Mechanism of Lamellar Body Formation by Lung Surfactant Protein B

Navdar Sever; Goran Miličić; Nicholas O Bodnar; Xudong Wu; Tom A Rapoport

doi:10.1016/j.molcel.2020.10.042

. Author manuscript; available in PMC: 2022 Jan 7.

Published in final edited form as: Mol Cell. 2020 Nov 25;81(1):49–66.e8. doi: 10.1016/j.molcel.2020.10.042

Mechanism of Lamellar Body Formation by Lung Surfactant Protein B

Navdar Sever ^1,², Goran Miličić ^1,², Nicholas O Bodnar ¹, Xudong Wu ¹, Tom A Rapoport ^1,^*

PMCID: PMC7797001 NIHMSID: NIHMS1645341 PMID: 33242393

SUMMARY

Breathing depends on pulmonary surfactant, a mixture of phospholipids and proteins, secreted by alveolar type II cells. Surfactant requires lamellar bodies (LB), organelles containing densely packed concentric membrane layers, for storage and secretion. LB biogenesis remains mysterious, but requires surfactant protein B (SP-B), which is synthesized as a precursor (pre-proSP-B) that is cleaved during trafficking into three related proteins. Here, we elucidate the functions and cooperation of these proteins in LB formation. We show that the N-terminal domain of proSP-B is a phospholipid binding and transfer protein, whose activities are required for proSP-B export from the ER and sorting to LBs, the conversion of proSP-B into lipoprotein particles, and neonatal viability in mice. The C-terminal domain facilitates ER export of proSP-B. The mature middle domain, generated after proteolytic cleavage of proSP-B, generates the striking membrane layers characteristic of LBs. Together, our results lead to a mechanistic model of LB biogenesis.

Graphical Abstract

graphic file with name nihms-1645341-f0001.jpg

eTOC:

Breathing depends on pulmonary surfactant, which is initially stored in lamellar bodies (LB), organelles containing concentric membrane layers. Sever et al. address the molecular mechanism of LB formation. They show that three related domains of surfactant protein B cooperate, with the mature middle domain generating the stacked membrane layers.

INTRODUCTION

Pulmonary surfactant, a mixture of proteins and lipids secreted by alveolar type 2 (AT2) epithelial cells, is critical to facilitate breathing and prevent alveolar collapse (reviewed in Bernhard, 2016; Lopez-Rodriguez et al., 2017). In the alveoli, surfactant forms a thin extracellular layer that drastically reduces surface tension, which enables the lungs to expand, allows modulation of alveolar size, and prevents fluid accumulation in the airways. Surfactant deficiency in preterm infants leads to neonatal respiratory distress syndrome, a condition that is fatal unless treated with intratracheal administration of a surfactant mixture, which is generally obtained from animal sources. Deficiency and impaired function of surfactant also accompanies acute respiratory distress syndrome (ARDS) (Echaide et al., 2017), a severe condition with heterogeneous causes. ARDS is a major cause of lethality in the current COVID-19 epidemic (Wu et al., 2020).

Surfactant is initially stored in AT2 cell-specific organelles, called lamellar bodies (LBs), as concentric stacks of membrane layers (Weaver et al., 2002; Stratton, 1976). LBs are lysosome-related organelles that have an onion-like appearance in thin-section electron microscopy (EM) images. Upon stimulation of exocytosis, the limiting membrane of an LB fuses with the plasma membrane, releasing the internal membrane sheets to the extracellular space, where they form a film with a phospholipid monolayer facing the air. LBs are therefore the essential source of extracellular surfactant and defects in their biogenesis have fatal consequences.

Although LBs have a critical biological role and their morphology has long been appreciated, it is unknown how they are formed. In particular, it is unclear how the characteristic, densely-spaced concentric membrane layers are generated. In other cases of membrane stacking, such as in the Golgi or in myelin sheets, the lipid bilayers are linked together by abundant membrane-bound proteins (Aggarwal et al., 2011; Rabouille and Linstedt, 2016). However, such linker proteins are not present in LBs (Ridsdale et al., 2011). Instead, LBs consist mostly of phospholipids, such as dipalmitoylphosphatidylcholine (DPPC). DPPC is thought to be transported into the organelle by the ATP-binding cassette (ABC) transporter ABCA3 (Matsumura et al., 2007; Li et al., 2019).

In addition to lipids, surfactant contains four proteins (SP-A, SP-B, SP-C, and SP-D), but only SP-B is essential for LB formation, postnatal viability, and breathing (Clark et al., 1995; Nogee et al., 1993; 2000; Stahlman et al., 2000). In the absence of SP-B, AT2 cells contain multi-vesicular bodies (MVBs), i.e. organelles with luminal vesicles. MVBs are therefore thought to be the precursors of LBs. The SP-B gene encodes a precursor protein (pre-proSP-B) that contains a signal sequence and three related domains with conserved disulfide bridges, called N-terminal (SP-BN), middle (SP-BM), and C-terminal (SP-BC) domains (Olmeda et al., 2013) (Figure S1A). The biosynthesis of SP-B begins with its translocation into the endoplasmic reticulum (ER) lumen (Lin et al., 1996a; 1996b). The signal sequence is removed after translocation, and the resulting luminal proSP-B protein is transported in vesicles first to the Golgi apparatus, then to endosomal compartments/MVBs, and finally to LBs. En route to LBs, proSP-B is sequentially processed by several proteases into the three SP-B proteins (Gerson et al., 2008; Korimilli et al., 2000; Ueno et al., 2004).

The functions of the individual SP-BN, SP-BM, and SP-BC proteins are poorly understood. SP-BM is considered to be the mature protein, as it is secreted together with the internal membranes of LBs and is present in extracellular surfactant. In addition, purified SP-BM can reduce surface tension in biophysical in vitro experiments (Smith et al., 1988). SP-BM is extremely hydrophobic, which has hampered its recombinant production and the elucidation of its structure and molecular function. In some species, such as humans, SP-BM is concentrated in an electron-dense region of the LBs, called the “projection core”, from which the membrane stacks appear to emanate (Stratton, 1978). In other species, such as mice, projection cores are not evident and SP-BM is found throughout the membrane stacks (Brasch et al., 2004). Whether SP-BM is directly involved in LB formation is unknown, because SP-BM cannot be targeted to LBs in vivo without being part of a precursor containing the preceding SP-BN domain. The SP-BN domain is indeed thought to play a role in targeting proSP-B to LBs (Lin et al., 1996a; 1996b). Despite its sequence relationship to SP-BM, its role of in the formation of LBs and surfactant has been little-studied. The role of SP-BC is also unclear; this domain is not essential, as a transgene lacking it rescues LB formation and the viability of SP-B knock-out mice (Akinbi et al., 1997).

All three SP-B domains are related in amino acid sequence to the saposins, which are present in the lysosomes of all cells and also made as a precursor. The saposin precursor is proteolytically processed into four proteins (saposins -A, -B, -C, -D) (Bruhn, 2005; Kolter and Sandhoff, 2005; Vaccaro et al., 1999). Crystal structures and biochemical data indicate that each saposin extracts a distinct lipid from luminal vesicles of MVBs and presents it to a specific lipase (Olmeda et al., 2013). Whether any of the SP-B domains can also transfer lipids, given their similarity to saposins, remains unclear.

Here, we elucidate the molecular mechanism by which the SP-B domains cooperate to form LBs. We show that the SP-BN domain functions in the precursor as a phospholipid-binding and -transfer protein, activities that are required for the intracellular sorting of proSP-B and the generation of lipoprotein particles in a low-pH compartment. After proteolytic cleavage, mature SP-BM is sufficient to generate LB-like structures.

RESULTS

SP-BN is a lipid-binding protein

Given that proSP-B is ultimately cleaved into SP-BN, SP-BM, and SP-BC, we decided to first study the structure and function of the purified individual proteins, and then use this knowledge to elucidate the roles of the different domains in intracellular targeting of proSP-B and in generating LBs. We first concentrated on the analysis of SP-BN. SP-BN was fused to N-terminal thioredoxin and a His₆ tag and expressed as a soluble protein in E. coli cells lacking thioredoxin reductase, a strain that allows disulfide bridge formation in the cytosol (Derman et al., 1993). The tags were proteolytically removed, and the protein was further purified by size-exclusion chromatography (SEC) and subjected to crystallization. SP-BN crystals were obtained at pH 5.4, mimicking the acidic environment in LBs (Chander et al., 1986), and diffracted to 2.2-Å resolution. The structure was solved by sulfur single-wavelength anomalous dispersion (SAD) (Table 1). The asymmetric unit contained four copies of SP-BN dimers, each of which had essentially the same structure. Like all saposin-like proteins, the SP-BN monomer consists of four helices with conserved intramolecular disulfide bridges and a loop between helices 2 and 3 (Figure 1A; Figure S1A). In different saposin-like proteins, the four helices form “pocket-knife” structures with open or closed conformations (Olmeda et al., 2013). Our structure is most similar to that of saposin B (Ahn et al., 2003) in terms of conformation of the monomers and their relative arrangement (Figure S1B). The two monomers are in an open conformation with a hydrophobic hollow in between (Figure 1A). The monomers associate through their helices 1 at the “floor”. The two loops between helices 2 and 3 contain additional short helices and form a “roof”. The structure of SP-BN differs from that of all other saposin-like proteins by the presence of three phospholipid (PL) molecules (Figure 1B); these originate from the E. coli cells in which the protein was expressed. Thin-layer chromatography (TLC) (Figure S1C) and mass spectrometry (Table S1) confirmed that the PL composition is the same as that in E. coli, consisting mostly of phosphatidylethanolamine (PE) and phosphatidylglycerol (PG). The PLs seen in each dimer are likely a mixture, explaining why the head groups are not well resolved. Thus, SP-BN appears to be a non-specific phospholipid-binding protein.

Table 1.

Data collection and refinement statistics.

	SP-BN	SP-BN (Y59A/H79A)	SP-BN (K46E/R51E)	SP-BC (low pH)	SP-BC (high pH)
Wavelength	1.771	0.979	0.979	0.979	0.979
Resolution range	70.82 – 2.2 (2.279 – 2.2)	66.46 – 2.315 (2.397 – 2.315)	59.13 – 1.88 (1.947 – 1.88)	51.96 – 1.9 (1.968 – 1.9)	43.33 – 1.755 (1.817 – 1.755)
Space group	P 21 21 21	C 1 2 1	P 1 21 1	P 1 21 1	P 21 21 21
Unit cell	93.707 108.156 114.741 90 90 90	90.335 169.527 73.714 90 126.897 90	49.793 66.124 59.62 90 97.37 90	52.193 79.008 71.917 90 106.428 90	38.396 58.36 64.673 90 90 90
Total reflections	1804081 (66076)	263140 (20487)	1681104 (111407)	150936 (14934)	95769 (7596)
Unique reflections	112655 (7105)	37983 (3077)	31237 (2186)	43545 (4075)	14990 (1393)
Multiplicity	16.0 (6.6)	6.9 (6.0)	53.8 (36.2)	3.5 (3.5)	6.4 (5.4)
Completeness (%)	95.15 (61.16)	97.45 (80.13)	92.19 (70.58)	97.42 (93.14)	98.91 (95.21)
Mean I/sigma(I)	24.77 (0.42)	13.34 (0.72)	40.86 (2.41)	9.54 (1.09)	16.52 (1.35)
Wilson B-factor	56.50	77.25	18.87	36.20	30.31
R-merge	0.2901 (1.905)	0.08439 (2.03)	0.1001 (1.511)	0.1057 (1.102)	0.09072 (1.229)
R-meas	0.2986 (2.065)	0.09118 (2.21)	0.101 (1.531)	0.1252 (1.301)	0.09891 (1.357)
R-pim	0.06995 (0.7777)	0.03419 (0.8571)	0.01347 (0.2406)	0.06629 (0.687)	0.03881 (0.5642)
CC1/2	0.994 (0.236)	0.995 (0.518)	1 (0.897)	0.991 (0.211)	0.998 (0.495)
CC*	0.998 (0.618)	0.999 (0.826)	1 (0.972)	0.998 (0.59)	1 (0.814)
Reflections used in refinement	108835 (7019)	37541 (3077)	28887 (2186)	43042 (4071)	14923 (1392)
Reflections used for R-free	3748 (230)	1826 (156)	1840 (133)	2127 (184)	703 (64)
R-work	0.2237 (0.4367)	0.2411 (0.3947)	0.1695 (0.2980)	0.2105 (0.3532)	0.1773 (0.3236)
R-free	0.2380 (0.4878)	0.2620 (0.3926)	0.2171 (0.3505)	0.2392 (0.3642)	0.2121 (0.3531)
CC(work)	0.943 (0.451)	0.945 (0.591)	0.907 (0.547)	0.947 (0.580)	0.963 (0.788)
CC(free)	0.935 (0.453)	0.967 (0.530)	0.874 (0.340)	0.938 (0.517)	0.955 (0.681)
Number of non-hydrogen atoms	6069	5626	3318	4862	1297
macromolecules	5098	4992	8	4680	1196
ligands	784	603	392	15	1
solvent	187	31	318	167	100
Protein residues	628	629	320	586	150
RMS(bonds)	0.003	0.002	0.006	0.004	0.013
RMS(angles)	0.69	0.64	0.70	0.74	1.04
Ramachandran favored (%)	99.02	95.43	100.00	97.72	97.95
Ramachandran allowed (%)	0.82	4.24	0.00	2.11	2.05
Ramachandran outliers (%)	0.16	0.33	0.00	0.18	0.00
Rotamer outliers (%)	0.34	1.04	0.65	1.55	0.76
Clashscore	2.53	3.86	1.61	4.50	1.72
Average B-factor	84.04	106.28	35.02	49.21	43.84
macromolecules	83.22	103.82	30.63	49.18	43.19
ligands	93.30	127.48	60.58	77.91	33.81
solvent	67.73	90.39	39.49	47.58	51.75
Number of TLS groups	35	33	19	30	10

Open in a new tab

Statistics for the highest-resolution shell are shown in parentheses

Two of the PL molecules in SP-BN are bound laterally inside the cavity and have their head groups on opposite ends of the channel (Figure 1B). The acyl chains of both PL molecules curl around and bind into a pocket of SP-BN. The central PL molecule has acyl chains that extend across the entire length of the hollow (Figure 1B). The occupancy in the crystal structure indicates that the head group of this PL molecule can be on either end of the hollow. Importantly, the acyl chains of the central PL molecule only make contact with the acyl chains of the lateral PL molecules, and not with the SP-BN protein. Thus, the central PL molecule encounters the same environment as in a phospholipid bilayer. The central PL molecule could therefore move with minimal energetic costs from a bilayer to SP-BN. Positively charged amino acid residues facing outwards from the roof may be involved in the interaction with negatively charged lipids that are characteristic of LBs (Figure 1C), and several residues generate the hydrophobic interior of the hollow (Figure 1D). The head groups of the bound PLs seem to be positioned by residues Y59 and H79 at the two ends of the hollow (Figure 1D).

Lipid transfer activity of SP-BN

Because our structure suggested that SP-BN can pick up PLs from membranes, we tested its lipid transfer activity in various assays. We first mixed purified SP-BN with liposomes containing fluorescently labeled PE and floated the vesicles in a Nycodenz gradient (Figure 2A). In the presence of SP-BN, a considerable fraction of the fluorescent lipid was retained in the bottom fractions together with the protein, indicating lipid transfer from the vesicles to the protein. The bottom fractions were then mixed with an excess of unlabeled liposomes and subjected to another round of flotation. This resulted in all the fluorescent lipid floating to the top of the gradient (Figure 2B), indicating lipid transfer from SP-BN to the added liposomes. SDS-PAGE showed that SP-BN does not float with the vesicles (Figure 2A and 2B; bottom panels), indicating that it has only weak affinity for lipid bilayers.

