Role of the N-Terminal Seven Residues of Surfactant Protein B (SP-B)

Breathing is enabled by lung surfactant, a mixture of proteins and lipids that forms a surface-active layer and reduces surface tension at the air-water interface in lungs. Surfactant protein B (SP-B) is an essential component of lung surfactant. In this study we probe the mechanism underlying the important functional contributions made by the N-terminal 7 residues of SP-B, a region sometimes called the “insertion sequence”. These studies employed a construct of SP-B, SP-B (1–25,63–78), also called Super Mini-B, which is a 41-residue peptide with internal disulfide bonds comprising the N-terminal 7-residue insertion sequence and the N- and C-terminal helices of SP-B. Circular dichroism, solution NMR, and solid state 2H NMR were used to study the structure of SP-B (1–25,63–78) and its interactions with phospholipid bilayers. Comparison of results for SP-B (8–25,63–78) and SP-B (1–25,63–78) demonstrates that the presence of the 7-residue insertion sequence induces substantial disorder near the centre of the lipid bilayer, but without a major disruption of the overall mechanical orientation of the bilayers. This observation suggests the insertion sequence is unlikely to penetrate deeply into the bilayer. The 7-residue insertion sequence substantially increases the solution NMR linewidths, most likely due to an increase in global dynamics.


Introduction
Lung surfactant is a complex of lipids and proteins that lines the air-water interface at the alveolar surface. It is essential for reducing surface tension and preventing alveolar collapse [1,2]. Lung surfactant has a complex composition and forms dynamic and intricate two dimensional and three dimensional structures, with a monolayer at the interface as well as associated multilayer structures underneath [3][4][5][6][7]. Deficiency or inactivation of lung surfactant leads to potentially lethal respiratory disorders such as neonatal respiratory distress syndrome (NRDS) in premature newborns [8][9][10] and acute respiratory distress syndrome (ARDS) in patients with severe injury or illness [11][12][13]. Surfactant replacement therapy has been quite successful in treating NRDS [14,15]. However, efforts to use replacement surfactant to treat ARDS have not demonstrated improvements in mortality rates thus far [9,12].
Approximately 90% by weight of surfactant is lipid, mainly phosphocholine (PC) and phosphoglycerol (PG), and 10% is surfactant proteins (SPs) [3,[16][17][18]. SPs are designated by their chronologic order of discovery as SP-A, SP-B, SP-C, and SP-D [19]. SP-A and SP-D are water soluble and important for host defense [20,21], whereas SP-B and SP-C are smaller, hydrophobic proteins that are critical for reducing surface tension during breathing [22,23]. SP-B is the only essential protein component of lung surfactant, as evidenced by the lethality of hereditary SP-B deficiency in humans [24,25] and the lethal effect of knocking out the SP-B gene in mice [26].
SP-B is a highly conserved member of the saposin superfamily of proteins and thus expected to possess 4 to 5 helices [23,27]. It is found in the lung as a covalently linked homodimer, with 79 amino acid residues in each monomer. Three intramolecular disulfide bonds are formed by six cysteines, and a seventh cysteine forms an intermolecular bridge to stabilize the SP-B homodimer structure. SP-B has a large proportion (52%) of hydrophobic amino acids, and also has cationic characteristics with a net charge of +7 (per monomer) at neutral pH. Its positive charge and highly hydrophobic nature are thought to provide the basis for interactions between SP-B and the negatively charged lipid components of lung surfactant. A number of potential mechanisms for SP-B have been proposed including providing a link between the monolayer and underlying bilayers, stabilizing lipid structures required for lowering surface tension, promoting transfer of lipids into and out of functional lipid structures, and promoting interfacial adsorption of surfactant from the hypophase to the airwater interface [6,7,[28][29][30][31]. However, SP-B's three-dimensional structure has not yet been determined, which is an impediment to understanding its detailed mechanism.
In an alternate approach, a number of studies have addressed structure/function relationships in fragments of SP-B that retain a substantial portion of the function of the full length protein [32][33][34][35][36]. NMR structures of some of the individual helices of SP-B have been determined [37,38]. A structure has also been determined for a larger fragment, termed Mini-B, which retains two of SP-B's four helices (residues 8-25 and 63-78) and much of its function, as assessed by measurements in surfactant-deficient rats [39,40].
