, J. Perez-

Pr

[50]

Gil, Segregated ordered lipid phases and protein-promoted membrane cohesivity are required for

E. Parra, L.H. Moleiro, I. Lopez-Montero, A. Cruz, F. Monroy, J. Perez-Gil, A combined action of pulmonary surfactant proteins SP-B and SP-C modulates permeability and dynamics of phospholipid

Jo u

[51]

rn

535–589.

al

pulmonary surfactant films to stabilize and protect the respiratory surface, Faraday Discuss. 161 (2013)

membranes, Biochem J. 438 (2011) 555–564.

28

Journal Pre-proof Declaration of interests

☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Jo u

rn

al

Pr

e-

pr

oo

f

☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:

29

Jo u

rn

al

Pr

e-

pr

oo

f

Journal Pre-proof

30

Journal Pre-proof HIGHLIGHTS Pulmonary Surfactant Protein SP-B Nanorings Induce the Multilamellar Organization of Surfactant Complexes Marta Martinez-Calle, Manuel Prieto, Bárbara Olmeda, Aleksander Fedorov, Luís L.M.S Loura and Jesús Pérez-Gil.

SP-B generates functional multilayered surfactant films at the alveolar surface.



SP-B induced aggregation of liposomes in vitro.



FRET study showed the structure of SP-B/membrane aggregates at molecular level.



Interaction of two SP-B oligomers mediates the connection between two bilayers.



Membrane stacking by SP-B/SP-B interactions ensures optimal function of surfactant.

Jo u

rn

al

Pr

e-

pr

oo

f



31