Figure 2. — **(A)** Liposomes with an ER-like PL composition and fluorescent NBD-PE were incubated at pH 5.4 with or without purified SP-BN and floated in a discontinuous Nycodenz gradient. Fractions were analyzed by their fluorescence and by SDS-PAGE and Coomassie-blue staining (lower panel). **(B)** The two bottom fractions of the gradient in (A) were combined, incubated with excess unlabeled liposomes, and re-centrifuged in a discontinuous Nycodenz gradient. **(C)** Liposomes with an LB-like PL composition with NBD-PE (donor fluorophore) and rhodamine-PE (acceptor fluorophore) were incubated with SP-BN at a 7:1 molar lipid to protein ratio and the indicated pH values. NBD fluorescence was followed over time and normalized with respect to fluorescence measured after detergent addition. **(D)** Liposomes with ER- or LB-like PL compositions were incubated with SP-BN for different time periods at 37°C. The zero-time sample was kept on ice for 60 min. All samples were centrifuged through a 100 kDa cut-off filter, extracted with chloroform/methanol, and analyzed by TLC. Lipids were visualized by Primulin staining. Standards were run in parallel. Note that most of PE/PG is derived from *E. coli* and pre-bound to SP-BN. **(E)** As in (C), but with wild-type SP-BN and K46E/R51E mutant and liposomes containing either negatively charged (DPPC/PG) or neutral (DPPC/DOPE) PLs. **(F)** As in (C), but with SP-BN (P50C) containing or lacking a disulfide bridge that locks the “roof”. The reactions were performed at a 4.5 molar ratio of lipid to protein at pH 5.6 with or without DTT. The inset shows SDS-PAGE of the protein with or without reduction with β-mercaptoethanol (β-ME). The arrow and star indicate the dimer and monomer species, respectively. See also Figures S1 and S2.

To follow lipid transfer kinetically, we used fluorescently labeled PLs in a dequenching assay. Emission of 7-nitrobenzoxadiazole (NBD) fluorescence of NBD-PE (donor fluorophore) is initially quenched by rhodamine attached to PE (acceptor fluorophore), but when the fluorescent lipids are transferred to unlabeled liposomes or SP-BN protein, an increase in NBD emission is observed (Bian et al., 2011). Lipid transfer was observed at low pH, but not neutral pH (Figure 2C), consistent with SP-BN functioning in an endosomal compartment.

We developed a filtration assay to determine if SP-BN transfers biologically relevant lipids lacking fluorescent labels. Liposomes were incubated with SP-BN and the mixture was subjected to centrifugation through 100-kDa molecular weight cut-off filters. We found that SP-BN moves into the filtrate, while the liposomes are left behind. The lipids were then extracted from the filtrate with chloroform/methanol and subjected to TLC. With this assay, we found that SP-BN picks up PLs from liposomes with either LB- or ER-like lipid composition (Figure 2D). Little or no lipid transfer was observed when the samples were kept on ice during the incubation (time points “0” in Figure 2D).

Molecular mechanism of phospholipid transfer by SP-BN

To analyze the mechanistic basis of lipid transfer by SP-BN, we applied our fluorescence-dequenching assay. Lipid transfer was observed with liposomes containing negatively charged, but not neutral PLs (Figure 2E; Figure S1D). We infer that this occurs due to an interaction between the negatively charged PL head groups and the positively charged amino acids that protrude from the “roof” of the SP-BN dimer (K46, R51) (Figure 1C), because mutation of these residues to negatively charged amino acids (K46E/R51E) drastically reduced lipid transfer in the fluorescence-dequenching assay (Figure 2E). A crystal structure of this mutant at 1.9-Å resolution was essentially identical to that of wild-type SP-BN (Figure 1E; Table 1), although the mutant protein crystallized in a different space group with only two dimers in the asymmetric unit. The dimer still contained all three PL molecules in approximately the same conformations as the wild-type protein (Figure 1F). Therefore, the K46E/R51E mutant transfers PLs more slowly than the wild-type protein, but binds them in a similar way.

Our structure suggests that PLs enter and exit the hollow of the SP-BN dimer through the roof. We tested this model by mutating P50 at the roof-top to a cysteine, so that the roof would be locked when a disulfide bridge is formed between the loops of the two monomers (Figure 1C). In support of our model, the P50C mutant was indeed inactive in lipid transfer, but became active when the disulfide bridge was reduced with dithiothreitol (DTT) (Figure 2F).

In the crystal structure, three hydrophobic residues (L36, L45, and V80) point into the interior of the PL-binding groove (Figure 1D). When these residues were mutated to charged residues (L36K/L45E/V80K mutant), PLs were no longer bound (Figure S1C) and no lipid transfer was seen in the fluorescence-dequenching assay (Figure S1E). The single L45E mutation had a less dramatic effect, as the protein co-purified with E. coli lipids, but did not transfer lipids in vitro (Figures S1C; S1E). The triple mutant could be purified similarly to the wild-type protein, suggesting that lipid binding is not required for the folding of SP-BN. This is supported by the observation that, upon detergent extraction of bound lipids, the wild-type protein still eluted at the same position in SEC (Figure S1F).

Residues Y59 and H79 are located at the two ends of the dimer hollow and come close to the head groups of the bound PLs (Figure 1D). The Y59A/H79A mutant was even more active than the wild-type protein in lipid transfer (Figure S1G). We determined a crystal structure of this mutant and found that the asymmetric unit contained four copies of the SP-BN dimer (Figure S2A; Table 1), like the wild-type protein. However, in the wild-type all four dimers had six acyl chains in the hollow (two from each PL molecule), whereas the acyl chain number varied in the mutant structure from three to seven (Figure S2B). Some of the acyl chains were also no longer inside the hollow (Figure S2A). These results indicate that Y59 and H79 are required for the correct positioning of the three PL molecules.

Taken together, our experiments indicate that SP-BN is a non-specific phospholipid-binding protein. SP-BN first interacts with negatively charged PL head groups of the bilayer through positively charged amino acids that point outwards from the roof. In the next step, the roof opens and allows PL molecules to exit and enter the hollow, with the central molecule likely favored kinetically. However, our experiments indicate that SP-BN is an unusual lipid transfer protein, as the protein does not seem to act as a catalyst of PL transfer between lipid bilayers; instead very high protein to lipid ratios (at least 1:10 molar ratio) were required. These results suggest that the acceptor for PL transfer may not be another bilayer, but rather another protein domain of SP-B.

The SP-BN domain in proSP-B transfers lipids to form lipoprotein particles

We considered the possibility that SP-BN’s activity might be enhanced when it is present in proSP-B together with the other two domains. To test this idea, we purified proSP-B from HEK293 cells. The protein was stably expressed as a fusion with an N-terminal HALO-tag and secreted into the culture media. After removal of the HALO-tag, the purified proSP-B protein migrated as a single band in SDS-PAGE (Figure 3A) and contained primarily PC (Figure 3B). The molar lipid to protein ratio was about the same as with isolated SP-BN (1–2 molecules per SP-BN monomer). Strikingly, at low pH, purified proSP-B exhibited the same activity as SP-BN when present at ~10-fold lower concentration (Figure 3C). This lipid transfer activity was reduced at neutral pH (Figures 3D) and with liposomes containing ER-like, instead of LB-like, PLs (Figure 3E). To prove that the lipid transfer activity of proSP-B is due to the SP-BN domain, we purified mouse proSP-B containing the K105E, R110E mutations (equivalent to K46E/R51E in SP-BN). These mutations indeed reduced lipid transfer activity, as in the isolated SP-BN domain (Figure 3F). Taken together, these results indicate that the SP-BN domain in proSP-B is activated for lipid transport when the precursor arrives in a low-pH compartment.

Figure 3. — **(A)** Coomassie-stained SDS-PAGE of purified human proSP-B. ProSP-B was secreted as a HALO-tagged fusion protein from HEK293 cells, purified with a HALO resin, and the tag was cleaved off and removed. **(B)** Lipids bound to purified proSP-B were extracted with chloroform/methanol and separated by TLC. A mock sample was treated analogously. The question mark indicates unidentified lipids. **(C)** Purified proSP-B or SP-BN was tested at pH 5.6 for lipid transfer with the fluorescence de-quenching assay. The PL composition of donor and acceptor vesicles resembled that of LBs and the molar protein to lipid ratio (P/L) was varied, as indicated. The data were normalized to fluorescence measured after addition of detergent. **(D)** As in (C), but at pH 7.4. **(E)** As in (C), but with liposomes that have an ER-like lipid composition. **(F)** As in (C), but with mouse proSP-B or SP-BN harboring the indicated mutations. These experiments were done in parallel with those in (C).

Because proSP-B has a higher lipid transfer activity than SP-BN in isolation, we wondered whether the other SP-B domains serve as lipid acceptors, generating lipoprotein particles, in which each proSP-B molecule is associated with a micelle of phospholipids; such lipoprotein particles could then serve as precursors for the stacked lipid bilayers in LBs. To test for the formation of lipoprotein particles, we incubated purified proSP-B with preformed liposomes containing a LB-like PL composition and fluorescent lipid, and subjected the mixture to SEC. In the absence of liposomes proSP-B eluted mostly as a dimer, but after incubation with liposomes at low pH and 37°C, a sizable fraction eluted at a size between that of the protein alone and the original liposomes (Figure 4A). This population also contains PLs, as shown by the absorbance of the fluorescent lipid at 560 nm (Figure 4A). We determined the amount of fluorescent lipid in these fractions by either measuring the absorbance at 560 nm or the fluorescence in a plate reader. The amount of proSP-B in these fractions was determined on the basis of a comparison with a titration of purified proSP-B in immunoblots. Together, these data indicate that each proSP-B monomer binds 15–25 PL molecules, ~10 times more than the isolated SP-BN protein (1.5 molecules per monomer). This is consistent with the fact that 10-fold higher protein to lipid ratios are required for SP-BN to show the same activity as proSP-B in the fluorescence dequenching assay (Figure 3C). Lipoprotein particle formation was not seen at neutral pH (Figure 4B), again consistent with proSP-B acting in an endosomal compartment. Furthermore, when the incubation was performed at 0°C, i.e. under conditions where lipid transfer is reduced, some proSP-B eluted together with the liposomes, indicative of binding, but a distinct lipoprotein peak was not observed (Figure 4C).

Lipoprotein particle formation was also demonstrated by native agarose gel electrophoresis, which separates molecules based on size and charge (Figure 4D). When purified proSP-B was incubated with liposomes at low pH and 37°C, a defined band with similar mobility as very-low-density lipoprotein (VLDL) was observed. In the absence of protein or at neutral pH, only diffuse bands were seen. ProSP-B harboring mutations that affect its lipid transfer activity (K105E, R110E; Figure 3F) formed lipoprotein particles less efficiently.

Negative-stain EM showed that the lipoprotein-containing SEC fraction consisted of particles with a variable diameter (~20–35 nm) (Figure 4E). Their size and rugged surface confirm that they are not vesicles with a normal phospholipid bilayer. The SEC data indicate that the particles have a molecular mass of ~0.6 million Daltons, which means that they contain approximately 12 molecules of proSP-B and 180 molecules of PLs. These lipoprotein particles were not seen with liposomes alone (Figure 4F) or with the proSP-B dimer fraction observed in the absence of liposomes (Figure 4G). Taken together, these results indicate that lipid transfer by the SP-BN domain of proSP-B results in the formation of lipoprotein particles in a low-pH compartment.

Lipid-binding and -transfer by the SP-BN domain determines intracellular trafficking of proSP-B

Our results raised the possibility that the lipid-binding and -transfer activities of the SP-BN domain are required for the intracellular sorting of proSP-B to LBs. We therefore tested proSP-B trafficking in an established mouse alveolar epithelial cell line (MLE-15). These cells express endogenous SP-B and accumulate mature SP-BM in LBs, but do not secrete surfactant (Wikenheiser et al., 1993). Antibodies raised against mouse SP-BN detected endogenous proSP-B, the precursor of SP-B lacking the signal sequence, both in the cellular membrane fraction and in the culture media (Figure 5A, lanes 1 and 9) (the latter is a protein population that was not diverted from the secretory pathway to LBs). Mature SP-BN was only detected inside the cells and mature, endogenous SP-BM was undetectable in most experiments (lanes 1 and 9). Transient over-expression of a C-terminally tagged version of mouse SP-B (mSP-B-Myc-Flag) again showed proSP-B (which runs slower in SDS gels because of the tag) both in cells and in the media (Figure 5A, lanes 2 and 10). Mature SP-BM was now detectable and found only inside the cells (lanes 2 and 10). Similar results were obtained when untagged human SP-B (hSP-B) was transiently expressed (lanes 3 and 11). These results show that exogenously expressed wild-type proSP-B is correctly processed in MLE-15 cells.

To avoid confusion with endogenous SP-B, we generated a SP-B knock-out MLE-15 cell line by CRISPR-Cas9 (Figure S3A; Figure 5A, lanes 5 and 13). Transient expression of mSP-B-Myc-Flag or hSP-B in the knock-out (KO) cells again showed the presence of proSP-B protein in cells and media, and mature SP-BM only inside the cells (Figure 5A, lanes 6, 7 and 14, 15). Intracellular proSP-B contained glycans sensitive to endoglycosidase H (endo H) treatment (Figure S3B), indicating its ER localization, whereas secreted proSP-B was endo H resistant (Figure S3B). ER localization of intracellular proSP-B is supported by immunostaining with an antibody that exclusively recognizes proSP-B (Figure 5B) and by co-localization with the ER marker calreticulin (Figure S3C). A considerable fraction of the ER-localized population was soluble in Triton X-100 and thus not grossly misfolded (Figure S3B). Mature hSP-BM was found in puncta (Figure 5B) and fractionated to a light membrane fraction in sucrose gradient centrifugation experiments (Figure S3D), consistent with its localization to LBs. The protein was also extractable in chloroform/methanol (Figure S3E), like SP-BM from lungs (Mathialagan and Possmayer, 1990). Thus, a fraction of wild-type proSP-B stays in the ER, another population is secreted, and yet another population is proteolytically processed and transported to LBs.

We next introduced three mutations into the hSP-B precursor that abrogated lipid binding in isolated SP-BN (L36K/L45E/V80K; corresponding to L96K/L105E/L140K in human proSP-B). The triple mutant protein was found inside the cells, but not in the media (Figure 5A, lanes 4, 8 and 12, 16). Mature SP-BM was not detected (Figure 5A, lanes 4 and 8). Endo H sensitivity and immunostaining confirmed that mutant proSP-B remained in the ER and was not processed to SP-BM (Figure 5B; Figures S3B; S3C). ER retention was also observed with proSP-B carrying only one of the three mutations (V80K in SP-BN, corresponding to V139K in mouse proSP-B) (Figure 5C, lanes 5 and 11). In contrast, the single-mutant equivalent to L36K (L95K in mouse proSP-B) behaved like the wild-type protein (Figure 5C, lanes 3 and 9). The P109C mutant (P50C in SP-BN), a mutation that greatly reduced lipid transfer activity of the isolated SP-BN protein (Figure 2F), behaved similar to the L96K/L105E/L140K and L140K mutants (Figure 5C, lanes 2 and 8). All mutant proSP-B proteins were expressed at levels similar to wild-type proSP-B and their intracellular population was extractable with TX-100 or alkali (Figure 5C), indicating that the mutations did not cause aggregation or membrane integration. Although the mutations P109C or L139K prevented ER export of full-length proSP-B (Figure 5C, lanes 2 versus 8 and 5 versus 11), they allowed export if both the SP-BM and SP-BC domains were deleted (constructs labeled ΔMC) (Figure 5D, lanes 3 versus 9 and 5 versus 11). Taken together, these results show that a defect in lipid binding to the SP-BN domain, rather than its misfolding, causes retention of proSP-B in the ER.

Next, we introduced mutations into proSP-B, which reduce lipid-transfer, but not lipidbinding, in the isolated SP-BN protein (Figures 1E; 2E). A mouse proSP-B mutant equivalent to K46E/R51E in SP-BN (K105E/R110E in mSP-B precursor) was localized to the ER and secreted into the media, but was only inefficiently processed into mature SP-BM (Figure 5C, lanes 6 and 12; reduced to ~10% of wild-type levels). A similar phenotype was observed with the L45E mutant (L104E in mouse proSP-B) (Figure 5C; lanes 4 and 10; Figure S1E). In both cases, the mutations are at the roof of the SP-BN dimer. Thus, efficient lipid transfer activity by the SP-BN domain is required for the targeting of proSP-B to the LB pathway and its proteolytic processing. Given that lipid transfer causes lipoprotein particle formation in a low-pH compartment (Figure 4), our results suggest that lipoprotein particles are an intermediate in the processing of proSP-B.