The region of SP-B's N-terminus preceding the first helix, termed the ''insertion sequence'' [41,42], is of particular interest. The segment of SP-B comprising residues 1-7, Phe-Pro-Ile-Pro-Leu-Pro-Tyr, resembles proline-rich cell-penetrating peptides [43]. Proline is very singular amongst the set of natural amino acids in that it lacks a backbone HN group to form hydrogen bonds, has a ring structure that restricts its allowable backbone dihedral angles, and has a propensity to form special secondary structures, such as polyproline II helix. The functional role of the insertion sequence has been probed in the context of N-terminal peptides of SP-B and it was found that mutation of any of the proline residues led to decreased surface activity [44]. More recently, the insertion sequence has been examined by adding it on to Mini-B, to create a construct of SP-B, termed ''Super Mini-B'' (SP-B (1-25,63-78)), which retains the N-terminal insertion sequence, two of the four SP-B helices, specifically the N-terminal helix and the C-terminal helix, the two disulfide bonds that help link the helices together, and an overall charge of +7 [41,42]. The 7 insertion sequence residues, along with the tryptophan at position 9, were proposed to stabilize the formation of ''nanosilos'', structures seen by atomic force microscopy imaging of monolayers deposited at high surface pressures [41]. In this study, we further probe the role of SP-B residues 1-7, by comparing the structure and lipid interactions of the construct that includes the insertion sequence, SP-B (1-25, 63-78) (Super Mini-B), with those of a construct lacking this sequence, SP-B (8-25,63-78) (Mini-B), as well as by characterizing the structural features of SP-B (1-7) by itself.

Solid-state NMR
2 H NMR was used to observe the perturbation of mechanicallyoriented phospholipid bilayers by SP-B (1-25,63-78). To prepare oriented samples, 1 mol% peptide, if present, was co-dissolved with 4 mg of a 7:3 (molar ratio) mixture of POPC-d 31 and POPG in a mixture of CH 3 OH/CHCl 3 (1:1 by volume). The solution, comprising a total volume of 250 ml, was spread onto 12 mica plates (12 mm by 12 mm) by depositing ,1 ml at the centre of each plate, allowing 3-5 minutes for spreading and drying, and then repeating until the deposition of the full volume was complete. The films on the mica plates were then dried for 2 hours in a fume hood after which residual solvent was removed by exposure to vacuum for ,8 hours. Films were then hydrated by spreading 5 mL of deuterium-depleted water onto each plate and then leaving the plates in a hydration chamber, along with saturated ammonia phosphate solution, at 4uC for 2 days. The plates were then carefully stacked, wrapped with plastic film, and sealed in heavy plastic wrap. Samples were stored at 4uC before the NMR experiments.
2 H-NMR spectra were obtained using a locally-assembled spectrometer operating at 61.4 MHz. The oriented samples were positioned in a flat coil (15 mm 615 mm 63 mm) with the bilayer normal parallel to the magnetic field of a 9.4 T superconducting solenoid. Spectra were derived from free-induction decays obtained by averaging 60000 transients accumulated using a quadrupole echo sequence [45] with a p/2 pulse length of 4.1 ms and 30 ms pulse separation. Transients were acquired using a digitizer dwell time of 1 ms and oversampling [46] by a factor of 4 to give an effective dwell time of 4 ms. All 2 H-NMR spectra were acquired at 23uC.