Testing proSP-B mutants in mice

To confirm the role of SP-BN’s lipid-binding and -transfer activity in vivo, we used CRISPR-Cas9 directed editing to generate mice carrying mutations in the SP-BN domain. We first generated mice with three mutations in the SP-BN domain (L95K/L104E/V139K, equivalent to L36K/L45E/V80K in SP-BN) (Figure 6A), which abrogated all SP-BN lipid binding in vitro and caused proSP-B to be retained in the ER of MLE-15 cells. The embryos were isolated one day before birth. Three embryos carried the three mutations in both alleles, as shown by PCR and DNA sequencing. Analysis by SDS-PAGE and immunoblotting showed that all three mice made proSP-B, but not mature SP-BM (Figure 6B, lanes 1,2,6). As expected, both proSP-B and mature SP-BM were detected in wild-type embryos (lanes 3,4) and these proteins were absent in embryos carrying insertions or deletions in both alleles (lanes 5,7). Thin-section EM showed that the lungs of the triple-mutant embryos lacked normal LBs (Figure 6D), like knock-out embryos (Figure 6E; controls shown in Figure 6C).

To test for viability, we repeated the CRISPR experiment and analyzed the genotypes of newborn mice (Table S2). Consistent with the gross defects in LB formation, triple-mutant mice died shortly after birth from respiratory failure, like knock-out animals (Clark et al., 1995). Thus, lipid binding to the SP-BN domain is required for the processing of proSP-B and the generation of functional protein in vivo.

Next we used CRISPR-Cas9 editing to generate mice carrying the K105E/R110E mutations, equivalent to K46E/R51E in isolated SP-BN (Figure 6A). These mutations compromise lipid transfer, but not lipid binding, of the SP-BN protein in vitro, and reduce proteolytic processing of proSP-B, but not its export from the ER, in MLE-15 cells. In two embryos with the expected knock-in mutations in proSP-B (Figure 6F), proteolytic processing of proSP-B was reduced to ~10% of wild-type levels (Figure 6F). Similar results were obtained with two knock-in mice that contained only one of the two desired mutations (K105E; corresponding to K46E in SP-BN; Figure 6F). Thus, as in MLE-15 cells, lipid transfer activity of the SP-BN domain is required for efficient targeting of proSP-B to the LB pathway. Interestingly, however, all mutant embryos contained normal LBs (Figures 6G; 6H). Accordingly, viable mice were obtained that contained the double mutation in one allele and a deletion/frame-shift in the other, the single K105E mutation in one allele and a deletion in the other, or the double mutation in one allele and the single K105E mutation in the other (Table S3). These mice were fertile and gave rise to viable offspring expressing exclusively the single or double mutant protein. We conclude that the low level of proSP-B processing is sufficient to sustain LB formation in vivo and neonatal viability.

The SP-BC domain facilitates ER export of proSP-B

Next we analyzed the function of the SP-BC domain, which is not essential for LB formation or breathing (Akinbi et al., 1997). Again, we first characterized the isolated protein. SP-BC was purified in the same way as SP-BN, but did not contain any bound lipids. In our assays, it also did not bind phospholipids, cholesterol, lyso-PC, fatty acids, or lipopolysaccharide. Crystal structures of SP-BC were obtained at pH 8.5 and pH 4.5, diffracting to 1.75 Å and 1.90 Å, respectively (Table 1). In the pH 8.5 structure, SP-BC forms a dimer, in which the two monomers interact through straight helix pairs and are oriented in a parallel manner (Figure S4A). The monomers are in a more open conformation than in any saposin structure and do not display a lipid binding surface, explaining why SP-BC did not bind lipid in our assays. In the pH 4.5 structure, the monomers are also in a wide-open conformation, but they are arranged in a dimer in an anti-parallel manner (Figure S4B). Two of the dimers seem to form a tetramer with a moderately hydrophobic pocket.

We tested the physiological role of SP-BC by following the intracellular trafficking of mutant proSP-B in MLE-15 cells. When the SP-BC domain was deleted (ΔC), only small amounts of proSP-B were secreted or proteolytically processed into mature SP-BM (Figure 5E; lane 3 versus 9). Similar results were obtained when the constructs carried HALO tags at their N-termini (Figure 5E; lane 4 versus 10 and 5 versus 11). All intracellular protein was endo H sensitive and as soluble in Triton X-100 as wild-type proSP-B (Figure S3B), indicating that the mutants were located in the ER, but not aggregated. Thus, the SP-BC domain is required for efficient ER export of proSP-B. Since the ER lumen has a neutral pH, these results also argue against the physiological significance of the tetramer seen in our low-pH structure.

To further test the role of SP-BC, we introduced point mutations into this domain in proSP-B. Most of the mutations reduced ER export of proSP-B and the generation of mature SP-BM (Figure 5F; lanes 7–10). However, one mutant (T348K), designed to disrupt the tetramer seen in our low-pH structure (Figure S4B), behaved like wild-type (Figure 5F; lane 7), again suggesting that tetramerization is not required. A parallel arrangement of the monomers, as observed in our high-pH dimer structure, is also more consistent with dimer formation of proSP-B (Figure 4A) and the parallel arrangement of monomers in our SP-BN structures (Figure 1). How exactly the SP-BC domain facilitates ER export of proSP-B remains to be clarified, but the small fraction that is still correctly processed (as with the K105E/R110E mutant) likely explains why a transgene lacking SP-BC rescues SP-B knock-out mice (Akinbi et al., 1997).

Reconstituted SP-BM generates LB-like structures

Finally, we analyzed the function of SP-BM. Despite its similarities to SP-BN and SP-BC, SP-BM could not be purified by the same protocol, likely because it is extremely hydrophobic; it associates with membranes in an alkali-resistant manner (Figure 5C). A structure of SP-BM is not yet available, but based on its behavior as an integral membrane protein, we propose a model in which each monomer has a large hydrophobic surface facing the acyl chains of phospholipids (Figure S4C). The predicted hydrophobic surface seems to be important, as mutations that render it more hydrophilic prevent ER exit of proSP-B in MLE-15 cells (Figure 5F; lane 4). On the other hand, mutation of several residues proposed to be involved in a peripheral membrane interaction of SP-BM (Olmeda et al., 2015; Martinez-Calle et al., 2020) did not drastically affect ER export (Figure 5F; lane 5; processing to SP-BM could not be tested because the mutations abolished recognition by the antibodies).

Next we tested the effect of SP-BM on the structure of membranes. We initially purified SP-BM from bovine lungs, taking advantage of its unusual property of being soluble in chloroform/methanol (Mathialagan and Possmayer, 1990). After removal of phospholipids, the protein was further purified by SEC in the detergent octylglucoside (Figure S5A). Non-reducing SDS-PAGE of the purified protein showed a single band corresponding in size to a dimer (Figure S5A), as expected from SP-BM monomers being disulfide-linked through the odd cysteine at position 48 (Beck et al., 2000). Purified SP-BM dimer was then mixed with a dried PL mixture consisting of DPPC and PG, the major lipids of LBs (King, 1982). The detergent was removed by dialysis, and the proteoliposomes were sedimented and analyzed by osmium staining and thin-section EM. Strikingly, we observed structures that resemble LB images obtained from human lung sections (Figure 7B versus 7A): they showed electron-dense areas, similar to projection cores, from which bubbles of membrane sheets emerge (Figure 7B). The membrane sheets are not directly apposed, indicating that they are not linked by protein, as in the case of myelin sheets or Golgi stacks. Because calcium plays important roles in the structure and function of pulmonary surfactant (Benson et al., 1984), we also added Ca²⁺ after dialysis; this caused the membrane sheets to stack more tightly (Figure 7C). With increasing protein to lipid ratios, more electron-dense areas with fewer associated membrane bubbles were observed (Figure S5B), consistent with the electron-rich regions containing SP-BM. Electron-dense areas with emanating membrane bubbles were scarce or absent when SP-BM was reconstituted with a lipid mixture mimicking the ER composition (Figure S5C). When protein was omitted, the membrane pellet was much smaller, probably because lipid was lost during dialysis.

Figure 7. — **(A)** Thin-section EM of a human LB (reproduced with permission from (Stratton, 1978)). PC, projection core; MC, matrix core. **(B)** Liposomes reconstituted from detergentsolubilized bovine SP-BM (10% by weight) and phospholipids were analyzed by thin-section EM. **(C)** As in (B) but in the presence of 5 mM CaCl₂. (D) As in (C), but with recombinant SP-BM. Two examples are shown. (E) Purified proSP-B was incubated with liposomes to form lipoprotein particles. The sample was incubated with proteinase K to generate the individual SP-B domains. After centrifugation, the membrane pellet containing mature SP-BM was analyzed by thin-section EM. (F) Scheme for the function of SP-B in LB formation. N, M, and C denote SP-BN, SP-BM, and SP-BC. See also Figures S5, S6, and S7.

To exclude the possibility that the LB-like structures were generated by a contaminant of the SP-BM preparation from lungs, we generated SP-BM recombinantly. Our purification procedure is based on the E. coli expression of a fusion of glutathione S-transferase (GST) and human proSP-B lacking the SP-BC domain (Figure S6A). The fusion protein was solubilized from inclusion bodies, the SP-BM domain was cleaved off with a specific protease, extracted into chloroform/methanol, and purified by SEC in octylglucoside (Figure S6B–D). CD spectroscopy showed that the protein is folded (Figure S6E), and mass spectrometry and immunoblotting with SP-BM antibodies confirmed the identity of the purified protein (Figures S6F; S6G). After reconstitution into liposomes with LB-like lipids, the dimer fraction obtained by SEC (Figure S6G; fraction #8) generated structures similar to those obtained with SP-BM purified from bovine lung, i.e. projection cores with emanating membrane bubbles (Figure 7D). In contrast, the higher molecular weight fractions (Figure S6G; fractions #3/4) gave rise to vesicles (Figure S7A; left panel), and the lower molecular weight fractions (Figure S6G; fractions #12/13) to multi-lamellar structures without projection cores (Figure S7A; right panel). While proSP-B generated lipoprotein particles (Figure 4), it did not form projection cores with emanating membrane bubbles when reconstituted in an analogous way as SP-BM into liposomes containing either an LB- or ER-like lipid composition (Figure S7B). Thus, proteolytic processing of proSP-B into mature SP-BM seems to be required to generate the membrane stacks characteristic of LBs.

To directly recapitulate proSP-B to SP-BM conversion in vitro, we first incubated purified proSP-B with liposomes to generate lipoprotein particles (see Figure 4). The sample was then treated with proteinase K as a surrogate for the normal processing proteases. Analysis of the samples by reducing and non-reducing SDS-PAGE showed that the generated mature SP-BM forms a dimer of disulfide-linked monomers (Figure S7C), as observed in vivo (Beck et al., 2000). Thus, the intermolecular disulfide bridge between SP-BM monomers forms spontaneously after proteolytic cleavage of proSP-B. Centrifugation showed that the mature SP-BM protein associates with membranes, whereas the mature SP-BN protein stayed in the supernatant (Figure S7D). Analysis of the membrane pellet by thin-section EM revealed that the membrane pellet contained numerous electron-dense areas from which membrane sheets emanate (Figure 7E). These structures are similar to those formed with purified SP-BM from natural or recombinant sources (Figures 7B–D), but are generated without detergent. Thus, our ability to reconstitute SP-B processing in vitro shows that cleavage of proSP-B into SP-BM causes lipoprotein particles to transition into LB-like membrane structures.

DISCUSSION

Here, we elucidate the molecular mechanism by which LBs are formed in alveolar type II cells. We show that the three SP-B domains cooperate in LB formation, but have distinct roles, despite their sequence similarities. The three domains initially function together in proSP-B, the precursor that is generated from pre-proSP-B by signal sequence cleavage in the ER lumen. The packaging of proSP-B into vesicles that bud from the ER requires lipid binding to the SP-BN domain and is facilitated by the SP-BC domain. SP-BN probably acts by recruiting phospholipids (PLs) that shield the hydrophobic surface of the SP-BM domain in the ER (Figure 7F, stage 1). The structure formed by proSP-B and PLs in the ER might already be a small lipoprotein particle, given that the predicted hydrophobic surface of SP-BM is required for ER exit of proSP-B. The process would be similar to the biogenesis of VLDL particles, where the microsomal triglyceride transfer protein moves triglycerides and other lipids from the ER bilayer to newly synthesized apolipoprotein B (Olofsson et al., 1999; Sirwi and Hussain, 2018), a process required for ER export of VLDL.

After ER exit, proSP-B is transported in vesicles to the Golgi apparatus, from where it reaches a low-pH compartment, likely endosomes or MVBs (Figure 7F, stage 2). The low pH activates PL transfer by the SP-BN domain, resulting in the formation of larger lipoprotein particles. Lipid transfer and thus lipoprotein formation are required for the sorting of proSP-B to the LB pathway. The SP-BN domain is an unusual lipid-binding and -transfer protein, as the acyl chains of the centrally bound PL molecule interact exclusively with the lateral PLs and therefore have essentially the same environment as in a lipid bilayer. In contrast, most other lipid transfer proteins bind the lipid in a proteinaceous pocket or tube (Wong et al., 2019). Furthermore, the SP-BN domain can transfer PLs to the SP-BM domain, while cytosolic lipid transfer proteins generally transfer lipids between bilayers of different organelles (Wong et al., 2019). In the lipoprotein particles generated by lipid transfer of the SP-BN domain, the acyl chains of the PLs likely interact with the hydrophobic face of the SP-BM domain (Figure 7F; stage 2). The major source of the PLs is the limiting membrane (Figure 7F, stage 2); the abundant surfactant lipid DPPC is flipped by the ABC transporter ABCA3 from the outer to the inner leaflet of the lipid bilayer (Matsumura et al., 2007), from where it can be extracted by proSP-B. This mechanism is similar to that proposed for the formation of high-density lipoprotein (HDL) particles. Here, the ABC transporter ABCA1 pumps PLs and cholesterol from the inner to the extracellular leaflet of the plasma membrane, and apolipoprotein A1 picks up the lipid and forms a bilayer within the discoidal HDL (Phillips, 2018).

Finally, following proteolytic cleavage of proSP-B, the mature SP-BM spontaneously dimerizes through formation of an intermolecular disulfide bridge, and generates the concentric lipid bilayer stacks characteristic of LBs (Figure 7F, stage 3). The lipoprotein particles serve as nucleation sites for the generation of the stacked bilayers. During the transition from lipoprotein particles to stacked membrane sheets, SP-BM remains in contact with the acyl chains of PLs and eventually localizes to the edges of bilayer discs (Figure 7F; stage 3). The postulated localization of SP-BM to bilayer edges is based on the fact that it behaves like an integral membrane protein and on properties of the related protein saposin A. Like proSP-B, saposin A forms lipoprotein particles (Li et al., 2016). In a crystal structure with the detergent lauryldimethylamine oxide (LDAO), two saposin A molecules with an open conformation, resembling an open “pocket-knife”, localize to the hydrophobic edges of a small, lipid-bilayer like nanodisc (Popovic et al., 2012). SP-BM is indeed predicted to have an extensive hydrophobic surface inside its “pocket-knife” structure (Figure S4C). In the proposed model, SP-BN and SP-BM function together to generate LBs de novo, explaining the lack of LBs in SP-B mutants. Although SP-BN acts as part of proSP-B, it is possible that the mature SP-BN protein also plays a role after cleavage of proSP-B, as suggested by its anti-microbial activity (Yang et al., 2010) and by the fact that it has lipid transfer activity in our assays. It should be noted that our reconstitution experiments were done with the major PLs of LBs and omitted other lipids, such as cholesterol (~5–8% of total lipids), because they are thought to play only a minor role (Diemel et al., 2002; Tanaka et al., 1986).

The existence of a protein-rich projection core in human LBs and in our reconstitutions suggests that SP-BM molecules can also self-assemble, consistent with the reported oligomerization of SP-BM (Olmeda et al., 2015; Martinez-Calle et al., 2020). In our model, the hydrophobic surface of each monomer faces the acyl chains of PLs (Figure 7F), whereas in the proposed alternative, the hydrophobic surfaces of two monomers would face each other, and the oligomer would form a ring that binds to the periphery of two membranes and connects them (Olmeda et al., 2015; Martinez-Calle et al., 2020). Our model is more consistent with the observation that membrane sheets are not linked by SP-BM in reconstituted LB-like structures, and that SP-BM can be extracted only with detergents or organic solvents, but not with alkali. SP-BM oligomerization within the bilayer would serve to anchor multiple bilayer discs at the same site, i.e. the projection core (Figure 7F, stage 3). In some species, such as mice, projection cores have not been observed and SP-BM is distributed throughout the membrane stacks (Brasch et al., 2004). Assuming that the mechanism of LB formation is conserved, the projection cores in these species might be smaller or initially formed and then dissolved. They could be dissolved if SP-BM formed diffusible oligomers that are completely surrounded by acyl chains of PLs, structures that may also be adopted by the saposin-like, poreforming proteins granulysin, NK-lysin, and amoebapore (Olmeda et al., 2013). In previous reconstitution experiments, SP-BM formed bilayer discs, rather than concentric membrane stacks (Poulain et al., 1992; Suzuki et al., 1989; Williams et al., 1991), perhaps because the protein was reconstituted directly from organic solvent. Bilayer discs are consistent with our model assuming that they are precursors to membrane stacks and that SP-BM localizes to their edges. Regardless of the exact mechanism of bilayer stacking, it is striking that the internal shape of LBs can be generated with SP-BM alone. The only other organelle that has been reconstituted with purified proteins is the tubular ER network (Powers et al., 2017).

Our improved understanding of the biological functions of SP-B domains, as well as the ability to produce the SP-B proteins recombinantly in E. coli, paves the way to generate large amounts of effective surfactant mixtures for the treatment of ARDS, a disease that has not been treatable with the available commercial surfactants (Echaide et al., 2017).

STAR * METHODS

RESOURCE AVAILABILITY

Lead Contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Tom Rapoport (tom_rapoport@hms.harvard.edu).

Materials Availability

All unique/stable reagents generated in this study are available from the Lead Contact with a completed Materials Transfer Agreement.

Data and Code Availability

The X-ray diffraction datasets generated during this study are available at SBGrid (https://sbgrid.org): DOIs 10.15785/SBGRID/757, 10.15785/SBGRID/758, 10.15785/SBGRID/759, 10.15785/SBGRID/760, 10.15785/SBGRID/761, 10.15785/SBGRID/762, 10.15785/SBGRID/763, 10.15785/SBGRID/773, 10.15785/SBGRID/768, 10.15785/SBGRID/769, 10.15785/SBGRID/770, 10.15785/SBGRID/771, 10.15785/SBGRID/766, and 10.15785/SBGRID/767. The structures generated during this study are available at the Protein Data Bank (PDB; https://www.rcsb.org/): PDB IDs 6VZ0, 6VZE, 6VYN, 6VZD, and 6W1B. The unprocessed gels (Coomassie, immunoblotting, agarose), TLC plates and microscopy images (fluorescence and EM) have been deposited to Mendeley Data: http://dx.doi.org/10.17632/s7w5tw5tvp.1

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Cells

Rosetta-Gami 2(DE3) Escherichia coli (E. coli) was grown in LB (10 g/L tryptone, 5 g/L yeast extract, 5 g/L NaCl) medium. BL21(DE3) E. coli was grown in 2xYT (16 g/L tryptone, 10 g/L yeast extract, 5 g/L NaCl) medium. Large cultures (> 1 L) were grown in Erlenmeyer flasks at 37°C and 300 rpm on an orbital shaker, and OD₆₀₀ measurements were taken using a 1-cm plastic cuvette.

Murine lung epithelial (MLE-15, female) and HEK293T (female) cells are maintained in liquid nitrogen and early passage aliquots are thawed periodically. MLE-15 cells are maintained in HITES medium supplemented with 2% fetal bovine serum (FBS) at 37°C (Wikenheiser et al., 1993). HEK293T cells are maintained in DMEM supplemented with 10% FBS at 37°C.

Mice

Female C57BL/6 (3–4 weeks old) and CD1 mice (8–10 weeks old) were purchased from Envigo. All mice were housed in accordance with Institutional Animal Care and Use Committee guidelines.

METHOD DETAILS

Cloning

Cloning of sgRNAs into pX330 was performed following a modified single-step digestion-ligation protocol from the Zhang lab, available online at genome-engineering.org. Synthetic SP-BN and SP-BC codon-optimized for E. coli were cloned into pET-32a(+) using BamHI and EagI sites. Human proSP-B lacking the signal sequence and the SP-BC domain (amino acids 24–279) was cloned into pGEX-6P-1 using EcoRI and NotI sites, and a thrombin cleavage site (LVPRGS) was introduced between residues 201–202 by QuikChange Site-Directed Mutagenesis Kit (Agilent). To introduce point mutations into human proSP-B, the commercial pCS6-SFTPB vector (transOMIC) was first used to amplify proSP-B ORF without the polyA tail, which was inserted back into the same vector backbone using EcoRI and HindIII sites. gBlocks (IDT) encompassing the SP-BM domain and including either L214E/V231E/Y253E/V270E or W209A/R212A/K216A/R217A/R236A/R272A mutations were used to replace the corresponding region of proSP-B using NEBuilder HiFi DNA Assembly (NEB). All other point mutations were introduced by QuikChange Site-Directed Mutagenesis (Agilent). Human proSP-B lacking the SP-C domain (amino acids 1–279) with a C-terminal Flag tag was cloned into pcDNA3.1+ using NEBuilder HiFi DNA Assembly (NEB). Human proSP-B lacking both SP-BM and SP-BC domains (SP-ΔMC; amino acids 1–200) was cloned into pCS6 using EcoRI and HindIII sites. Human proSP-B (24–381) was cloned into the lentiviral pHAGE2 vector using EcoRI and NotI sites. Mouse proSP-B (24–382) containing K105E/R110E mutations was cloned into pHAGE2 using BamHI and NotI sites. All mutations were verified by DNA sequencing.

Protein purifications

SP-BN and SP-BC were expressed with N-terminal His₆ and thioredoxin tags, followed by a Tobacco Etch virus (TEV) protease cleavage site (ENLYFQS). The fusion proteins were expressed in E. coli cells lacking thioredoxin reductase (Rosetta-Gami 2 (DE3) cells) from an isopropyl β-D-1-thiogalactopyranoside (IPTG)-inducible promoter of the pET-32a(+) plasmid. The cells were grown to mid log phase in the presence of 100 μg/ml ampicillin and induced with 1 mM IPTG at 16°C overnight. Cell pellets from a 6 L culture were homogenized in ~100 ml 20 mM Tris/HCl, pH 8, 300 mM NaCl, 20 mM imidazole, 10% glycerol and protease inhibitors. The extract was centrifuged for 45 min at 40,000 g in a Beckman Ti45 rotor. The supernatant was incubated with 5 ml Ni-NTA beads for 1 hr, and the protein was eluted with 300 mM imidazole in the same buffer. The sample was supplemented with 5 mM glutathione (GSH)/0.5 mM oxidized glutathione (GSSG) and 0.05–0.1 mg/ml TEV protease and dialyzed overnight against 20 mM Tris/HCl, pH 8, 200 mM NaCl, 5% glycerol, 5 mM GSH, 0.5 mM GSSG using a membrane with a molecular weight cut-off of 3.5 kDa. The sample was incubated with ~1 ml Ni-NTA resin to remove the tags, concentrated, and applied to a Superdex 75 column using the same buffer without GSH and GSSG. Peak fractions were concentrated and used for assays or crystallization. All SP-BN mutant proteins were purified as described above.

SP-BM was purified from bovine lungs as follows. Lungs from one bovine calf were cut into small pieces and passed twice through a household mincer. Organic extraction was performed according to Bligh and Dyer (1959). The organic phase was dried using a rotary evaporator and dissolved in ~15 ml chloroform/methanol/0.1 M HCl (1:1:0.05) and loaded on a Sephadex LH-20 column equilibrated with the same solvent. Fractions that contain SP-BM based on Western blotting were pooled, dried using a rotary evaporator and solubilized in 20 mM Tris, pH 8, 200 mM NaCl, 2% glycerol, 6% n-octyl-β-D-glucopyranoside (octylglucoside) for 1 h. After centrifugation at 10,000 g for 10 min, the supernatant was applied to a Superdex 200 column using the same buffer except with 1.5% octylglucoside. Fractions containing SP-BM were pooled and concentrated.

Recombinant SP-BM was purified as follows. A fusion construct containing N-terminal His₆ and GST tags and human SP-BN and SP-BM domains with a thrombin cleavage site in between was expressed in E. coli cells (BL21 (DE3)) from an IPTG inducible promoter of the pGEX-6P-1 plasmid. The cells were grown in 2xYT medium at 37°C to mid log phase and induced with 1 mM IPTG at 30°C overnight. Cell pellets from a 6 L culture were homogenized at 25,000 psi (Avestin Emulsiflex -C3) in 150 ml 20 mM Tris/HCl pH 8, 200 mM NaCl and protease inhibitors. The sample was centrifuged for 10 min at 15,000g at 4°C (Sorvall RC5C Plus centrifuge, SLA-600TC rotor) to sediment the inclusion bodies. The pellet was resuspended and washed twice in 150 mL 20 mM Tris/HCl pH 8, 200 mM NaCl, 1% Triton. The pellet was resuspended in 50 mL 20 mM Tris/HCl pH 8, 200 mM NaCl, 0.6% sarkosyl and homogenized at 25,000 psi in an Emulsiflex instrument. The sample was incubated at room temperature for 1 hr and spun for 10 min at 15,000g at 4°C to remove any insoluble material. 2,000 U of thrombin (Thomas Scientific Inc.) was added to the supernatant and the sample was dialyzed against 2 L of 20 mM Tris/HCl pH 8.0, 200 mM NaCl for 4 hrs at room temperature using a membrane with a cut-off of 3.5 kDa. Following the dialysis and cleavage, the solution was complemented with 20 mM β-mercaptoethanol (β-ME) and incubated for 30 min at room temperature. To extract the SP-BM, 190 mL chloroform:methanol (1:2 (v/v)) was added and shaken for 1 min followed by incubation for 1 hr at room temperature. Additional 64 mL chloroform were added, shaken for 1 min, then 64 mL of buffer (20 mM Tris/HCl pH 8.0, 200 mM NaCl, 20 mM β-ME, 0.1% sarkosyl) for 1 min. The organic phase was collected and the aqueous phase was re-extracted with 90 mL chloroform. The combined organic phase was reduced to ~8 mL with a rotary evaporator (Büchi R-114) and applied to a Sephadex LH-20 size exclusion column equilibrated in chloroform:methanol:0.1M HCl (1:1:0.05 (v/v/v)). Fractions containing the protein were combined and the organic solvent was removed with a rotavapor. Protein was solubilized in 20 mM Tris/HCl pH 8.0, 200 mM NaCl, 6% β-octylglucoside (Anatrace) and subjected to a Superdex 200 Increase 5/150 GL column (GE Healthcare Life Sciences) equilibrated in 20 mM Tris/HCl pH 8.0, 200 mM NaCl, 1.5 % β-octylglucoside. The peak fractions were concentrated using a membrane with a 3.5 kDa cut-off.

To purify human proSP-B or mouse proSP-B containing the K105E/R110E mutations, each protein was expressed as a fusion with an N-terminal HALO tag. The gene was cloned into the lentiviral pHAGE2-DN-CMV-FLAG-HT7-PreScission vector and HEK293T cells were infected with lentiviral particles. Stable cell lines were selected by growth in the presence of 10 μg/ml blasticidin. The cells were grown in 15-cm plates until confluent, after which the medium was replaced with Opti-MEM I (Thermo Fisher) and collected every day for 3–4 days. About 400 ml of medium was incubated with 0.5 ml HALO resin (Promega) at room temperature for 4 hrs. The beads were washed with 20 mM Tris pH 8, 200 mM NaCl, 5% glycerol. They were resuspended in 0.5 ml buffer and incubated overnight at 4° C with 20 μg/ml of a fusion between GST and 3C protease (GST-3C). The supernatant was incubated with 100 μl GSH beads to remove the protease and concentrated with a 30 kDa cut-off filter.

Lipid transfer assays

To measure lipid transfer by flotation, liposomes containing an ER-like PL composition (65% 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), 15% 1,2-dioleoyl-snglycero-3-phospho-L-serine (DOPS), 17% 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE)) and 1.5% fluorescent NBD-PE at 0.9 mM final concentration of total lipids were incubated with 0.4 mM purified SP-BN in 50 μl of 0.1 M citrate buffer pH 5.4, 200 mM NaCl for 1 h at 37°C. The sample was mixed with an equal volume of 80% Nycodenz in the same buffer, and overlayed with 50 μl 30% Nycodenz, 30 μl 15% Nycodenz and 30 μl buffer. The sample was centrifuged in a TLS55 rotor (Beckman) at 48,000 rpm for 1 h at 4°C. Fractions of 40μl were taken from the top and analyzed in a fluorescence plate reader (FlexStation 3, Molecular Devices). An aliquot was analyzed by SDS-PAGE and Coomassie-blue staining. To measure lipid transfer by fluorescence de-quenching, liposomes contained an ER-like PL composition (see above) or an LB-like composition (70% dipalmitoyl phosphatidylcholine (DPPC), 30% phosphatidylglycerol (PG)). Donor liposomes also contained 1.5% NBD-PE and 1.5% rhodamine-PE. The assay was performed in 0.1 M citrate buffer pH 5.4, 200 mM NaCl or 0.1 M Tris buffer pH 7.4 or 7.5, 200 mM NaCl and contained in 50 μl 0.2 mM total lipid of donor vesicles, 0.6 mM lipid of acceptor vesicles, and different concentrations of SP-BN or proSP-B. The samples were incubated at 37°C in a fluorescence plate reader (excitation 460 nm; emission 538 nm). All readings were normalized to fluorescence measured after addition of 0.5% Triton X-100.

To determine lipid transfer by the filtration assay, liposomes of ER- or LB-like PL composition (final lipid concentration 1 mM) were incubated with SP-BN (final concentration 0.1 mM) in 100 μl final volume for different time periods at 37°C. The samples were placed on ice and centrifuged in an Eppendorf centrifuge at 14,000 g through 100 kDa molecular weight cut-off filters (Amicon Ultra 0.5 ml). Forty μl of the filtrate was extracted with chloroform/methanol, the organic phase was dried, and the pellet resuspended in 10 μl chloroform. TLC was performed with silica plates and 60% chloroform/35% methanol/5% H₂O as solvent. For the experiment in Figure S1C, a TLC plate impregnated with a solution of 1.2% boric acid in 50% ethanol was used with chloroform:methanol:water:ammonium hydroxide 120:75:6:2 as the solvent. Lipids were stained with Primuline and visualized in a fluorescence scanner.

Reconstitution of SP-BM into liposomes

A mixture of 100 μg DPPC + 50 μg egg PG in chloroform was evaporated to dryness under nitrogen and then in vacuo overnight. To the dried residue, 15 μg SP-BM was added, and the volume was brought to 100 μl with SEC buffer (20 mM Tris, pH 8.0, 200 mM NaCl, 2% glycerol, 1.5% OG). The mixture was incubated at 40°C for 1 h and then diluted 5 times with 20 mM Tris, pH 7.4, 200 mM NaCl, 1 mM EDTA, and dialyzed against 2–5 changes of the same buffer. 100 μl aliquots were then incubated overnight at 37°C with or without CaCl₂ (final concentration of 5 mM).

Thin-section EM

Reconstituted samples were fixed in suspension by adding an equal volume of fixative (2.5% paraformaldehyde, 5% glutaraldehyde, 0.06% picric acid in 0.2 M cacodylate buffer). After 1 h at room temperature, the samples were spun at 10,000–20,000 g for 15 min and stored overnight in fixative at 4°C. The samples were washed in 0.1 M cacodylate buffer and postfixed with 1% osmium tetroxide/1.5% potassium ferrocyanide for 1 h, washed in water 3 times and incubated in 1% aqueous uranyl acetate for 1 h, followed by 2 washes in water and subsequent dehydration in grades of alcohol (10 min each in 50%, 70%, 90%; and 2 times 10min in 100%). The samples were then put in propylene oxide for 1 h and infiltrated overnight in a 1:1 mixture of propylene oxide and TAAB Epon (Marivac Canada Inc. St. Laurent, Canada). The following day the samples were embedded in TAAB Epon and polymerized at 60°C for 48 h.

Ultrathin sections (about 60 nm) were cut on a Reichert Ultracut-S microtome, picked up on to copper grids stained with lead citrate and examined in a TecnaiG² Spirit BioTWIN and images were recorded with an AMT 2k CCD camera.