The 2 H-NMR spectrum of POPC-d 31 in a liquid crystalline bilayer oriented with its normal at an angle b with respect to the applied magnetic field is a superposition of spectral doublets with quadrupole splittings given by where e 2 qQ h~1 67kHz is the quadrupole coupling constant for carbon-deuterium bonds and S i CD is the orientational order parameter for deuterons on the acyl chain methylene group denoted by i. For a given carbon-deuterium bond, the orientational order parameter is given by where h is the angle between the direction of the carbondeuterium bond and the bilayer normal which is the symmetry axis for fast, axially symmetric reorientation of the lipid acyl chain in the liquid crystalline phase. The average is over all motions and chain conformational changes that modulate the quadrupole interaction on the approximately 10 25 s timescale of the 2 H-NMR. For a saturated phospholipid acyl chain, the orientational order parameter is largest at the headgroup end of the chain, where motions are most constrained, and decreases with position along the chain to the methyl group near the bilayer centre where motions are least constrained. The dependence of orientational order parameter on position along the chain is characterized by the orientational order parameter profile [47,48].
For a mechanically-oriented sample with bilayer normal parallel to the magnetic field, b = 0u and the spectrum is a superposition of sharp doublets with quadrupole splittings given by For unoriented bilayers, such as in a multilamellar vesicle sample, the bilayer normal directions are spherically distributed and the spectrum is a superposition of Pake doublets with prominent edges, for each position along the chain, corresponding  Table 1. Splittings, D n, and corresponding order parameters, S CD , for the resolved peaks of the 2H NMR spectra of POPC-d 31      to b = 90u and thus half the splitting of the corresponding doublet in the oriented bilayer spectrum.

Circular dichroism (CD)
The The CD spectra were acquired on a Jasco J-810 spectropolarimeter (Applied Photophysics, UK). The absorbance at 222 nm was checked for each sample to ensure that it did not exceed an optical density of 1.0. A 1 mm cell was used and spectra were collected between wavelengths of 200 and 260 nm at 308 K. Baselines were established using the appropriate buffer and the means of 4 spectra were analyzed. Secondary structure content was calculated from the spectra using CDPro software (http://lamar.colstate.edu/ ssreeram/CDPro) developed by Woody and co-workers (Sreerama and Woody, 1993). CD values were converted to mean residue ellipticity (MRE). Basis set 2 of the CDPro software was used and analysis was performed using CONTIN/LL method [49].

Solution NMR
The same samples used for the CD observations of SP-B (1-25,63-78) were also employed in the solution NMR experiments. Data was acquired on a Bruker Avance 600 MHz spectrometer equipped with z-gradients and a triple-resonance TXI probe. The The chemical shifts were referenced with respect to an internal DSS standard (0.0 ppm). The data was processed with iNMR (http://www.inmr.net) and analyzed using Sparky [50].
Diffusion-ordered spectroscopy (DOSY) experiments were performed on the same Bruker Avance 600 MHz spectrometer employing pulsed field gradient (PFG) NMR [51] with the same SP-B (1-25, 63-78) SDS sample. The pulse sequence used a stimulated echo with bipolar gradient pulses and one spoil gradient [52] followed by a 3-9-19 pulse for water suppression [53]. The 1 H signals were attenuated to ,5% of their initial amplitudes by increasing the gradient strength from ,2% to 95% in 32 steps. Experiments were performed at 37uC. The pseudo 2D DOSY spectra were processed using iNMR and the diffusion constants extracted using the DECRA package of DOSYToolbox [54].

H NMR of SP-B (1-25,63-78) in oriented lipid bilayers
The effects of SP-B (1-25,63-78) on lipid bilayer orientation and chain orientational order were investigated by 2 H NMR. Figure 1 shows spectra of POPC-d 31 /POPG (7:3) with and without SP-B (1-25,63-78). In the absence of peptide, the doublets are sharp and well resolved, indicating that the lipid bilayers are well aligned. With the incorporation of 1 mol% SP-B (1-25,63-78), the distribution of spectral area across the spectrum shifts in a way that suggests the emergence of a weak spectral component corresponding to a more random distribution of bilayer normal directions -i.e. Pake quadrupole doublets split by half of the corresponding oriented sample doublet splittings. Doublets in the oriented component of the spectrum also broaden in a way consistent with increased mosaic spread [55]. The presence of SP-B (1-25,63-78) thus appears to disrupt the mechanical orientation in a small fraction of the bilayer material.