Negative-stain EM

Formvar/Carbon 400 mesh copper grids (Ted Pella) were plasma cleaned in EMS100X for 30 s at 30 mA. The grids were incubated with 3 μl sample of SEC fractions for 30 s. Excess solution was blotted away and the grids were stained with 1.5% uranyl formate for 30 s. Micrographs were acquired on a Phillips CM10 microscope equipped with a Gatan UltraScan 894 2k CCD camera and operated at 100 kV.

Experiments with MLE-15 cells

To generate SP-B knock-out MLE-15 cells, 1,160,250 cells were plated on a 60-mm dish and transfected the next day with 2.5 μg each of two sgRNA-expressing plasmids and 0.5 μg of an EYFP plasmid that confers hygromycin resistance, using Lipofectamine 3000 (Invitrogen). One day after transfection, cells were split 1:5 and media supplemented with 300 μg/ml hygromycin were added every other day for a total of 10 days to select for positive clones. The transfected and selected cells were trypsinized and diluted to achieve 1 cell per 100 μl for seeding in 96-well plates. Singlecell wells were progressively expanded in size up to 6-well plates, when they were split into two, so that half of the cells could be used for PCR and sequencing of genomic DNA by incubating in 150 μl of 50 mM NaOH at 95–100°C for 30 min, neutralizing with 50 μl of 1 M Tris/HCl, pH 8, and removing insoluble debris by spinning for 5 min.

For transient expression of various SP-B constructs, 525,000 cells were plated per well in a 6-well plate and transfected the next day with 2.5 μg of total plasmid using Lipofectamine 3000. Two days after transfection, cells were scraped in medium and centrifuged at 1000 g for 5 min. Medium was saved (~2 ml), cells were washed with PBS and resuspended in 500 μl 20 mM Tris/HCl, pH 8, 5 mM EDTA and protease inhibitors. After passing through a 22G needle 25 times, samples were centrifuged again at 1000 g for 5 min. The post-nuclear supernatant (PNS) was centrifuged at 16,000 g for 15 min to obtain the cellular membranes, which were resuspended in PBS (typically 50 μl). Equal volumes (typically 15 μl) of media and membranes were loaded on SDS gels. In some experiments, the resuspended membranes were mixed with an equal volume of 0.2 M Na₂CO₃ (pH > 11) or TX-100 was added to a final concentration of 1%. After 30-min incubation on ice, insoluble material was sedimented by centrifugation at 16,000 g for 30 min and washed with PBS. In Figure S3B, equal volumes of media and resuspended membranes were treated with Endo H (NEB) following manufacturer’s instructions. For immunoblotting human proSP-B with mouse F-2 mAb, samples were reduced by adding β-ME to the SDS-loading buffer. Immunoblots using rabbit anti-mature SP-B and rabbit anti-mouse SP-BN (for detecting both mouse SP-BN and mouse proSP-B) were done under non-reducing conditions. Immunoblots were visualized with an Amersham Imager 600.

Lamellar body enrichment

Lamellar bodies were isolated by modifications of the method of (Chander et al., 1983) and (Matsumura et al., 2007). Confluent cells (five 100-mm dishes) were harvested and sonicated in PBS containing 1 M sucrose and a protease inhibitor mixture using a probe sonicator. Sonication was performed with two bursts of 15 s each with a 5 s interval and a 40% maximum power output. The sample was then centrifuged at 1,000 g for 10 min to obtain PNS. A sucrose gradient consisting of 1 ml each of 0.8, 0.7, 0.6, 0.5, 0.4, 0.3 M, and 0.2 M sucrose was layered successively above 3.5 ml of PNS. The gradient was spun in a Beckman SW 55 Ti rotor at 80 000 g for 3 h. After centrifugation, 1 ml of each fraction was collected from the top.

Immunofluorescence

52,500 SP-B KO MLE-15 cells per well were set up in an 8-well chamber slide and transfected the next day with 0.25 μg of human proSP-B (wild type or L96K/L105E/L140K mutant) using Lipofectamine 3000. Two days after transfection, cells were fixed with 4% formaldehyde, permeabilized with 0.2% saponin and stained with mouse anti-human SP-B (F-2) and rabbit anti-mature SP-B or rabbit anti-calreticulin (Abcam) antibodies. Images were collected with a Yokogawa spinning disk confocal on a Nikon Ti-E inverted microscope equipped with a 100x objective lens. Single confocal sections are displayed, and gamma, brightness, and contrast were adjusted (identically for compared image sets) using ImageJ.

Experiments with mice

A cocktail of 0.61 pmol/μl of an equimolar mixture of crRNA and tracrRNA, 100 ng/μl Cas9 protein (Aida et al, 2015) and 10 ng/μl single-stranded oligodeoxynucleotide (ssODN, IDT) was injected into the pronuclei of E0.5 embryos (C57BL6/Envigo). Postinjection surviving embryos were re-implanted into recipient CD1 pseudo-pregnant females and either allowed to develop to term or harvested at E18.5. Tails were cut for PCR and sequencing of genomic DNA as above. For immunoblotting, lungs were cut into small pieces, placed in 1 ml RIPA buffer with protease inhibitors, and homogenized in a bead beater using an equal volume of glass beads. Lysates were cleared by centrifugation first at 1000 g and then at 10,000 g. For electron microscopy, lung pieces were placed in EM fixative and processed as above.

Lipid analysis by mass spectrometry

Lipids were extracted by adding 50% acetonitrile to purified proteins and analyzed using a Thermo Q Exactive Plus mass spectrometer coupled to a Thermo Ultimate 3000 uHPLC. Lipids were identified using Thermo Lipidsearch 4.1.

CD spectroscopy

The CD spectra were acquired on a Jasco J-815 spectropolarimeter equipped with a PFD-425S/15 Peltier unit. Spectra were accumulated 5 times from 260–200 nm, using a scanning rate of 50 nm/min. Measurements were performed at 20°C in a 1 mm Quartz cuvette. Molecular ellipticities were calculated for the mean residue weight (MRW). Protein concentrations were 0.1– 0.3 mg/ml.

Analysis of lipoprotein particles

ProSP-B (20 μg), purified as above, was mixed with 1.2 mM liposomes containing a LB-like lipid composition (70% DPPC, 30% egg PG) and 1.5% fluorescent rhodamine-PE and incubated in either 100 mM citrate pH 5.6, 200 mM NaCl or 20 mM Tris/HCl pH 7.5, 200 mM NaCl for 30 min at 37°C or 0°C. The samples were then applied to a Superose 6 Increase 3.2/300 GL column equilibrated in either pH 5.6 or pH 7.5 buffer. Absorbance was monitored at both 280 nm and 560 nm. Fractions were collected and equal volumes of each fraction were analyzed by SDS-PAGE and immunoblotting.

For agarose gel electrophoresis, 3.6 μg of either human proSP-B (wild-type) or mouse proSP-B K105E/R110E was incubated in a final volume of 10 μl with 1.2 mM liposomes containing LB-type lipids, NBD-PE, and rhodamine-PE in either 100 mM citrate pH 5.4, 200 mM NaCl or 100 mM Tris/HCl pH 7.4, 200 mM NaCl. After addition of 2.5 μl 80% glycerol, electrophoresis was performed with 0.6% agarose gels in barbital buffer overnight at 4°C. VLDL was pre-stained as described by Greenspan and Gutman (1993). The gel was imaged in Amersham Imager 600 using 460-nm excitation wavelength and Cy2 filter.

ProSP-B processing in vitro

ProSP-B (20 μg), purified as above, was mixed with 1.2 mM liposomes containing a LB-like lipids (70% DPPC, 30% egg PG) and 1.5% fluorescent rhodamine-PE and incubated in 100 mM citrate pH 5.6, 200 mM NaCl for 30 min at 37°C. An aliquot was analyzed by negative-stain EM to ascertain the formation of lipoprotein particles. The remainder of the sample was incubated with proteinase K (final concentration of 10 μg/mL) for 30 min at 37°C. An aliquot was then analyzed directly by SDS-PAGE and immunoblotting. Another aliquot was centrifuged for 15 min at 4°C at 18,000 g, and the supernatant was analyzed by SDS-PAGE and immunoblotting. Fixative (2.5% paraformaldehyde, 5% glutaraldehyde, 0.06% picric acid in 0.2 M cacodylate buffer) was added to the pellet, the sample again centrifuged for 15 min at 4°C at 18,000 g, and incubated for 1 hr at room temperature and at 4°C overnight. The sample was then analyzed by thin-section EM.

Crystallization

Crystal conditions were screened using the hanging drop method in a 96 well format with a Mosquito robot (TTP Labtech). Crystallization trays were incubated at 20°C with the exception of trays for SP-BN K46E/R51E, which were incubated at 14°C. In all cases, crystals appeared in 1–3 days.

Crystals of wild-type SP-BN were obtained by mixing the protein solution (7–12 mg/mL 1:1 with mother liquor consisting of 100 mM citrate buffer pH 5.4, 150 mM sodium chloride, 17% (w/v) PEG 3,350. Crystals were incubated briefly in cryoprotection solution consisting of mother liquor with 25% glycerol before flash freezing in liquid nitrogen.

Crystals of SP-BN Y59A/H79A were obtained by mixing the protein solution (62 mg/mL) 1:1 with mother liquor consisting of 100 mM sodium/potassium phosphate pH 6.2, 200 mM sodium chloride, 47.5% (w/v) PEG 200. Crystals were flash frozen in liquid nitrogen without additional cryoprotection.

Crystals of SP-BN K46E/R51E were obtained by mixing the protein solution (49 mg/mL) 1:1 with mother liquor consisting of 100 mM citrate buffer pH 5.4, 23–25% (w/v) PEG 4,000, 14–15% 2-propanol. Crystals were incubated briefly in cryoprotection solution consisting of 100 mM citrate buffer pH 5.4, 26% (w/v) PEG 4,000, 25% glycerol before flash freezing in liquid nitrogen.

Crystals of SP-BC at pH 8.5 were obtained by mixing the protein solution (25 mg/mL) 1:1 with mother liquor consisting of 100 mM Tris pH 8.5, 200 mM lithium sulfate, 1,260 mM ammonium sulfate. Crystals were harvested directly from the 96-well screen, incubated briefly in cryoprotection solution consisting of mother liquor with 25% glycerol, and flash frozen in liquid nitrogen.

Crystals of SP-BC at ~pH 4.5 were obtained by mixing the protein solution (25 mg/mL) 1:1 with mother liquor consisting of 170 mM ammonium sulfate, 21–23% PEG 4,000, 15% glycerol. Crystals were incubated briefly in cryoprotection solution consisting of mother liquor with 25% glycerol before flash freezing in liquid nitrogen.

Structure determination

Data were collected on beamlines 24-ID-C and 24-ID-E at the Advanced Photon Source (Argonne National Laboratory). Data were processed with XDS (Kabsch, 2010) and analyzed with Aimless (Evans and Murshudov, 2013). Datasets have been deposited at SBgrid (Morin et al., 2013).

Data for wild-type SP-BN were collected at a wavelength of 1.7712 Å. The crystals belonged to space group P2₁2₁2₁. Seven datasets collected from three crystals (DOIs: 10.15785/SBGRID/757, 10.15785/SBGRID/758, 10.15785/SBGRID/759, 10.15785/SBGRID/760, 10.15785/SBGRID/761, 10.15785/SBGRID/762, 10.15785/SBGRID/763) were combined using the scale-and-merge program within Phenix (Adams et al., 2010) to produce a merged dataset with a resolution of 2.2 Å and an anomalous completeness of 95.21%. A thorough sub-structure search yielded initial phases using single-wavelength anomalous dispersion of sulfur atoms with Phenix Autosol (Adams et al., 2010). An initial structure was obtained with Phenix AutoBuild (Terwilliger et al., 2008), and the structure was completed with iterative rounds of manual adjustment in Coot (Emsley and Cowtan, 2004) and refinement in Phenix (Adams et al., 2010). TLS parameters were enabled toward the end of refinement, and a riding hydrogen model was employed. Lipid densities were modeled as phosphatidylethanolamine (16:1/18:1), the most abundant species identified by mass spectrometry of the protein sample. The central lipids were modeled with two alternative conformations (i.e., with the head groups at either end of the central cavity), each with an occupancy of 0.5. When the central lipid was modeled with the head group on one side of the hydrophobic cavity, refinement produced strong positive difference density connected to the acyl chains on the other side of the cavity. We therefore modeled the central lipids with two alternative conformations. Although there was still some residual difference density, this approach produced better free R factors than modeling a single conformation for the central lipid.

Data for SP-BN Y59A/H79A were collected at a wavelength of 0.9791 Å. The crystals belonged to space group C121. Data used for structure determination (DOI: 10.15785/SBGRID/773) were collected from a single crystal, which diffracted to 2.3 Å. Initial phases were obtained by molecular replacement, searching for four copies of a single dimer of wild-type SP-BN. Lipids were removed from the search model to avoid biasing the resulting solution. An initial structure was obtained with Phenix AutoBuild (Terwilliger et al., 2008), and the structure was completed with iterative rounds of manual adjustment in Coot (Emsley and Cowtan, 2004) and refinement in Phenix. TLS parameters were enabled toward the end of refinement, and a riding hydrogen model was employed. Lipid densities were modeled as phosphatidylethanolamine (18:0/18:0), with some acyl chains and head groups truncated during manual adjustment owing to poor density, which was likely reflective of the increased heterogeneity of the bound lipids in this structure.

Data for SP-BN K46E/R51E were collected at a wavelength of 0.9792 Å. The crystals belonged to space group P12₁1. Four datasets collected from one crystal (DOIs: 10.15785/SBGRID/768, 10.15785/SBGRID/769, 10.15785/SBGRID/770, 10.15785/SBGRID/771) were combined to form a merged dataset with a resolution of 1.88 Å. Initial phases were obtained by molecular replacement, searching for two copies of a single dimer of wild-type SP-BN. Lipids were removed from the search model to avoid biasing the resulting solution. An initial structure was obtained with Phenix AutoBuild (Terwilliger et al., 2008), and the structure was completed with iterative rounds of manual adjustment in Coot (Emsley and Cowtan, 2004) and refinement in Phenix. TLS parameters were enabled toward the end of refinement, and a riding hydrogen model was employed. Lipid densities were modeled as phosphatidylethanolamine (16:1/18:1). As for wild-type SP-BN, the central lipids were modeled with two alternative conformations (i.e., with the head groups at either end of the central cavity). Given the higher resolution of this dataset, alternative conformation occupancies were refined.

Data for SP-BC at high pH were collected at a wavelength of 0.9792 Å. The crystals belonged to space group P2₁2₁2₁. Data used for structure determination (DOI: 10.15785/SBGRID/766) were collected from a single crystal, which diffracted to 1.75 Å. A solution was obtained using the ab initio phasing software ARCIMBOLDO (Rodríguez et al., 2009), using a search for four helical fragments of fourteen residues each. An initial structure was obtained with Phenix AutoBuild (Terwilliger et al., 2008), and the structure was completed with iterative rounds of manual adjustment in Coot (Emsley and Cowtan, 2004) and refinement in Phenix. A metal ion was observed to participate in crystal contacts and was assigned as zinc in a trigonal bipyramidal configuration coordinated by two histidines, two aspartates, and a water molecule; this assignment was validated with the program CheckMyMetal (Zheng et al., 2017).

Data for SP-BC at low pH were collected at a wavelength of 0.9791 Å. The crystals belonged to space group P12₁1. Data used for structure determination (DOI: 10.15785/SBGRID/767) were collected from a single crystal, which diffracted to 1.90 Å. Initial phases were obtained by molecular replacement, searching for eight copies of the two helices of SP-BC at high pH that are connected by a single disulfide bond (i.e., residues A26 to V45 of the C terminal domain). An initial structure was obtained with Phenix AutoBuild (Terwilliger et al., 2008), and the structure was completed with iterative rounds of manual adjustment in Coot (Emsley and Cowtan, 2004) and refinement in Phenix. Three sulfate molecules were placed into strong difference density due to the presence of this ion in the crystallization solution and plausible coordinating residues. The program PISA predicts that the protein forms a tetramer.

QUANTIFICATION AND STATISTICAL ANALYSIS

Basic quantifications

Proteins are quantified in solution based on absorbance at 280 nm using the extinction coefficient provided by the Protparam tool (Wilkins et al., 1999). DNA is quantified based on ε = 20(mg/mL)⁻¹cm⁻¹. Lipids co-purified with proteins are quantified by comparing to known amounts of standards on a TLC plate. The amount of fluorescent lipid in lipoprotein fractions is quantified by either measuring the absorbance at 560 nm or the fluorescence in a plate reader. The amount of proSP-B in these fractions is determined on the basis of a comparison with a titration of purified proSP-B in immunoblots. Density of sucrose after density gradient centrifugation is measured by a refractometer. Reagents used otherwise were quantified by weight before solution preparation.

Statistical details

All experiments were repeated 2–3 times. Crystallographic data collection and refinement statistics are listed in Table 1. All cell-culture experiments were repeated with separate starter cultures in different flasks or multiplex well plates. CRISPR-Cas9 editing was performed on 5–16 embryos in each experiment. Immunoblots were performed on 7–9 embryos. EM images representative of each genotype are shown. High-magnification EM images of SP-BM reconstitutions are representative of at least 5 images.

Supplementary Material

NIHMS1645341-supplement-1.pdf^{(140.5MB, pdf)}

Key Resources Table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Mouse monoclonal anti-SP-B (clone F-2)	Santa Cruz Biotechnology	Cat#sc-133143; RRID: AB_2285686
Rabbit polyclonal anti-mature SP-B	Seven Hills Bioreagents	Cat#WRAB-48604
Rabbit polyclonal anti-mouse SP-BN	This paper	N/A
Bacterial and Virus Strains
Escherichia coli: BL21 (DE3) strain	Sigma-Aldrich	Cat#69450–3
Escherichia coli: Rosetta-gami 2 (DE3) strain	Sigma-Aldrich	Cat#71351–3
pHAGE2-DN-CMV-FLAG-HT7-PreScission	Laboratory of Adrian Salic	N/A
Biological Samples
Bovine lungs	Research 87	N/A
Chemicals, Peptides, and Recombinant Proteins
16:0–18:1 PC (POPC)	Avanti Polar Lipids	Cat#850457C
18:1 PS (DOPS)	Avanti Polar Lipids	Cat#840035C
18:1 (Δ9-Cis) PE (DOPE)	Avanti Polar Lipids	Cat#850725C
16:0 PC (DPPC)	Avanti Polar Lipids	Cat#850355C
Egg PG	Avanti Polar Lipids	Cat#841138C
16:0 NBD PE	Avanti Polar Lipids	Cat#810144C
16:0 Liss Rhod PE	Avanti Polar Lipids	Cat#810158C
Mouse SP-BN	This paper	N/A
Mouse SP-BC	This paper	N/A
Human GST-SPBN-thrombin-SPBM	This paper	N/A
Human proSP-B	This paper	N/A
Deposited Data
X-ray diffraction datasets for SP-BN (wild type)	SBGrid	DOIs: 10.15785/SBGRID/757, 10.15785/SBGRID/758, 10.15785/SBGRID/759, 10.15785/SBGRID/760, 10.15785/SBGRID/761, 10.15785/SBGRID/762, 10.15785/SBGRID/763
X-ray diffraction dataset for SP-BN Y59A/H79A	SBGrid	DOI: 10.15785/SBGRID/773
X-ray diffraction dataset for SP-BN K46E/R51E	SBGrid	DOIs: 10.15785/SBGRID/768, 10.15785/SBGRID/769, 10.15785/SBGRID/770, 10.15785/SBGRID/771
X-ray diffraction dataset for SP-BC (high pH)	SBGrid	DOI: 10.15785/SBGRID/766
X-ray diffraction dataset for SP-BC (low pH)	SBGrid	DOI: 10.15785/SBGRID/767
SP-BN (wild type) structure	Protein Data Bank	6VYN
SP-BN Y59A/H79A structure	Protein Data Bank	6W1B
SP-BN K46E/R51E structure	Protein Data Bank	6VZD
SP-BC high pH structure	Protein Data Bank	6VZE
SP-BC low pH structure	Protein Data Bank	6VZ0
Unprocessed images	Mendeley datasets	http://dx.doi.org/10.17632/s7w5tw5tvp.1
Experimental Models: Cell Lines
Human: HEK293T cells	Laboratory of Adrian Salic	ATCC: CRL-3216
Mouse: MLE-15 cells	Laboratory of Jeffrey Whitsett	N/A
Mouse: MLE-15 SP-B KO cells	This paper	N/A
Experimental Models: Organisms/Strains
Mouse: C57BL/6: C57BL/6NHsd	Envigo	Envigo: 4410F
Oligonucleotides
ssDNA donor for SP-B L95K/L104E/V139K knock-in: AAAGACCAGGCTAATCCTCCCTTCTCTGCTCTCCCAGGAAGCAATTCGGAAGTTCaaGGAACAAGAATGTGATATCCTTCCCgaGAAGCTGCTTGTGCCCCGGTGTCGCCAAGTGCTTGATGTCTACCTGCCCCTGGTTATTGACTACTTCCAGAGCCAGATTAACCCCAAAGCTATCTGCAATCATAAAGGCCTGTGCCCACGTGGGCAGGCTAAGCCAGAACAGAATCCAGGGATGCCGGAT	This paper	N/A
ssDNA donor for SP-B K105E/R110E knock-in: CTCCCAGGAAGCAATCCGGAAGTTCCTGGAACAAGAATGTGATATCCTTCCCCTTgAGCTGCTTGTGCCCgaGTGTCGCCAAGTGCTTGATGTCTACCTGCCCCTGGTTATTGACTACTTCCAGAGC	This paper	N/A
CRISPR guide RNA #1 for SP-B knock-out and L95K/L104E/V139K knock-in: ATTCTTGTTCCAGGAACTTC	This paper	N/A
CRISPR guide RNA #2 for SP-B knock-out and L95K/L104E/V139K knock-in: AAAGCCATCTGCAATCATGT	This paper	N/A
CRISPR guide RNA for SP-B K105E/R110E knock-in: CGGGGCACAAGCAGCTTCAA	This paper	N/A
Forward primer for genotyping SP-B: CCAGGCTAATCCTCCCTTCT	Laboratory of Bernard Thebaud	N/A
Reverse primer for genotyping K105E/R110E knock-in mice: CCCACTTAGGCACATGCAC	Laboratory of Bernard Thebaud	N/A
Reverse primer for genotyping SP-B knockout cells and L95K/L104E/V139K knock-in mice: TTGCCATCTCTGCCTCCTAG	This paper	N/A
Recombinant DNA
Synthetic, codon-optimized (for E. coli) mouse SP-BN	Thermo Fisher Scientific (GeneArt)	N/A
Synthetic, codon-optimized (for E. coli) mouse SP-BC	Thermo Fisher Scientific (GeneArt)	N/A
Plasmid: pX330-U6-Chimeric_BB-CBh-hSpCas9	Cong et al., 2013	Addgene Plasmid #42230
Plasmid: pcDNA3.1 Hygro EYFP H148Q/I152L	Galietta et al., 2001	Addgene Plasmid #25874
Plasmid: pET-32a(+)	Sigma-Aldrich	Cat#69015
Plasmid: pGEX-6P-1	GE Healthcare	Cat#28954648
Plasmid: pCMV6 mouse Sftpb	OriGene	Cat#MR226883; NM_147779
Plasmid: pCS6 human SFTPB cDNA	transOMIC	Cat#TCH1303; BC032785
Plasmid: pCS6 human SFTPB ORF	This paper	N/A
Plasmid: pCS6 human SP-BN (1–200)	This paper	N/A
Plasmid: pGEX-6P-1 human SPB-N-thrombin-SPBM	This paper	N/A
Plasmid: pcDNA3.1 human SPBΔC(1–279)-Flag	This paper	N/A
pHAGE2-Flag-HT7 human proSP-B (24–381)	This paper	N/A
pHAGE2-Flag-HT7 mouse proSP-B (24–382) K105E/R110E	This paper	N/A
Software and Algorithms
ImageJ	Schneider et al., 2012	https://imagej.nih.gov/ij/
Benchling		https://www.Benchling.com
CHOPCHOP	Labun et al., 2019	https://chopchop.cbu.uib.no
XDS	Kabsch, 2010
Aimless	Evans and Murshudov, 2013
ARCIMBOLDO	Rodríguez et al., 2009
Phenix	Adams et al., 2010
Phenix AutoBuild	Terwilliger et al., 2008
Coot	Emsley and Cowtan, 2004	https://www2.mrc-lmb.cam.ac.uk/personal/pemsley/coot/
CheckMyMetal	Zheng et al., 2017
CavityPlus	Xu et al., 2018
Chimera	Petterson et al., 2004	https://www.cgl.ucsf.edu/chimera/
Swiss-model	Waterhouse et al., 2018	https://swissmodel.expasy.org
Other
Sephadex LH-20	GE Healthcare	Cat#17009001

Open in a new tab

Highlights:

Mechanism of lamellar body (LB) formation by lung surfactant protein B (SP-B)
Three related domains in the SP-B precursor (proSP-B) cooperate in LB formation
The N-terminal domain transfers lipids in proSP-B to form lipoprotein particles
The middle domain generates the stacked membrane layers characteristic of LBs

ACKNOWLEDGMENTS

We thank M. Ericsson and her team of the Harvard Medical School EM facility for performing thin-section EM, S. Trauger at the Harvard Center for Mass Spectrometry for lipid analysis, Nikon Imaging Center at Harvard Medical School for support, and D. Pellman, R. Farese, and M. Z. Bao for reading the manuscript. CRISPR-Cas9 knock-ins were generated at Mouse Gene Manipulation Core (IDDRC grant #1U54HD090255). The structural work is based upon research conducted at the Northeastern Collaborative Access Team beamlines, which are funded by the National Institute of General Medical Sciences from the National Institutes of Health (P30 GM124165). The Eiger 16M detector on 24-ID-E is funded by a NIH-ORIP HEI grant (S10OD021527). This research used resources of the Advanced Photon Source, a U.S. Department of Energy (DOE) Office of Science User Facility operated for the DOE Office of Science by Argonne National Laboratory under Contract No. DE-AC02-06CH11357. N.O.B. is the recipient of a NIGMS Medical Scientist Training Program (MSTP) grant (T32GM007753), X.W. was supported by a Jane Coffin Child fellowship, and T.A.R. by an NIH grant from the NHLBI (R01 HL150520). T.A.R. is a Howard Hughes Medical Institute Investigator.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. 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.

DECLARATION OF INTERESTS

N.S., G.M., N.O.B. and T.A.R. have a pending patent application.