There are two possible explanations for the greater linewidth of 1D and 2D NMR spectra of SP-B (1-25,63-78) compared to the spectra from SP-B (8-25,63-78). One is that there is an increase in the size of the SP-B (1-25,63-78)/SDS micelle complexes being observed, compared to the complexes formed with SP-B (8-25,63-78). Alternatively, the broader lines could derive from conformational fluctuations on the millisecond to microsecond timescale [59].
The ability of the N-terminal 7 residues of SP-B to disrupt order deep within the bilayer could be an important component of SP-B's suggested function in mediating transfer of lipids between the surfactant monolayer at the surface and the underlying multilayers. Mutation of the N-terminal prolines to alanines in SP-B  leads to less effective reinsertion of surface-active material into the expanding interface film [44] and deletion of the insertion sequence from SP-B , reduces this peptide's ability to promote the formation of a fluid isotropic phase in the lipids [62].
When a peptide substantially reduces the orientational order of the lipid chains, in particular deep in the bilayer, this indicates that the peptide locates close to the polar/apolar interface [61,63]. This is because a peptide positioned close to the lipid head groups can increase the spacing between lipids without constraining the motions of the acyl chains, that, with the increase in headgroup spacing, have more orientational freedom. By contrast, peptides that insert deeply into the bilayer have little effect on acyl chain order [63,64] because the peptide itself confers stearic constraints on acyl chain motions. That Super Mini-B (SP-B (1-25,63-78)) is so effective at disrupting the order deep in the bilayer interior is thus an indication that the N-terminal 7 residues likely do not penetrate deeply enough to constrain the amplitude of lipid acyl chain reorientation. This is not because the 7 residues cannot form a long enough structure; for an extended conformation such as a b-strand, the translation per residue is about 3.5 Å [65] and so 7 residues could, in theory extend out to 24.5 Å , which is close to the width of the hydrophobic region of a lipid bilayer. Instead, it appears that the insertion sequence residues preferentially position closer to the surface of the bilayer. Consistent with this interpretation, fluorescence quenching experiments found that residues 1-6 of SP-B (1-25) locate near the bilayer surface, while residues 7-9 insert more deeply [66].  [42]. However, diffusion NMR measurements ( Figure 6) indicate that the size of the SP-B (1-25,63-78)/SDS complexes is similar to that previously measured for SP-B (8-25,63-78) complexes and titration experiments also failed to provide evidence of peptide self-association. Thus, the most likely explanation for the increase in linewidth is a change in the global dynamics of SP-B (1-25,63-78) compared to SP-B (8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(63)(64)(65)(66)(67)(68)(69)(70)(71)(72)(73)(74)(75)(76)(77)(78)). An increase in dynamics on the millisecond to microsecond timescale would account for the increase in line widths [59] and would also be consistent with the increase in motional flexibility of the lipid chains observed in the 2 H-NMR experiments.
The insertion sequence contains three proline residues and thus we wanted to explore the possibility that a polyproline helical structure is formed. The circular dichroism data does not clearly indicate what type of structure this 7-residue piece formsalthough as expected the CD spectra do exclude aor 3 10 helix (Table 2, Figure 3). The CD spectra are not inconsistent with the 7 residues taking on a polyproline helical type secondary structure in that while the spectrum of SP-B (8-25,63-78) exhibits a canonical helical shape, with minima at 209 nm and 220 nm, the spectrum of SP-B (1-25,63-78) displays an altered shape with a slight shift of the 209 minimum to a shorter wavelength and a less pronounced minimum at 220 nm. However, the deconvolution of the CD spectra ( Table 2) suggests the insertion sequence could also consist of a large portion of b and/or unordered structures.
In summary, we report that the 7-residue insertion sequence of SP-B does not substantially disrupt the helical region of the peptide, but does cause alterations in the solution NMR linewidth, most likely via alterations in the global mobility of the peptide We also find that the inclusion of the insertion sequence leads to dramatic increases in acyl chain disorder in the centre of the bilayer, as compared to the more modest degree of disorder produced by SP-B peptides lacking the insertion sequence. This degree of acyl chain disruption is consistent with a location of the insertion sequence close to the bilayer surface and is likely important in promoting the bilayer disruptions important to SP-B's role in adsorption and re-spreading of lung surfactant.