REFERENCES

Adams PD, Afonine PV, Bunkoczi G, Chen VB, Davis IW, Echols N, Headd JJ, Hung LW, Kapral GJ, Grosse-Kunstleve RW, et al. (2010). PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr. D Biol. Crystallogr 66, 213–221. [DOI] [PMC free article] [PubMed] [Google Scholar]
Aggarwal S, Yurlova L, and Simons M (2011). Central nervous system myelin: structure, synthesis and assembly. Trends Cell Biol. 21, 585–593. [DOI] [PubMed] [Google Scholar]
Ahn VE, Faull KF, Whitelegge JP, Fluharty AL, and Privé GG (2003). Crystal structure of saposin B reveals a dimeric shell for lipid binding. Proc. Natl. Acad. Sci. USA 100, 38–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
Aida T, Chiyo K, Usami T, Ishikubo H, Imahashi R, Wada Y, Tanaka KF, Sakuma T, Yamamoto T, and Tanaka K (2015). Cloning-free CRISPR/Cas system facilitates functional cassette knock-in in mice. Genome Biol. 16, 87. [DOI] [PMC free article] [PubMed] [Google Scholar]
Akinbi HT, Breslin JS, Ikegami M, Iwamoto HS, Clark JC, Whitsett JA, Jobe AH, and Weaver TE (1997). Rescue of SP-B knockout mice with a truncated SP-B proprotein. Function of the C-terminal propeptide. J. Biol. Chem 272, 9640–9647. [DOI] [PubMed] [Google Scholar]
Beck DC, Ikegami M, Na CL, Zaltash S, Johansson J, Whitsett JA, and Weaver TE (2000). The role of homodimers in surfactant protein B function in vivo. J. Biol. Chem 275, 3365–3370. [DOI] [PubMed] [Google Scholar]
Benson BJ, Williams MC, Sueishi K, Goerke J, and Sargeant T (1984). Role of calcium ions in the structure and function of pulmonary surfactant. Biochem. Biophys. Acta 793, 18–27. [DOI] [PubMed] [Google Scholar]
Bernhard W (2016). Lung surfactant: Function and composition in the context of development and respiratory physiology. Ann. Anat 208, 146–150. [DOI] [PubMed] [Google Scholar]
Bian X, Klemm RW, Liu TY, Zhang M, Sun S, Sui X, Liu X, Rapoport TA, and Hu J (2011). Structures of the atlastin GTPase provide insight into homotypic fusion of endoplasmic reticulum membranes. Proc. Natl. Acad. Sci. USA 108, 3976–3981. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bligh EG, and Dyer WJ (1959). A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol 37, 911–917. [DOI] [PubMed] [Google Scholar]
Brasch F, Johnen G, Winn-Brasch A, Guttentag SH, Schmiedl A, Kapp N, Suzuki Y, Müller KM, Richter J, Hawgood S, et al. (2004). Surfactant protein B in type II pneumocytes and intra-alveolar surfactant forms of human lungs. Am. J. Respir. Cell Mol. Biol 30, 449–458. [DOI] [PubMed] [Google Scholar]
Bruhn H (2005). A short guided tour through functional and structural features of saposin-like proteins. Biochem. J 389, 249–257. [DOI] [PMC free article] [PubMed] [Google Scholar]
Chander A, Dodia CR, Gil J, and Fisher AB (1983). Isolation of lamellar bodies from rat granular pneumocytes in primary culture. Biochim. Biophys. Acta 753, 119–129. [DOI] [PubMed] [Google Scholar]
Chander A, Johnson RG, Reicherter J, and Fisher AB (1986). Lung lamellar bodies maintain an acidic internal pH. J. Biol. Chem 261, 6126–6131. [PubMed] [Google Scholar]
Clark JC, Wert SE, Bachurski CJ, Stahlman MT, Stripp BR, Weaver TE, and Whitsett JA (1995). Targeted disruption of the surfactant protein B gene disrupts surfactant homeostasis, causing respiratory failure in newborn mice. Proc. Natl. Acad. Sci. USA 92, 7794–7798. [DOI] [PMC free article] [PubMed] [Google Scholar]
Cong L, Ran FA, Cox D, Lin S, Barretto R, Habib N, Hsu PD, Wu X, Jiang W, Marrafinni LA, and Zhang F (2013). Multiplex genome engineering using CRISPR/Cas systems. Science 339, 819–823. [DOI] [PMC free article] [PubMed] [Google Scholar]
Derman AI, Prinz WA, Belin D, and Beckwith J (1993). Mutations That Allow Disulfide Bond Formation in the Cytoplasm of Escherichia coli. Science 262, 1744–1747. [DOI] [PubMed] [Google Scholar]
Diemel RV, Snel MM, van Golde LM, Putz G, Haagsman HP, and Batenburg JJ (2002). Effects of cholesterol on surface activity and surface topography of spread surfactant films. Biochemistry 41, 15007–15016. [DOI] [PubMed] [Google Scholar]
Echaide M, Autilio C, Arroyo R, and Perez-Gil J (2017). Restoring pulmonary surfactant membranes and films at the respiratory surface. Biochim. Biophys. Acta. Biomembr 1859, 1725–1739. [DOI] [PubMed] [Google Scholar]
Emsley P, and Cowtan K (2004). Coot: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr 60, 2126–2132. [DOI] [PubMed] [Google Scholar]
Evans PR, and Murshudov GN (2013). How good are my data and what is the resolution? Acta Crystallogr. D Biol. Crystallogr 69, 1204–1214. [DOI] [PMC free article] [PubMed] [Google Scholar]
Galietta LJ, Haggie PM, and Verkman AS (2001). Green fluorescent protein-based halide indicators with improved chloride and iodide affinities. FEBS Lett. 499, 220–224. [DOI] [PubMed] [Google Scholar]
Gerson KD, Foster CD, Zhang P, Zhang Z, Rosenblatt MM, and Guttentag SH (2008). Pepsinogen C proteolytic processing of surfactant protein B. J. Biol. Chem 283, 10330–10338. [DOI] [PMC free article] [PubMed] [Google Scholar]
Greenspan P, and Gutman RL (1993). Detection by Nile Red of agarose gel electrophoresed native and modified low density lipoprotein. Electrophoresis 14, 65–68. [DOI] [PubMed] [Google Scholar]
Kabsch W (2010). Xds. Acta Crystallogr. D Biol. Crystallogr 66, 125–132. [DOI] [PMC free article] [PubMed] [Google Scholar]
King RJ (1982). Pulmonary surfactant. J. Appl. Physiol. Respir. Environ. Exerc. Physiol 53, 1–8. [DOI] [PubMed] [Google Scholar]
Kolter T, and Sandhoff K (2005). Principles of lysosomal membrane digestion: stimulation of sphingolipid degradation by sphingolipid activator proteins and anionic lysosomal lipids. Annu. Rev. Cell Dev. Biol 21, 81–103. [DOI] [PubMed] [Google Scholar]
Korimilli A, Gonzales LW, and Guttentag SH (2000). Intracellular localization of processing events in human surfactant protein B biosynthesis. J. Biol. Chem 275, 8672–8679. [DOI] [PubMed] [Google Scholar]
Labun K, Montague TG, Krause M, Torres Cleuren YN, Tjeldsen H and Valen E (2019). CHOPCHOP v3: expanding the CRISPR web toolbox beyond genome editing. Nucleic Acids Res. 47, W171–W174. [DOI] [PMC free article] [PubMed] [Google Scholar]
Li Y, Kinting S, Hoppner S, Forstner ME, Uhl O, Koletzko B, and Griese M (2019). Metabolic labelling of choline phospholipids probes ABCA3 transport in lamellar bodies. Biochim. Biophys. Acta Mol. Cell Biol. Lipids 1864, 158516. [DOI] [PubMed] [Google Scholar]
Liekkinen J, Enkavi G, Javanainen M, Olmeda B, Perez-Gil J and Vattulainen I (2020). Pulmonary surfactant lipid reorganization induced by the adsorption of the oligomeric surfactant protein B complex. J. Mol. Biol 432, 3251–3268. [DOI] [PubMed] [Google Scholar]
Lin S, Akinbi HT, Breslin JS, and Weaver TE (1996a). Structural requirements for targeting of surfactant protein B (SP-B) to secretory granules in vitro and in vivo. J. Biol. Chem 271, 19689–19695. [DOI] [PubMed] [Google Scholar]
Lin S, Phillips KS, Wilder MR, and Weaver TE (1996b). Structural requirements for intracellular transport of pulmonary surfactant protein B (SP-B). Biochim. Biophys. Acta 1312, 177–185. [DOI] [PubMed] [Google Scholar]
Lopez-Rodriguez E, Gay-Jordi G, Mucci A, Lachmann N, and Serrano-Mollar A (2017). Lung surfactant metabolism: early in life, early in disease and target in cell therapy. Cell Tissue Res. 367, 721–735. [DOI] [PubMed] [Google Scholar]
Martínez-Calle M, Prieto M, Olmeda B, Fedorov A, Loura LMS and Pérez-Gil J (2020). Pulmonary surfactant protein SP-B nanorings induce the multilamellar organization of surfactant complexes. Biochim. Biophys. Acta Biomembr 1862, 183216. [DOI] [PubMed] [Google Scholar]
Mathialagan N, and Possmayer F (1990). Low-molecular-weight hydrophobic proteins from bovine pulmonary surfactant. Biochim. Biophys. Acta 1045, 121–127. [DOI] [PubMed] [Google Scholar]
Matsumura Y, Sakai H, Sasaki M, Ban N, and Inagaki N (2007). ABCA3-mediated choline-phospholipids uptake into intracellular vesicles in A549 cells. FEBS Lett. 581, 3139–3144. [DOI] [PubMed] [Google Scholar]
Morin A, Eisenbraun B, Key J, Sanschagrin PC, Timony MA, Ottaviano M, and Sliz P (2013). Collaboration gets the most out of software. eLife 2, e01456. [DOI] [PMC free article] [PubMed] [Google Scholar]
Nogee LM, de Mello DE, Dehner LP, and Colten HR (1993). Brief report: deficiency of pulmonary surfactant protein B in congenital alveolar proteinosis. N. Engl. J. Med 328, 406–410. [DOI] [PubMed] [Google Scholar]
Nogee LM, Wert SE, Proffit SA, Hull WM, and Whitsett JA (2000). Allelic heterogeneity in hereditary surfactant protein B (SP-B) deficiency. Am. J. Respir. Crit. Care Med 161, 973–981. [DOI] [PubMed] [Google Scholar]
Olmeda B, García-Álvarez B, and Pérez-Gil J (2013). Structure-function correlations of pulmonary surfactant protein SP-B and the saposin-like family of proteins. Eur. Biophys. J 42, 209–222. [DOI] [PubMed] [Google Scholar]
Olmeda B, García-Álvarez B, Gomez MJ, Martínez-Calle M, Cruz A and Pérez-Gil J (2015). A model for the structure and mechanism of action of pulmonary surfactant protein B. FASEB J. 29, 4236–4247. [DOI] [PubMed] [Google Scholar]
Olofsson SO, Asp L, and Boren J (1999). The assembly and secretion of apolipoprotein B-containing lipoproteins. Curr. Opin. Lipidol 10, 341–346. [DOI] [PubMed] [Google Scholar]
Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, and Ferrin TE (2004). UCSF Chimera—a visualization system for exploratory research and analysis. J. Comput. Chem 25, 1605–1612. [DOI] [PubMed] [Google Scholar]
Phillips MC (2018). Is ABCA1 a lipid transfer protein? J. Lipid Res 59, 749–763. [DOI] [PMC free article] [PubMed] [Google Scholar]
Popovic K, Holyoake J, Pomès R, and Privé GG (2012). Structure of saposin A lipoprotein discs. Proc. Natl. Acad. Sci. USA 109, 2908–2912. [DOI] [PMC free article] [PubMed] [Google Scholar]
Poulain FR, Allen L, Williams MC, Hamilton RL, and Hawgood S (1992). Effects of surfactant apolipoproteins on liposome structure: implications for tubular myelin formation. Am. J. Physiol 262, L730–L739. [DOI] [PubMed] [Google Scholar]
Powers RE, Wang S, Liu TY, and Rapoport TA (2017). Reconstitution of the tubular endoplasmic reticulum network with purified components. Nature 543, 257–260. [DOI] [PMC free article] [PubMed] [Google Scholar]
Rabouille C, and Linstedt AD (2016). GRASP: A Multitasking Tether. Front. Cell Dev. Biol 4, 1. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ridsdale R, Na C-L, Xu Y, Greis KD, and Weaver T (2011). Comparative proteomic analysis of lung lamellar bodies and lysosome-related organelles. PloS One 6, e16482. [DOI] [PMC free article] [PubMed] [Google Scholar]
Rodríguez DD, Grosse C, Himmel S, González C, de Ilarduya IM, Becker S, Sheldrick GM, and Usón I (2009). Crystallographic ab initio protein structure solution below atomic resolution. Nat. Methods 6, 651–653. [DOI] [PubMed] [Google Scholar]
Schneider CA, Rasband WS, and Eliceiri KW (2012). NIH Image to ImageJ: 25 years of image analysis. Nat. Methods 9, 671–675. [DOI] [PMC free article] [PubMed] [Google Scholar]
Sirwi A, and Hussain MM (2018). Lipid transfer proteins in the assembly of apoBcontaining lipoproteins. J. Lipid Res 59, 1094–1102. [DOI] [PMC free article] [PubMed] [Google Scholar]
Smith GB, Taeusch HW, Phelps DS, and Keough KMW (1988). Mixtures of low molecular weight surfactant proteins and dipalmitoyl phosphatidylcholine duplicate effects of pulmonary surfactant in vitro and in vivo. Pediatr. Res 23, 484–490. [DOI] [PubMed] [Google Scholar]
Stahlman MT, Gray MP, Falconieri MW, Whitsett JA, and Weaver TE (2000). Lamellar body formation in normal and surfactant protein B-deficient fetal mice. Lab. Invest 80, 395–403. [DOI] [PubMed] [Google Scholar]
Stratton CJ (1976). The three-dimensional aspect of mammalian lung multilamellar bodies. Tissue Cell 8, 693–712. [DOI] [PubMed] [Google Scholar]
Stratton CJ (1978). The ultrastructure of multilamellar bodies and surfactant in the human lung. Cell Tissue Res. 193, 219–229. [DOI] [PubMed] [Google Scholar]
Suzuki Y, Fujita Y, and Kogishi K (1989). Reconstitution of tubular myelin from synthetic lipids and proteins associated with pig pulmonary surfactant. Am. Rev. Respir. Dis 140, 75–81. [DOI] [PubMed] [Google Scholar]
Tanaka Y, Takei T, Aiba T, Masuda K, Kiuchi A, and Fujiwara T (1986). Development of synthetic lung surfactants. J. Lipid Res 27, 475–485. [PubMed] [Google Scholar]
Terwilliger TC, Grosse-Kunstleve RW, Afonine PV, Moriarty NW, Zwart PH, Hung LW, Read RJ, and Adams PD (2008). Iterative model building, structure refinement and density modification with the PHENIX AutoBuild wizard. Acta Crystallogr. D Biol. Crystallogr 64, 61–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ueno T, Linder S, Na C-L, Rice WR, Johansson J, and Weaver TE (2004). Processing of pulmonary surfactant protein B by napsin and cathepsin H. J. Biol. Chem 279, 16178–16184. [DOI] [PubMed] [Google Scholar]
Vaccaro AM, Salvioli R, Tatti M, and Ciaffoni F (1999). Saposins and their interaction with lipids. Neurochem. Res 24, 307–314. [DOI] [PubMed] [Google Scholar]
Waterhouse A, Bertoni M, Bienert S, Studer G, Tauriello G, Gumienny R, Heer FT, de Beer TAP, Rempfer C, Bordoli L, Lepore R, and Schwede T (2018). SWISS-MODEL: homology modelling of protein structures and complexes. Nucleic Acids Res. 46, W296–W303. [DOI] [PMC free article] [PubMed] [Google Scholar]
Weaver TE, Na C-L, and Stahlman M (2002). Biogenesis of lamellar bodies, lysosome-related organelles involved in storage and secretion of pulmonary surfactant. Semin. Cell Dev. Biol 13, 263–270. [DOI] [PubMed] [Google Scholar]
Wikenheiser KA, Vorbroker DK, Rice WR, Clark JC, Bachurski CJ, Oie HK, and Whitsett JA (1993). Production of immortalized distal respiratory epithelial cell lines from surfactant protein C/simian virus 40 large tumor antigen transgenic mice. Proc. Natl. Acad. Sci. USA 90, 11029–11033. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wilkins MR, Gasteiger E, Bairoch A, Sanchez JC, Williams KL, Appel RD, and Hochstrasser DF (1999). Protein identification and analysis tools in the ExPASy server. Methods Mol. Biol 112, 531–552. [DOI] [PubMed] [Google Scholar]
Williams MC, Hawgood S, and Hamilton RL (1991). Changes in lipid structure produced by surfactant proteins SP-A, SP-B, and SP-C. Am. J. Respir. Cell Mol. Biol 5, 41–50. [DOI] [PubMed] [Google Scholar]
Wong LH, Gatta AT, and Levine TP (2019). Lipid transfer proteins: the lipid commute via shuttles, bridges and tubes. Nat. Rev. Mol. Cell Biol 20, 85–101. [DOI] [PubMed] [Google Scholar]
Wu C, Chen X, Cai Y, Xia J, Zhou X, Xu S, Huang H, Zhang L, Zhou X, Du C et al. (2020). Risk factors associated with acute respiratory distress syndrome and death in patients with coronavirus disease 2019 pneumonia in Wuhan, China. JAMA Intern. Med e200994. doi: 10.1001/jamainternmed.2020.0994 [DOI] [PMC free article] [PubMed] [Google Scholar]
Xu Y, Wang S, Hu Q, Gao S, Ma X, Zhang W, Shen Y, Chen F, Lai L, and Pei J (2018). CavityPlus: a web server for protein cavity detection with pharmacophore modelling, allosteric site identification and covalent ligand binding ability prediction. Nucleic Acids Res. 46, W374–W379. [DOI] [PMC free article] [PubMed] [Google Scholar]
Yang L, Johansson J, Ridsdale R, Willander H, Fitzen M, Akinbi HT, and Weaver TE (2010). Surfactant protein B propeptide contains a saposin-like protein domain with antimicrobial activity at low pH. J. Immunol 184, 975–983. [DOI] [PMC free article] [PubMed] [Google Scholar]
Zheng H, Cooper DR, Porebski PJ, Shabalin IG, Handing KB, and Minor W (2017). CheckMyMetal: a macromolecular metal-binding validation tool. Acta Crystallogr. D Struct. Biol 73, 223–233. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

NIHMS1645341-supplement-1.pdf^{(140.5MB, pdf)}

Data Availability Statement

[R1] Adams PD, Afonine PV, Bunkoczi G, Chen VB, Davis IW, Echols N, Headd JJ, Hung LW, Kapral GJ, Grosse-Kunstleve RW, et al. (2010). PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr. D Biol. Crystallogr 66, 213–221. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R2] Aggarwal S, Yurlova L, and Simons M (2011). Central nervous system myelin: structure, synthesis and assembly. Trends Cell Biol. 21, 585–593. [DOI] [PubMed] [Google Scholar]

[R3] Ahn VE, Faull KF, Whitelegge JP, Fluharty AL, and Privé GG (2003). Crystal structure of saposin B reveals a dimeric shell for lipid binding. Proc. Natl. Acad. Sci. USA 100, 38–43. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R4] Aida T, Chiyo K, Usami T, Ishikubo H, Imahashi R, Wada Y, Tanaka KF, Sakuma T, Yamamoto T, and Tanaka K (2015). Cloning-free CRISPR/Cas system facilitates functional cassette knock-in in mice. Genome Biol. 16, 87. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R5] Akinbi HT, Breslin JS, Ikegami M, Iwamoto HS, Clark JC, Whitsett JA, Jobe AH, and Weaver TE (1997). Rescue of SP-B knockout mice with a truncated SP-B proprotein. Function of the C-terminal propeptide. J. Biol. Chem 272, 9640–9647. [DOI] [PubMed] [Google Scholar]

[R6] Beck DC, Ikegami M, Na CL, Zaltash S, Johansson J, Whitsett JA, and Weaver TE (2000). The role of homodimers in surfactant protein B function in vivo. J. Biol. Chem 275, 3365–3370. [DOI] [PubMed] [Google Scholar]

[R7] Benson BJ, Williams MC, Sueishi K, Goerke J, and Sargeant T (1984). Role of calcium ions in the structure and function of pulmonary surfactant. Biochem. Biophys. Acta 793, 18–27. [DOI] [PubMed] [Google Scholar]

[R8] Bernhard W (2016). Lung surfactant: Function and composition in the context of development and respiratory physiology. Ann. Anat 208, 146–150. [DOI] [PubMed] [Google Scholar]

[R9] Bian X, Klemm RW, Liu TY, Zhang M, Sun S, Sui X, Liu X, Rapoport TA, and Hu J (2011). Structures of the atlastin GTPase provide insight into homotypic fusion of endoplasmic reticulum membranes. Proc. Natl. Acad. Sci. USA 108, 3976–3981. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R10] Bligh EG, and Dyer WJ (1959). A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol 37, 911–917. [DOI] [PubMed] [Google Scholar]

[R11] Brasch F, Johnen G, Winn-Brasch A, Guttentag SH, Schmiedl A, Kapp N, Suzuki Y, Müller KM, Richter J, Hawgood S, et al. (2004). Surfactant protein B in type II pneumocytes and intra-alveolar surfactant forms of human lungs. Am. J. Respir. Cell Mol. Biol 30, 449–458. [DOI] [PubMed] [Google Scholar]

[R12] Bruhn H (2005). A short guided tour through functional and structural features of saposin-like proteins. Biochem. J 389, 249–257. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R13] Chander A, Dodia CR, Gil J, and Fisher AB (1983). Isolation of lamellar bodies from rat granular pneumocytes in primary culture. Biochim. Biophys. Acta 753, 119–129. [DOI] [PubMed] [Google Scholar]

[R14] Chander A, Johnson RG, Reicherter J, and Fisher AB (1986). Lung lamellar bodies maintain an acidic internal pH. J. Biol. Chem 261, 6126–6131. [PubMed] [Google Scholar]

[R15] Clark JC, Wert SE, Bachurski CJ, Stahlman MT, Stripp BR, Weaver TE, and Whitsett JA (1995). Targeted disruption of the surfactant protein B gene disrupts surfactant homeostasis, causing respiratory failure in newborn mice. Proc. Natl. Acad. Sci. USA 92, 7794–7798. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R16] Cong L, Ran FA, Cox D, Lin S, Barretto R, Habib N, Hsu PD, Wu X, Jiang W, Marrafinni LA, and Zhang F (2013). Multiplex genome engineering using CRISPR/Cas systems. Science 339, 819–823. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R17] Derman AI, Prinz WA, Belin D, and Beckwith J (1993). Mutations That Allow Disulfide Bond Formation in the Cytoplasm of Escherichia coli. Science 262, 1744–1747. [DOI] [PubMed] [Google Scholar]

[R18] Diemel RV, Snel MM, van Golde LM, Putz G, Haagsman HP, and Batenburg JJ (2002). Effects of cholesterol on surface activity and surface topography of spread surfactant films. Biochemistry 41, 15007–15016. [DOI] [PubMed] [Google Scholar]

[R19] Echaide M, Autilio C, Arroyo R, and Perez-Gil J (2017). Restoring pulmonary surfactant membranes and films at the respiratory surface. Biochim. Biophys. Acta. Biomembr 1859, 1725–1739. [DOI] [PubMed] [Google Scholar]

[R20] Emsley P, and Cowtan K (2004). Coot: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr 60, 2126–2132. [DOI] [PubMed] [Google Scholar]

[R21] Evans PR, and Murshudov GN (2013). How good are my data and what is the resolution? Acta Crystallogr. D Biol. Crystallogr 69, 1204–1214. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R22] Galietta LJ, Haggie PM, and Verkman AS (2001). Green fluorescent protein-based halide indicators with improved chloride and iodide affinities. FEBS Lett. 499, 220–224. [DOI] [PubMed] [Google Scholar]

[R23] Gerson KD, Foster CD, Zhang P, Zhang Z, Rosenblatt MM, and Guttentag SH (2008). Pepsinogen C proteolytic processing of surfactant protein B. J. Biol. Chem 283, 10330–10338. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R24] Greenspan P, and Gutman RL (1993). Detection by Nile Red of agarose gel electrophoresed native and modified low density lipoprotein. Electrophoresis 14, 65–68. [DOI] [PubMed] [Google Scholar]

[R25] Kabsch W (2010). Xds. Acta Crystallogr. D Biol. Crystallogr 66, 125–132. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R26] King RJ (1982). Pulmonary surfactant. J. Appl. Physiol. Respir. Environ. Exerc. Physiol 53, 1–8. [DOI] [PubMed] [Google Scholar]

[R27] Kolter T, and Sandhoff K (2005). Principles of lysosomal membrane digestion: stimulation of sphingolipid degradation by sphingolipid activator proteins and anionic lysosomal lipids. Annu. Rev. Cell Dev. Biol 21, 81–103. [DOI] [PubMed] [Google Scholar]

[R28] Korimilli A, Gonzales LW, and Guttentag SH (2000). Intracellular localization of processing events in human surfactant protein B biosynthesis. J. Biol. Chem 275, 8672–8679. [DOI] [PubMed] [Google Scholar]

[R29] Labun K, Montague TG, Krause M, Torres Cleuren YN, Tjeldsen H and Valen E (2019). CHOPCHOP v3: expanding the CRISPR web toolbox beyond genome editing. Nucleic Acids Res. 47, W171–W174. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R30] Li Y, Kinting S, Hoppner S, Forstner ME, Uhl O, Koletzko B, and Griese M (2019). Metabolic labelling of choline phospholipids probes ABCA3 transport in lamellar bodies. Biochim. Biophys. Acta Mol. Cell Biol. Lipids 1864, 158516. [DOI] [PubMed] [Google Scholar]

[R31] Liekkinen J, Enkavi G, Javanainen M, Olmeda B, Perez-Gil J and Vattulainen I (2020). Pulmonary surfactant lipid reorganization induced by the adsorption of the oligomeric surfactant protein B complex. J. Mol. Biol 432, 3251–3268. [DOI] [PubMed] [Google Scholar]

[R32] Lin S, Akinbi HT, Breslin JS, and Weaver TE (1996a). Structural requirements for targeting of surfactant protein B (SP-B) to secretory granules in vitro and in vivo. J. Biol. Chem 271, 19689–19695. [DOI] [PubMed] [Google Scholar]

[R33] Lin S, Phillips KS, Wilder MR, and Weaver TE (1996b). Structural requirements for intracellular transport of pulmonary surfactant protein B (SP-B). Biochim. Biophys. Acta 1312, 177–185. [DOI] [PubMed] [Google Scholar]

[R34] Lopez-Rodriguez E, Gay-Jordi G, Mucci A, Lachmann N, and Serrano-Mollar A (2017). Lung surfactant metabolism: early in life, early in disease and target in cell therapy. Cell Tissue Res. 367, 721–735. [DOI] [PubMed] [Google Scholar]

[R35] Martínez-Calle M, Prieto M, Olmeda B, Fedorov A, Loura LMS and Pérez-Gil J (2020). Pulmonary surfactant protein SP-B nanorings induce the multilamellar organization of surfactant complexes. Biochim. Biophys. Acta Biomembr 1862, 183216. [DOI] [PubMed] [Google Scholar]

[R36] Mathialagan N, and Possmayer F (1990). Low-molecular-weight hydrophobic proteins from bovine pulmonary surfactant. Biochim. Biophys. Acta 1045, 121–127. [DOI] [PubMed] [Google Scholar]

[R37] Matsumura Y, Sakai H, Sasaki M, Ban N, and Inagaki N (2007). ABCA3-mediated choline-phospholipids uptake into intracellular vesicles in A549 cells. FEBS Lett. 581, 3139–3144. [DOI] [PubMed] [Google Scholar]

[R38] Morin A, Eisenbraun B, Key J, Sanschagrin PC, Timony MA, Ottaviano M, and Sliz P (2013). Collaboration gets the most out of software. eLife 2, e01456. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R39] Nogee LM, de Mello DE, Dehner LP, and Colten HR (1993). Brief report: deficiency of pulmonary surfactant protein B in congenital alveolar proteinosis. N. Engl. J. Med 328, 406–410. [DOI] [PubMed] [Google Scholar]

[R40] Nogee LM, Wert SE, Proffit SA, Hull WM, and Whitsett JA (2000). Allelic heterogeneity in hereditary surfactant protein B (SP-B) deficiency. Am. J. Respir. Crit. Care Med 161, 973–981. [DOI] [PubMed] [Google Scholar]

[R41] Olmeda B, García-Álvarez B, and Pérez-Gil J (2013). Structure-function correlations of pulmonary surfactant protein SP-B and the saposin-like family of proteins. Eur. Biophys. J 42, 209–222. [DOI] [PubMed] [Google Scholar]

[R42] Olmeda B, García-Álvarez B, Gomez MJ, Martínez-Calle M, Cruz A and Pérez-Gil J (2015). A model for the structure and mechanism of action of pulmonary surfactant protein B. FASEB J. 29, 4236–4247. [DOI] [PubMed] [Google Scholar]

[R43] Olofsson SO, Asp L, and Boren J (1999). The assembly and secretion of apolipoprotein B-containing lipoproteins. Curr. Opin. Lipidol 10, 341–346. [DOI] [PubMed] [Google Scholar]

[R44] Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, and Ferrin TE (2004). UCSF Chimera—a visualization system for exploratory research and analysis. J. Comput. Chem 25, 1605–1612. [DOI] [PubMed] [Google Scholar]

[R45] Phillips MC (2018). Is ABCA1 a lipid transfer protein? J. Lipid Res 59, 749–763. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R46] Popovic K, Holyoake J, Pomès R, and Privé GG (2012). Structure of saposin A lipoprotein discs. Proc. Natl. Acad. Sci. USA 109, 2908–2912. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R47] Poulain FR, Allen L, Williams MC, Hamilton RL, and Hawgood S (1992). Effects of surfactant apolipoproteins on liposome structure: implications for tubular myelin formation. Am. J. Physiol 262, L730–L739. [DOI] [PubMed] [Google Scholar]

[R48] Powers RE, Wang S, Liu TY, and Rapoport TA (2017). Reconstitution of the tubular endoplasmic reticulum network with purified components. Nature 543, 257–260. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R49] Rabouille C, and Linstedt AD (2016). GRASP: A Multitasking Tether. Front. Cell Dev. Biol 4, 1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R50] Ridsdale R, Na C-L, Xu Y, Greis KD, and Weaver T (2011). Comparative proteomic analysis of lung lamellar bodies and lysosome-related organelles. PloS One 6, e16482. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R51] Rodríguez DD, Grosse C, Himmel S, González C, de Ilarduya IM, Becker S, Sheldrick GM, and Usón I (2009). Crystallographic ab initio protein structure solution below atomic resolution. Nat. Methods 6, 651–653. [DOI] [PubMed] [Google Scholar]

[R52] Schneider CA, Rasband WS, and Eliceiri KW (2012). NIH Image to ImageJ: 25 years of image analysis. Nat. Methods 9, 671–675. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R53] Sirwi A, and Hussain MM (2018). Lipid transfer proteins in the assembly of apoBcontaining lipoproteins. J. Lipid Res 59, 1094–1102. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R54] Smith GB, Taeusch HW, Phelps DS, and Keough KMW (1988). Mixtures of low molecular weight surfactant proteins and dipalmitoyl phosphatidylcholine duplicate effects of pulmonary surfactant in vitro and in vivo. Pediatr. Res 23, 484–490. [DOI] [PubMed] [Google Scholar]

[R55] Stahlman MT, Gray MP, Falconieri MW, Whitsett JA, and Weaver TE (2000). Lamellar body formation in normal and surfactant protein B-deficient fetal mice. Lab. Invest 80, 395–403. [DOI] [PubMed] [Google Scholar]

[R56] Stratton CJ (1976). The three-dimensional aspect of mammalian lung multilamellar bodies. Tissue Cell 8, 693–712. [DOI] [PubMed] [Google Scholar]

[R57] Stratton CJ (1978). The ultrastructure of multilamellar bodies and surfactant in the human lung. Cell Tissue Res. 193, 219–229. [DOI] [PubMed] [Google Scholar]

[R58] Suzuki Y, Fujita Y, and Kogishi K (1989). Reconstitution of tubular myelin from synthetic lipids and proteins associated with pig pulmonary surfactant. Am. Rev. Respir. Dis 140, 75–81. [DOI] [PubMed] [Google Scholar]

[R59] Tanaka Y, Takei T, Aiba T, Masuda K, Kiuchi A, and Fujiwara T (1986). Development of synthetic lung surfactants. J. Lipid Res 27, 475–485. [PubMed] [Google Scholar]

[R60] Terwilliger TC, Grosse-Kunstleve RW, Afonine PV, Moriarty NW, Zwart PH, Hung LW, Read RJ, and Adams PD (2008). Iterative model building, structure refinement and density modification with the PHENIX AutoBuild wizard. Acta Crystallogr. D Biol. Crystallogr 64, 61–69. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R61] Ueno T, Linder S, Na C-L, Rice WR, Johansson J, and Weaver TE (2004). Processing of pulmonary surfactant protein B by napsin and cathepsin H. J. Biol. Chem 279, 16178–16184. [DOI] [PubMed] [Google Scholar]

[R62] Vaccaro AM, Salvioli R, Tatti M, and Ciaffoni F (1999). Saposins and their interaction with lipids. Neurochem. Res 24, 307–314. [DOI] [PubMed] [Google Scholar]

[R63] Waterhouse A, Bertoni M, Bienert S, Studer G, Tauriello G, Gumienny R, Heer FT, de Beer TAP, Rempfer C, Bordoli L, Lepore R, and Schwede T (2018). SWISS-MODEL: homology modelling of protein structures and complexes. Nucleic Acids Res. 46, W296–W303. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R64] Weaver TE, Na C-L, and Stahlman M (2002). Biogenesis of lamellar bodies, lysosome-related organelles involved in storage and secretion of pulmonary surfactant. Semin. Cell Dev. Biol 13, 263–270. [DOI] [PubMed] [Google Scholar]

[R65] Wikenheiser KA, Vorbroker DK, Rice WR, Clark JC, Bachurski CJ, Oie HK, and Whitsett JA (1993). Production of immortalized distal respiratory epithelial cell lines from surfactant protein C/simian virus 40 large tumor antigen transgenic mice. Proc. Natl. Acad. Sci. USA 90, 11029–11033. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R66] Wilkins MR, Gasteiger E, Bairoch A, Sanchez JC, Williams KL, Appel RD, and Hochstrasser DF (1999). Protein identification and analysis tools in the ExPASy server. Methods Mol. Biol 112, 531–552. [DOI] [PubMed] [Google Scholar]

[R67] Williams MC, Hawgood S, and Hamilton RL (1991). Changes in lipid structure produced by surfactant proteins SP-A, SP-B, and SP-C. Am. J. Respir. Cell Mol. Biol 5, 41–50. [DOI] [PubMed] [Google Scholar]

[R68] Wong LH, Gatta AT, and Levine TP (2019). Lipid transfer proteins: the lipid commute via shuttles, bridges and tubes. Nat. Rev. Mol. Cell Biol 20, 85–101. [DOI] [PubMed] [Google Scholar]

[R69] Wu C, Chen X, Cai Y, Xia J, Zhou X, Xu S, Huang H, Zhang L, Zhou X, Du C et al. (2020). Risk factors associated with acute respiratory distress syndrome and death in patients with coronavirus disease 2019 pneumonia in Wuhan, China. JAMA Intern. Med e200994. doi: 10.1001/jamainternmed.2020.0994 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R70] Xu Y, Wang S, Hu Q, Gao S, Ma X, Zhang W, Shen Y, Chen F, Lai L, and Pei J (2018). CavityPlus: a web server for protein cavity detection with pharmacophore modelling, allosteric site identification and covalent ligand binding ability prediction. Nucleic Acids Res. 46, W374–W379. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R71] Yang L, Johansson J, Ridsdale R, Willander H, Fitzen M, Akinbi HT, and Weaver TE (2010). Surfactant protein B propeptide contains a saposin-like protein domain with antimicrobial activity at low pH. J. Immunol 184, 975–983. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R72] Zheng H, Cooper DR, Porebski PJ, Shabalin IG, Handing KB, and Minor W (2017). CheckMyMetal: a macromolecular metal-binding validation tool. Acta Crystallogr. D Struct. Biol 73, 223–233. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Mechanism of Lamellar Body Formation by Lung Surfactant Protein B

Navdar Sever

Goran Miličić

Nicholas O Bodnar

Xudong Wu

Tom A Rapoport

SUMMARY

Graphical Abstract

eTOC:

INTRODUCTION

RESULTS

SP-BN is a lipid-binding protein

Table 1.

Figure 1. Crystal structures of SP-BN.

Lipid transfer activity of SP-BN

Figure 2. Phospholipid transfer by SP-BN.

Molecular mechanism of phospholipid transfer by SP-BN

The SP-BN domain in proSP-B transfers lipids to form lipoprotein particles

Figure 3. ProSP-B has enhanced lipid transfer activity.

Figure 4. ProSP-B forms lipoprotein particles.

Lipid-binding and -transfer by the SP-BN domain determines intracellular trafficking of proSP-B

Figure 5. Lipid-binding and -transfer by SP-BN is required for proSP-B trafficking in alveolar cells.

Testing proSP-B mutants in mice

Figure 6. Testing proSP-B mutants in mice.

The SP-BC domain facilitates ER export of proSP-B

Reconstituted SP-BM generates LB-like structures

Figure 7. Reconstituted SP-BM generates LB-like structures.

DISCUSSION

STAR * METHODS

RESOURCE AVAILABILITY

Lead Contact

Materials Availability

Data and Code Availability

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Cells

Mice

METHOD DETAILS

Cloning

Protein purifications

Lipid transfer assays

Reconstitution of SP-BM into liposomes

Thin-section EM

Negative-stain EM

Experiments with MLE-15 cells

Lamellar body enrichment

Immunofluorescence

Experiments with mice

Lipid analysis by mass spectrometry

CD spectroscopy

Analysis of lipoprotein particles

ProSP-B processing in vitro

Crystallization

Structure determination

QUANTIFICATION AND STATISTICAL ANALYSIS

Basic quantifications

Statistical details

Supplementary Material

Highlights:

ACKNOWLEDGMENTS

Footnotes

REFERENCES

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases