Aerosol delivery of dry powder synthetic lung surfactant to surfactant-deficient rabbits and preterm lambs on non-invasive respiratory support

Background: The development of synthetic lung surfactant for preterm infants has focused on peptide analogues of native surfactant proteins B and C (SP-B and SP-C). Non-invasive respiratory support with nasal continuous positive airway pressure (nCPAP) may benefit from synthetic surfactant for aerosol delivery. Methods: A total of three dry powder (DP) surfactants, consisting of phospholipids and the SP-B analogue Super Mini-B (SMB), and one negative control DP surfactant without SMB, were produced with the Acorda Therapeutics ARCUS® Pulmonary Dry Powder Technology. Structure of the DP surfactants was compared with FTIR spectroscopy, in vitro surface activity with captive bubble surfactometry, and in vivo activity in surfactant-deficient adult rabbits and preterm lambs. In the animal experiments, intratracheal (IT) aerosol delivery was compared with surfactant aerosolization during nCPAP support. Surfactant dosage was 100 mg/kg of lipids and aerosolization was performed using a low flow inhaler. Results: FTIR spectra of the three DP surfactants each showed secondary structures compatible with peptide folding as an α-helix hairpin, similar to that previously noted for surface-active SMB in other lipids. The DP surfactants with SMB demonstrated in vitro surface activity <1 mN/m. Oxygenation and lung function increased quickly after IT aerosolization of DP surfactant in both surfactant-deficient rabbits and preterm lambs, similar to improvements seen with clinical surfactant. The response to nCPAP aerosol delivery of DP surfactant was about 50% of IT aerosol delivery, but could be boosted with a second dose in the preterm lambs. Conclusions: Aerosol delivery of DP synthetic surfactant during non-invasive respiratory support with nCPAP significantly improved oxygenation and lung function in surfactant-deficient animals and this response could be enhanced by giving a second dose. Aerosol delivery of DP synthetic lung surfactant has potential for clinical applications.


Introduction
Lung surfactant is a lipid-protein mixture that is synthesized by alveolar type II cells and secreted into the alveoli to reduce surface tension at the air-liquid interface. Lack or dysfunction of lung surfactant leads to respiratory failure, as seen in very preterm infants with respiratory distress syndrome (RDS) due to lung immaturity and surfactant deficiency and in (near-) term infants with surfactant dysfunction due to meconium aspiration syndrome. Mammalian lung surfactant harvested by lavage or lung extraction has been instrumental in reducing mortality and morbidity in preterm infants with RDS 1 . However, animal surfactant has some major drawbacks, such as limited supplies, complex production and sterilization processes, high production costs, and the need for intratracheal (IT) delivery. This has stimulated the design and testing of synthetic lung surfactant formulations with peptide mimics of the pivotal surfactant protein B (SP-B) and the less essential surfactant protein C (SP-C) 2 . The realization that the etiology of chronic lung injury (a.k.a. bronchopulmonary dysplasia, BPD) is multifactorial and intubation and mechanical ventilation are associated with a higher risk of BPD has led to a preference for non-invasive respiratory support with nasal continuous positive airway pressure (nCPAP) in preterm infants 3 . Therefore, clinical neonatal care might benefit from a synthetic lung surfactant formulation that can be delivered by inhalation and without the need for endotracheal intubation of a preterm infant.
Using 3D-saposin motifs for SP-B predicted from homology modeling, we produced truncated SP-B peptide mimics with an α-helix hairpin structure and high surface activity in surfactantdeficient animals 2,4 . One of these 'short-cut' peptides is Super Mini-B (SMB), a 41-residue SP-B mimic ( Figure 1) based on the amino-acid sequence, structure, folding and association of native SP-B (79-residues). Surface activity of SMB is similar to that of its parent protein 4-6 . SMB consists of the N-terminal (~residues 1-25) and C-terminal α-helices (~residues 63-78) of native SP-B ( Figure 1) and is connected with a -PKGG-turn to form a α-helix hairpin (α-helix/turn/α1-helix, a.k.a. αtα) 7 . SMB has two disulfide bonds that lie close together (i.e., Cys-8 to Cys-77 and Cys-11 to Cys-71) and covalently link both α-helices to reinforce the formation of an α-helix hairpin. The structure of SMB has been confirmed with spectroscopy, homology modeling and molecular dynamics simulations in lipid mimics and lipid bilayers 5,6,8 . Formulated with surfactant lipids, SMB has demonstrated stability of the α-helix hairpin and continuous high surface activity in fresh and stored preparations 4,5,8 . In this context, it should be noted that previous work demonstrated that aerosol delivery of synthetic lung surfactant with SP-B and SP-C peptides could improve oxygenation and lung function in spontaneously breathing, lavaged, surfactant-deficient rabbits supported with nCPAP 9 . Nevertheless, the above highly active surfactant formulations were all liquids, may require cold-chain technology to ensure shelf-life and may therefore not be practical in low-resource settings.
In the present study, we report the development of surface-active dry powder (DP) synthetic lung surfactant formulations, their in vitro surface activity in the captive bubble surfactometer and their in vivo surface activity following aerosol delivery in spontaneously breathing, surfactant-deficient rabbits and preterm lambs supported with nCPAP. These DP surfactants consisted of 2-3 phospholipids combined with a SP-B peptide mimic (SMB) and necessary additives, and one DP surfactant formulation without SMB as negative control. Although SMB has primarily been studied in liquids (see above), our finding that SMB maintains its engineered α-helix turn in a wide range of environments 5,8 suggests that this SP-B mimic may also retain its structure and surface activity in the anhydrous state characteristic of dry powder aerosols.

Methods
Preparation of DP synthetic lung surfactant formulations for aerosol delivery DP synthetic surfactant powders were designed and formulated at Acorda Therapeutics, utilizing its ARCUS ® Pulmonary Dry Powder Technology. The challenge that was overcome via the utilization of the ARCUS ® technology was to produce phospholipid-based DP lung surfactant formulations, with and without SP-B peptide mimetic, that both replicated the fluidizing surface activity of endogenous lung surfactant yet were both physically and chemically stable as well as dispersible and aerosolizable in a solid-state DP form, in contrast to the liquidbased forms currently utilized for surfactant replacement therapy. An extensive screening process 10 was conducted utilizing various combinations of phospholipids, fatty acids and stabilizing excipients that resulted in a lead set of optimized peptide-free candidate DP synthetic surfactant powders. SP-B peptide mimiccontaining formulations were later produced utilizing this lead set of powders. Lead DP synthetic surfactant powders were formulated by adding the following components to the organic solvent used for spray drying: 49 wt% DPPC and 21 wt% POPG-Na (Avanti Polar Lipids, Alabaster, AL); an optional third phospholipid, e.g., 7 wt% POPC (Avanti Polar Lipids, Alabaster, . SMB (41 amino-acid residues; 1-letter amino-acid notation) with the N-and C-terminal Phe-1 and Ser-41 are indicated, and also the sulfur-containing cysteines (Cys-8, Cys-11, Cys-34, Cys-40) and methionines (Met-21 and Met-28) in red. The two disulfide-linkages are shown between Cys-11 and Cys-34 and between Cys-8 and Cys-40.

Amendments from Version 1
In this revision, we implemented the thoughtful suggestions of the reviewers. We also improved the layout of Figure 2 that shows the FTIR spectra of Super Mini-B (SMB) in dry powder (DP) synthetic surfactant lipid formulations and of Figure 5 that indicates lung function of lavaged, surfactant-deficient rabbits treated with aerosolized DP synthetic surfactants.
SMB was synthesized at the Los Angeles Biomedical Research Institute using a standard Fmoc protocol with a Symphony Multiple Peptide Synthesizer (Protein Technologies, Inc., Tucson, AZ 87514) or a CEM Liberty microwave synthesizer (CEM Corporation, Mathews, NC 28104), cleaved-deprotected, purified using reverse phase HPLC, freeze-dried, and had its mass confirmed by MALDI TOF mass spectrometry and concentration quantitated using UV absorbance 4,5 .
Respirable dry synthetic surfactant particles were produced at Acorda Therapeutics using a GEA Niro PSD-1 spray dryer (Niro Inc., Copenhagen, Denmark) or Buchi B-290 mini (Buchi Corpration, New Castle, DE 19720) spray dryer. After spray drying, the synthetic surfactant powders were filled into size 00 capsules. Testing of aerosol properties and solid-state properties of the DP surfactant formulations included particle sizing, X-ray diffraction, thermogravimetric analysis and differential scanning calorimetry. Respirable dry particles with a mass median aerodynamic diameter of between 1 and 4 microns were preferred. A set of lead optimized DP synthetic surfactant formulations were then provided to Los Angeles Biomedical Research Institute in a blinded manner for further in vitro and in vivo characterization.

Aerosol delivery
Because there is currently no commercially available system designed specifically for inhalation therapy during nCPAP, a proprietary DP delivery device was designed and produced by Acorda Therapeutics (Chelsea, MA 02150). This low flow aerosolization chamber (LFAC) consists of a cylindrical chamber with one or more holes at one end that can accommodate a perforated capsule containing powder and disperse it at low flow rates (4-10 l/min) and can also be incorporated into a ventilatory system like nCPAP in lieu of being utilized as a stand-alone device. Depending on the flow rate, DP surfactant is delivered as a fine mist over 1-3 minutes per capsule filled with about 30 mg of surfactant. Dosing was based on the lipid content of the DP surfactant and amounted 100 mg/kg bodyweight. In consideration of the potential applications of this work, in addition to its efficacy in aerosolizing and delivering DP synthetic surfactants, additional design criteria for the device included the following (1) simplicity of design and use, (2) minimum number of parts, and (3) low cost of goods/manufacture. Attenuated-Total-Reflectance Fourier-transform infrared (ATR-FTIR) spectrometry ATR-FTIR spectra were recorded at 37°C using a Bruker Vector 22 FTIR spectrometer (Pike Technologies, Fitchburg, WI 53719) with a deuterium triglyceride sulfate (DTGS) detector. The spectra were averaged over 256 scans at a gain of 4 and a resolution of 2 cm -15 . For FTIR spectra of SMB in DP synthetic lung surfactant, the powder was solvated with deuterated water (D 2 O) and transferred onto a germanium ATR crystal. The aqueous solvent was then removed by flowing nitrogen gas over the sample to produce a thick lipid-peptide (lipid:peptide ratios of 10:0.3, mole:mole) 5 . The multilayer film was then hydrated to ≥35% with deuterated water vapor in nitrogen for 1 h before acquiring the spectra 11 . The spectra for the SMB peptide in the lipid film were obtained by subtracting the spectrum of a peptide-free control sample from that of the peptide-bound sample. The relative amounts of α-helix, β-turn, β-sheet, or random (disordered) structures in lipid-peptide films were estimated using Fourier deconvolution (GRAMS AI 8, version 8.0, Thermo Fisher Scientific, Waltham, MA 02451). The respective areas of component peaks were calculated using curve-fitting software (Igor Pro, version 1.6, Wavemetrics, Lake Oswego, OR 97035) 12 . FTIR frequency limits were: α-helix (1662-1650 cm -1 ), β-sheet (1637-1613 cm -1 ), turn/bend (1682-1662 cm -1 ), and disordered or random (1650-1637 cm -1 ) 13 .

Captive bubble surfactometry
Surface activity of the DP surfactant formulations was measured with a captive bubble surfactometer built by Schürch and coworkers 14,15 . DP surfactant formulations were dissolved in distilled water at a concentration of 35 mg/mL and quasi-static and dynamic compression cycling of the air bubble was performed at an average surfactant lipid concentration of 70 µg/mL in the bubble chamber (~1.0 mL volume) 5,16 . All measurements were performed in quadruplicate.

In vivo experiments: surfactant-deficient rabbits
Animal experiments were performed according to protocols that were reviewed and approved by the Institutional Animal Care and Use Committee of the Los Angeles Biomedical Research Institute at Harbor-UCLA Medical Center (LA BioMed protocol # 020645) 16 . All procedures and anesthesia were in accordance with the American Veterinary Medical Association (AMVA) guidelines. Treatment allocation was done by dynamic randomization and minimum sample size per treatment group was based on effect size 16 .
Lung lavaged rabbits are surfactant-deficient for at least 6-8 hours, but need respiratory support and surfactant treatment for survival. A total of 30 young adult, male New Zealand white rabbits with a weight of 1.0-1.4 kg were purchased from IFPS Inc. (Norco, CA 92860). Veterinary care; anesthesia, sedation, and muscle paralysis; and provision of maintenance fluid have been reported previously 5,16 . After inducing anesthesia, a venous line was placed via a marginal ear vein, an orotracheal tube was inserted, mechanical ventilation was started, and a carotid arterial line was inserted via a small incision in the skin of the anterior neck. Heart rate, arterial blood pressures and rectal temperature were monitored continuously (Labchart ® Pro, ADInstruments Inc., Colorado Springs, CO 80906). Respiratory support was provided with volume-controlled ventilation (Harvard Apparatus, Holliston, MA 01746) using a tidal volume 7.5 mL/kg, a positive end-expiratory pressure of 3 cm H 2 O, an inspiratory/expiratory ratio of 1:2, 100% oxygen, and a respiratory rate sufficient to maintain the partial pressure of carbon dioxide (PaCO 2 ) at ∼40 mmHg 5,16 . Airway flow and pressures and tidal volume were monitored with a pneumotachograph (Hans Rudolph Inc., Shawnee, KS 66227) connected to the orotracheal tube during mechanical ventilation. After a partial pressure of oxygen in arterial blood (PaO 2 ) >500 mmHg was reached at a peak inspiratory pressure <15 cm H 2 O in 100% oxygen, surfactant-deficiency was obtained with repeated lung lavages with 30 mL/kg of warmed normal saline. When the PaO 2 was stable at <100 mmHg (average four lavages), one of two surfactant treatment approaches was chosen in a random fashion: (1) insufflation of surfactant via the orotracheal tube using the LFAC device with continuation of mechanical ventilation (n=13), or (2) extubation after establishment of adequate spontaneous breathing, followed by placement on nasal continuous positive airway pressure (nCPAP) using a F&P bubble CPAP system (Fisher & Paykel Healthcare Inc., Irvine, CA 92618) with short home-made nasal prongs and insufflation of surfactant with the LFAC device at its lowest possible flow setting via the nCPAP system 9 (n=17). DP surfactant was administered at a dose of 100 mg lipids/kg body weight. Oxygenation was followed by measuring arterial pH and blood gases and, if the rabbit was intubated and mechanically ventilated, dynamic lung compliance at 15 min intervals over a 2 h period. At 2 h after surfactant administration, rabbits were euthanized with 100 mg/kg of pentobarbital intravenously. Animals supported on nCPAP were re-intubated via a tracheotomy and mechanically ventilated to measure dynamic lung compliance. Oxygenation and dynamic lung compliance were used as end-points.
In vivo experiments: preterm lambs Preterm lambs at 135 days of gestation (term is 150 d) are surfactant-deficient, but able to breathe spontaneously while supported with CPAP delivered with nasal prongs (nCPAP) 17,18 . A total of 12 date-mated pregnant ewes with twin pregnancies were treated with 12 mg of betamethasone by intramuscular injection 48 and 24 h prior to delivery by cesarean section. Anesthesia of the ewes was induced with 10 mg/kg of ketamine IM and maintained with oxygen and isoflurane after placement of a cuffed endotracheal tube and starting mechanical ventilation. The uterus was then exposed through a ventral midline incision, the fetus(es) were located and their legs and abdomen were exposed through a hysterotomy, followed by insertion of an arterial and a venous umbilical catheter, exteriorization, insertion of an orotracheal tube (ID 4.0 mm) for mechanical ventilation, and cutting of the umbilical cord. Prior to the cesarean section, heparinized blood was collected from the jugular vein of the ewe to be administered to the lambs in case of hypotension or blood loss. After the cesarean section, ewes were euthanized by infusing 100 mg/kg of pentobarbital IV.
The lambs were dried, weighed, placed in a warmer on a heating pad and under heating lamps, and started on mechanical ventilation after a sustained inflation at 40 cm H 2 O pressure for 20 sec. Initial ventilator settings were: peak inspiratory pressure (PIP) of 30 cm H 2 O, positive end-expiratory pressure (PEEP) of 8 cm H 2 O, inspiratory time of 0.4 sec, and respiratory rate of 60/min. PIP and respiratory rate were adjusted to achieve tidal volumes of 6 (5-10) mL/kg, a pH >7.20 and PaCO 2 <60 mmHg. Both nostrils of the lambs were sprayed with 10% lidocaine spray and custom-made binasal prongs (shortened Portex tracheal tubes, ID 3.0 mm) were inserted 6-7 cm for CPAP delivery 17 . Spontaneous breathing was stimulated with 20 mg/kg of caffeine and 5 mg/kg of doxapram IV, followed by continuous infusion of 2.5 mg/kg/h of doxapram. Heart rate, arterial blood pressure, oxygen saturation (pulse oximetry) and rectal temperature were recorded continuously. Anesthesia was provided by administering 30 mg/kg/h of propofol IV as needed. Maintenance fluid consisted of Lactated Ringer's solution at a rate of 10 mL/kg/h. Airway flow and pressures and tidal volume were monitored during mechanical ventilation with a pneumotachograph (Hans Rudolph Inc., Shawnee, KS 66227).
Respiratory failure was monitored by observing oxygen saturation and confirmed by PaO 2 levels <150 mmHg in 100% oxygen, at which time lambs were treated with synthetic DP lung surfactant. Surfactant was given either (1) with the LFAC device attached to the orotracheal tube, followed by extubation to nasal intermittent positive pressure ventilation (NIPPV) and then to nCPAP (heated and humidified bubble CPAP with PEEP 8-9 cm H 2 O and 100% oxygen), or (2) during nCPAP (after extubation and transition via NIPPV) with the LFAC device attached to one of the nasal prongs of the CPAP system and a short period of mononasal CPAP support, or (3) same as (2) with a second dose of surfactant 1 h after the initial dose. So, the location of aerosol delivery of synthetic DP lung surfactant (IT or via nCPAP) and the number of surfactant doses during nCPAP (1 or 2) were the only differences between the three treatment groups. Aerosol delivery of surfactant powder usually took about 15 min from start to finish. After surfactant treatment, arterial pH and blood gas measurements were done every 15-30 minutes. At 3 h after surfactant treatment, the lambs were sacrificed with 100 mg/kg of pentobarbital IV. After euthanasia, the thorax was opened by a midline incision, an endotracheal tube (ID 4.0 mm) was reinserted, and a pressure volume curve was recorded. End-points of the experiment included: gas exchange (arterial pH and blood gases) and pulmonary mechanics (dynamic compliance during mechanical ventilation, postmortem pressure-volume curve).

Statistical analysis
All data are expressed as mean ± SEM. Student's t-tests were used for comparisons of discrete data points and functional data were analyzed with one-way analysis of variance (ANOVA) with Scheffe's post-hoc test. Differences with a P value <0.05 were considered statistically significant.

DP synthetic lung surfactant formulations
Aerosol properties and solid-state properties of the DP surfactant formulations are summarized in Table 1. Geometric particle size (gPSD) and fine particle fraction (FPF) were well within the margins for respiratory particles. The FPF of DP-2L remained relatively constant across all stability time points whereas a drop in FPF was observed for DP-PA and DP-3L. The solid-state properties are similar among the various formulations. The DP formulations are semi-crystalline with a characteristic diffraction peak at 21° 2theta, which is attributed to the presence of phospholipids in a bilayer structure. Residual solvent (ethanol and water) content is less than 1.5 wt%. The DP formulations contain phase transitions at ~45-53°C, calculated as the intercept of a step transition, and 61-67°C, calculated as a peak extremum of an endotherm. The stability data of these formulations indicate that the powders would be viable for long-term storage.

FTIR results
The secondary structures for SMB in DP-2L, DP-3L, and DP-PA were studied with conventional 12 C-FTIR spectroscopy of samples solvated in D 2 O. Representative FTIR spectra of the amide I band for SMB in these environments were all similar ( Figure 2), each showing a primary component centered at ~1654-1656 cm -1 , with a small low-field shoulder at ~1622-1626 cm -1 . Because prior FTIR studies of proteins and peptides 13,19 have assigned bands in the range of ~1650-1659 cm -1 as α-helical, while those at ~1613-1637 cm -1 denote β-sheet, SMB probably assumes α-helical and β-sheet structures and possibly other conformations in these environments. Self-deconvolutions of the Figure 2 spectra confirmed that SMB is polymorphic, primarily adopting α-helix but with significant contributions from β-sheet, loop-turn and disordered components (Table 2). Interestingly, the relative proportions of secondary conformations determined from FTIR spectra of SMB (i.e., α-helix > loop-turn ~ disordered ~ β-sheet) in both lipid mimetics and surfactant lipids of varying polarity are all comparable (Table 2) 5,8 , suggesting overall stability of the SMB structure that is remarkably conserved. It should also be noted that the proportions of these secondary conformations are entirely compatible with SMB principally assuming an α-helix hairpin 8 , and which has been shown to be surface active in in vivo experiments involving intratracheal instillation of liquid surfactants 5 .

Captive bubble surfactometry
All three DP surfactant formulations with 3 wt% SMB, i.e. DP-2L, DP-3L and DP-PA, exhibited a minimum surface tension below 1 mN/m during quasi-static cycling and were comparable with the positive control, i.e. Curosurf ® , indicating that there was no loss in in vitro surface activity during the surfactant production process (Figure 3). Minimum surface tension values of DP surfactant without SMB (DP-C), our negative control, far exceeded those of the surfactant formulations with SMB and amounted to ~18 mN/m (p<0.001), comparable to equivalent non-spray dried surfactant lipid formulations.
First, second, fifth and tenth surface tension-area relations produced by dynamic cycles (20 cycles/min) are shown in Figure 4. gPSD, average geometric particle size (gPSD) in microns (d50); FPF, fine particle fraction of the total dose < 5.6 and/or 3.4 microns; XRPD, X-ray powder diffraction; SC-D, semicrystalline material with diffraction peaks attributed to the presence of phospholipids in a bilayer structure; DSC,differential scanning calorimetry; Low T1, characteristic temperature(s) of the first thermal event(s) observed during a DSC scan at 20 °C/min; Low T2, characteristic temperature(s) of the second set of thermal event(s) observed during a DSC scan at 20°C/min; TGA-120 (%), volatiles loss by 120°C. . Minor contributions are also indicated by arrows on the DP-PA spectrum at the 1625 cm -1 band due to ß-sheet, at the 1672 cm -1 band due to turns and bends, and at the 1643 cm -1 band due to disordered or random conformations. The DP-3L spectrum is based on a representative sample of two productions. Peptide concentrations were 3.0 wt% for lipid-peptide ensembles. The areas under each absorbance curve are normalized using a SigmaPlot 14.0® macro. FTIR sample crystal using a stream of dry nitrogen gas. The dried film was then hydrated with deuterated water vapor for one hour prior to spectral measurement. b See Figure 2. ATR-FTIR spectra were estimated for proportions of the secondary structure for SMB in surfactant lipid films using self-deconvolution of the peptide amide I band (see Methods).
The dry powders DP-2L, DP-3L, and DP-PA ( Figure 4) and the clinical surfactant Curosurf ® (Figure 4) all needed an area compression of approximately 30% to reach a minimum surface tension of 1 mN/m in the first cycle. Dynamic cycling showed a wide difference between the compression and expansion curves for both DP-2L and Curosurf ® , with a relatively smaller hysteresis area for DP-PA and an even smaller hysteresis area for DP-3L. The lipid control (DP-C) (Figure 4) failed to reach low surface tension values as seen during quasi-static cycling in Figure 3.
In vivo experiments: surfactant-deficient rabbits A total of 30 surfactant-deficient rabbits (mean weight 1.3 ± 0.02 kg) received 100 mg lipids/kg bodyweight of one of the three DP synthetic lung surfactant containing SMB via intratracheal (IT) or nCPAP aerosol delivery. Of the 30, 13 rabbits underwent IT aerosol delivery while supported with mechanical ventilation: 6 received DP-2L, 5 DP-PA and 2 DP-3L. IT delivery was followed by continuation of mechanical ventilation and we observed a quick and large improvement in oxygenation and lung compliance after aerosolization of the three DP synthetic surfactants containing SMB (DP-2L, DP-A and DP-3L) ( Figure 5). Oxygenation in the IT groups increased from 10-16% to 75-85% and lung compliance from 46-52% to 69-71% of baseline values measured prior to lung lavage. The surfactant response was comparable to our recent experience with liquid synthetic surfactant preparations containing 3 wt% of SMB or its sulfur-free derivative B-YL or the clinical surfactant Curosurf ®4,5,16 . The remaining 17 rabbits received DP synthetic surfactant by aerosol delivery via the nCPAP system: 8 received DP-2L, 6 DP-PA and 3 DP-3L. Nasal CPAP led to a consistent, but smaller increase in oxygenation and dynamic compliance ( Figure 5). Oxygenation in the nCPAP groups increased from 12-13% to 41-43% and lung compliance from 44-50% to 63-67% of baseline values measured prior to lung lavage. Improvement in lung function in the rabbits receiving mechanical ventilation was independent of the active DP synthetic surfactant administered and the differences among the rabbits treated with surfactant while on nCPAP support were not statistically different either. The differences in oxygenation and lung compliance between IT and nCPAP aerosol delivery of DP-2L, DP-PA and DP-3L were statistically significant for each surfactant (p<0.05).
In vivo experiments: preterm lambs A total of 12 pregnant ewes delivered 22 preterm lambs; one ewe carried a single fetus and one ewe had a stillbirth. Of the lambs, three died quickly after birth despite mechanical ventilation and resuscitation efforts, so 19 preterm lambs (weight 3.1 ± 0.2 kg, gestational age 135.5 ± 0.3 days, 11 males and 8 females) received DP synthetic surfactant. In total, 6 lambs were dosed IT and then weaned to nCPAP, 7 lambs received one dose of surfactant while supported with nCPAP and 6 lambs received two doses of surfactant 1 h apart while supported with nCPAP. Based on the available in vitro data and experience with surfactant-deficient rabbits, all preterm lambs received DP-2L surfactant at 100 mg lipids/kg bodyweight/dose. All lambs were first weaned to NIPPV before initiating nCPAP support, and, if necessary, were returned for short periods to NIPPV in case of insufficient breathing effort or rising PaCO 2 values. The response to surfactant treatment was monitored with pulse oximetry and arterial pH and blood gases every 15 min during the 1 st and 2 nd hour and every 30 min during the 3 rd hour after initial surfactant treatment. Oxygenation quickly improved, reaching PaO 2 values of 385 ± 20 mmHg in 100% oxygen at 3 h after IT aerosol delivery of DP-2L versus 204 ± 19 mm Hg after 1 dose and 310 ± 16 mm Hg after 2 doses of DP-2L administered via the nCPAP system (p<0.02) (Figure 6). PaCO 2 values  Oxygenation in the IT groups increased from 10-16% to 75-85% and lung compliance from 46-52% to 69-71% of baseline values measured prior to lung lavage. In the nCPAP groups oxygenation increased from 12-13% to 41-43% and lung compliance from 44-50% to 63-67% of baseline values measure prior to lung lavage. Differences in oxygenation and lung compliance between DP-2L, DP-PA and DP-3L surfactant given IT were not statistically significant and the same held true for DP-2L, DP-PA and DP-3L surfactant given via nCPAP. Differences in oxygenation and lung compliance between IT and nCPAP aerosol delivery were statistically significant for each of the three DP synthetic surfactants (p<0.05).
decreased more slowly and were 58.0 ± 2.2, 56.8 ± 2.8 and 54.8 ± 2.1 mm Hg (NS), respectively, at 3 h after surfactant treatment ( Figure 6) and this reduction was accompanied by an equivocal increase of pH values (data not shown). Data from postmortem pressure-volume curves are shown in Figure 7. Only the differences in lung volumes at pressures of 40 and 20 cm H 2 O between IT or two doses of nCPAP surfactant treatment and one dose of nCPAP surfactant were statistically significant (p=0.003 and 0.035, respectively).

Discussion
Here, we designed and formulated a group of three surfaceactive dry powder (DP) synthetic lung surfactant preparations with phospholipids DPPC and POPG (7:3 wt:wt) and 3 wt% of the SP-B peptide mimic Super Mini-B (SMB) as surfaceactive ingredients. These preparations were tested in vitro with ATR-FTIR spectrometry and captive bubble surfactometry and in vivo in two animal models of surfactant-deficiency, i.e., lavaged young adult rabbits and preterm lambs supported with mechanical ventilation or nCPAP. IT and nCPAP delivery of aerosolized DP synthetic lung surfactant was realized with a proprietary DP delivery device. In vitro data showed that the three SMB-containing DP preparations DP-2L, DP-3L, and DP-PA were surface-active. All three surface-active DP synthetic surfactant preparations improved lung function in ventilated and spontaneous breathing surfactant-deficient rabbits receiving nCPAP. The oxygenation response after nCPAP delivery and support amounted to about 50% of the oxygenation response after IT delivery and mechanical ventilation, whereas the improvement in lung compliance was close to 40% in both approaches. Based on the in vitro data and rabbit experiments, DP-2L synthetic surfactant was chosen for treatment of the preterm lambs. In the preterm lambs we found a similar difference in the oxygenation response to IT and nCPAP aerosol delivery during non-invasive respiratory support with nCPAP. This difference could be reduced by giving a second dose of DP surfactant during nCPAP support.
Although we used positive and negative controls in the in vitro studies, we did not include controls in the current in vivo studies as previous work by our group has repeatedly reported the response to lipids alone as negative control and the clinical surfactants Curosurf ® and Infasurf ® as positive controls after IT instillation in lavaged, ventilated young adult rabbits 4,5,9,16 . After completing the first 6 sheep experiments, we changed the treatment protocol in the preterm lamb studies by adding a group that received two doses of surfactant while supported with nCPAP in order to test whether repeated dosing was a viable and effective approach. The aerosol delivery device we used relies on the use of capsules with a limited amount of DP surfactant, so dosing would take about 10-15 min in the rabbit experiments and 20-30 min in the preterm lamb experiments. Adaptation of this inhaler to facilitate continuous surfactant delivery would enhance the ease of use 16 .
It is noteworthy that the three SMB-containing DP preparations, administered to animals in the dry state as aerosols, showed good surfactant activities ( Figure 5-Figure 7), which were comparable to earlier in vivo findings with intratracheal instillation of liquid surfactant lipids with SMB 5 . SMB reproduces the topology of the N-and C-terminal domains of SP-B as it contained the α-helical residues 1-25 and 63-78 joined by a custom β-turn -PKGG-and cross-linked with two disulfide bridges (Table 2) to form an α-helix hairpin. Both homology (I-TASSER) and MD-simulated models of SMB previously demonstrated a globular protein, exhibiting a surface of polar residues and a core of hydrophobic residues that was severely constrained by the disulfide bonds 8 . Earlier 100-nsec MD simulations also showed that SMB inserted into DPPC/POPG lipid bilayers as a peripheral protein, where the peptide partitions into POPGenriched domains through selective interactions of its cationic N-and C-helices and the anionic lipid head group 6 . Notably, a preliminary 499-nsec MD simulation of SMB in surfactant lipid bilayers [i.e., DPPC/POPC/POPG] showed the absence of water within disulfide-linked SMB, and only weakly H-bonded water to several polar residues at the aqueous-bilayer interface 8 . An additional drying step that temporarily removed only surfacebound water would probably minimally perturb the structure and function of SMB in surfactant lipids, which is supported by FTIR spectroscopy showing α-helix hairpins for SMB in either dry powders ( Figure 1; Table 2)  The ultimate goal of this study was to design and develop a lowcost synthetic lung surfactant and a device for aerosol delivery for use in a low technical setting. Using a DP synthetic lung surfactant with an advanced SP-B peptide mimic and a relatively simple low flow inhaler we were able to obtain a pulmonary response in our animal studies that supports translation of this approach to the sketched clinical setting.

Conclusion
DP synthetic lung surfactant with DPPC and POPG (7:3 wt: wt) and 3 wt% of the SP-B mimic peptide SMB is stable, highly surface active and effective in improving lung function in surfactant-deficient rabbits and preterm lambs.

Gates Open Research
investigated at refrigeration temperature for the equivalent of 12 months?
3) The accelerated stability testing conducted at 40°C for the dry powders could use further explanation in the Methods section. It seems to only be briefly mentioned in Table 1 but powder stability is cited as one of the advantages of this surfactant delivery strategy.

4)
Have the authors investigated whether the DP formulation after accelerated stability testing was still active (e.g., via captive bubble surfactometry)? Might the SP-B peptide have undergone some degree of thermal degradation even as a dry powder? 5) In Figure 3, are the error bars for the Curosurf data and the minimum surface tension measurements absent or so small as to be contained within the data point markers?
6) Please explain the rationale behind the uneven DP group sizes chosen for the rabbit study, in particular the much smaller n for DP-3L. For rabbits that underwent intratracheal aerosol delivery, 6 received DP-2L, 5 DP-PA, and 2 DP-3L. For rabbits that underwent delivery via the nCPAP system, 8 received DP-2L, 6 DP-PA, and 3 DP-3L.

If applicable, is the statistical analysis and its interpretation appropriate? Partly
Are all the source data underlying the results available to ensure full reproducibility?
No source data required

Are the conclusions drawn adequately supported by the results? Partly
No competing interests were disclosed.

Competing Interests:
Reviewer Expertise: Lung surfactant protein mimics (based on peptoids), Antimicrobial peptides, Molecular biophysics, Surface & colloid science We have read this submission. We believe that we have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. : Adding a SP-C analogue would indeed be a logical next step. Response Table 1 reads "5°CC 12 mth". Should this read "5°C" instead as in the powder stability was investigated at refrigeration temperature for the equivalent of 12 months?

A row in
: Thank you for noticing, corrected. Response 3. The accelerated stability testing conducted at 40°C for the dry powders could use further explanation in the Methods section. It seems to only be briefly mentioned in Table 1 but powder stability is cited as one of the advantages of this surfactant delivery strategy.
: Further stability testing is being planned. Response 4. Have the authors investigated whether the DP formulation after accelerated stability testing was still active (e.g., via captive bubble surfactometry)? Might the SP-B peptide have undergone some degree of thermal degradation even as a dry powder?
: We have indeed only tested "fresh" DP formulations with captive bubble surfactometry. Response 5. In Figure 3, are the error bars for the Curosurf data and the minimum surface tension measurements absent or so small as to be contained within the data point markers?
: The standard errors were small and contained within the data point markers. This info Response was added to the legend of figure 3.
6. Please explain the rationale behind the uneven DP group sizes chosen for the rabbit study, in particular the much smaller n for DP-3L. For rabbits that underwent intratracheal aerosol delivery, 6 received DP-2L, 5 DP-PA, and 2 DP-3L. For rabbits that underwent delivery via the nCPAP system, 8 received DP-2L, 6 DP-PA, and 3 DP-3L.
: We started our rabbit experiments with 2 DP surfactants (DP-2L and DP-PA) on hand Response and obtained DP-3L at a later stage. The DP surfactant preparations appear to be stable in anhydrous state and yet posses high surface activity. The investigators have also designed a low flow aerosolization chamber (LFAC) to deliver the dry powder synthetic SMB via nCPAP. activity. The investigators have also designed a low flow aerosolization chamber (LFAC) to deliver the dry powder synthetic SMB via nCPAP.

No competing interests
Extensive testing of 3 different synthetic surfactant formulations, DP-2L, DP-3L and DP-PA in vitro comparing with synthetic dry powder without SMB as a negative control and clinical surfactant, Curosurf as positive control and authors report similar surface tension reducing properties similar to Curosurf. The authors went on to perform testing, initially in surfactant deficient young adult rabbits comparing in vivo arterial oxygenation and dynamic compliance for all three dry powder formulations delivered via intra-tracheal route and during nCPAP. They report significant improvement in oxygenation and lung compliance when the dry powder is delivered intra-tracheal route as compared to nCPAP, while reporting no significant difference between the 3 formulations given by either route. They then conducted experiments in preterm lambs, delivered at 135 days, comparing changes in PaO and PaCO after DP-2L formulation was administered via intra-tracheal route or during nCPAP (1 or 2 doses) and report significant improvement in oxygenation. Administration of a second dose further improved oxygenation.
Overall the manuscript reads well with the methods and results sections presented clearly and succinctly. The results of the study are promising for the development of novel synthetic lung surfactant with a device for aerosol delivery. Even though the intra-tracheal route was shown to have the greatest impact on oxygenation and compliance, 2 doses of SMB administered via nCPAP was reported to improve oxygenation and compliance, further studies are needed to evaluate the potential clinical benefits of SMB and LFAC.

Specific comments:
Introduction: 'The realization that chronic lung injury (a.k.a. bronchopulmonary dysplasia, BPD) is the result of chronic inflammation due to tracheal intubation and mechanical ventilation has led to a preference for non-invasive respiratory support with nasal continuous positive airway pressure (nCPAP) in preterm infants." Suggest adding: The etiology of BPD is multifactorial and although the intubation and mechanical ventilation is associated with higher risk of BPD, extremely preterm infants who have never been intubated or mechanically ventilated can still have BPD (new BPD) .
"This approach would especially be useful in preterm infants born in low-resource settings." Authors correctly discuss the benefits of noninvasive ventilation and surfactant therapy. Minimizing the duration of invasive ventilation, especially during the transitional period following preterm delivery is key in terms of reducing inflammation and lung injury. Aerosol delivery of surfactant is truly a non-invasive surfactant treatment (NIST) technique. I think, if surfactant in sufficient amounts could be delivered into the lung via aerosol, this combination of noninvasive ventilation and NIST should improve outcomes for all preterm infants and not just for infants born in resource limited settings. I would consider revising this sentence to include all preterm neonates with RDS.

Methods:
Aerosol delivery: we are not sure as to why the authors called 10 l/min as low flow Aerosolization chamber. Please explain. May just call it as Aerosolization chamber.
In vivo experiments: surfactant-deficient rabbits Please describe the details for providing nCPAP in these rabbits.

In vivo experiments: preterm lambs
In this section, authors describe surfactant delivery during nCPAP. However, the nasal prongs were inserted 6-7 cm for CPAP delivery. This is more of a naso-pharyngeal CPAP 9NP-CPAP) rather than 2 2 1 inserted 6-7 cm for CPAP delivery. This is more of a naso-pharyngeal CPAP 9NP-CPAP) rather than nCPAP. In preterm neonates or term neonates, nasal prongs are typically inserted into the nostrils only 0.5 to 1 cm. Please clarify. We would have anticipated lower amount of surfactant delivery in smaller animals, like, rabbits (~weight 1 kg) compared to lambs (~3 kg). Authors may want to discuss about the potential for redosing to improve pulmonary function in smaller animals.
Although the investigators mention in their discussion, not including clinical surfactant as a control, following results of their previous study in young adult rabbits, comparison of DP-2L administered via nCPAP with Curosurf administered via intra-tracheal route would have provided important information for future clinical application of SMB delivered by aerosol.

If applicable, is the statistical analysis and its interpretation appropriate? Yes
Are all the source data underlying the results available to ensure full reproducibility? Yes We have read this submission. We believe that we have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
Author Response 07 Mar 2019 , LA BioMed at Harbor-UCLA Medical Center, USA Frans Walther 1. Introduction: 'The realization that chronic lung injury (a.k.a. bronchopulmonary dysplasia, BPD) is the result of chronic inflammation due to tracheal intubation and mechanical ventilation has led to a preference for non-invasive respiratory support with nasal continuous positive airway pressure (nCPAP) in preterm infants." Suggest adding: The etiology of BPD is multifactorial and although the intubation and mechanical ventilation is associated with higher risk of BPD, extremely preterm infants who have never been intubated or mechanically ventilated can still have BPD (new BPD).
: We agree with the reviewer and rephrased this sentence. Response 2. "This approach would especially be useful in preterm infants born in low-resource settings." Authors correctly discuss the benefits of noninvasive ventilation and surfactant therapy. Minimizing the duration of invasive ventilation, especially during the transitional period following preterm delivery is key in terms of reducing inflammation and lung injury. Aerosol delivery of surfactant is truly a non-invasive surfactant treatment (NIST) technique. I think, if surfactant in sufficient amounts could be delivered into the lung via aerosol, this combination of noninvasive ventilation and NIST should improve outcomes for all preterm infants and not just for infants born in resource limited settings. I would consider revising this sentence to include all preterm neonates with RDS.
: Our starting point for this research was to develop a synthetic surfactant treatment for Response use in low technical settings. We agree with the reviewer and decided to remove this sentence.
3. Aerosol delivery: we are not sure as to why the authors called 10 l/min as low flow Aerosolization chamber. Please explain. May just call it as Aerosolization chamber.
: We wrote "10 L/min or less" because we usually started with a higher flow to get the Response surfactant capsule spinning and then reduced the flow as far as possible. We changed the text therefore to "(4-10 l/min)". 4. In vivo experiments: surfactant-deficient rabbits. Please describe the details for providing nCPAP in these rabbits.
: We added the following text to this section: "using a F&P bubble CPAP system (Fisher Response & Paykel Healthcare Inc., Irvine, CA 92618) with short home-made nasal prongs and insufflation of surfactant with the LFAC device at its lowest possible flow setting via the nCPAP system ." 5. In vivo experiments: preterm lambs. In this section, authors describe surfactant delivery during nCPAP. However, the nasal prongs were inserted 6-7 cm for CPAP delivery. This is more of a naso-pharyngeal CPAP (NP-CPAP) rather than nCPAP. In preterm neonates or term neonates, nasal prongs are typically inserted into the nostrils only 0.5 to 1 cm. Please clarify.
: We inserted the home-made prongs deeper into the nose of the preterm lambs to get a Response 9 : We inserted the home-made prongs deeper into the nose of the preterm lambs to get a Response better fixation and optimize aerosol delivery based on the work by Dargaville et al. (reference 17) and added this reference.
6. "All three surface-active DP synthetic surfactant preparations improved lung function in ventilated and spontaneous breathing surfactant-deficient rabbits." Change to: "All three surface-active DP synthetic surfactant preparations improved lung function in ventilated and spontaneous breathing surfactant-deficient rabbits receiving nCPAP." : in vivo in vitro phospholipids with the synthetic surfactant protein Super Mini-B (SMB). The manuscript builds upon the groups previous publications of liquid forms of the SMB. By creating a dry version that is clinically stable at room temperature and can be easily aerosolized, they can provide the beneficial effects of surfactant in limited resource environments.
The authors demonstrate that the three surfactant/phospholipid mixtures were stable at high temperatures for multiple months, with the DP-2L combination the most stable. They then demonstrated that the DP-2L for multiple months, with the DP-2L combination the most stable. They then demonstrated that the DP-2L mixture has the closest dynamic cycling curves to Curosurf.
The authors used standard protocols for saline lavaged rabbits and preterm lambs that were appropriate for the studies they conducted. The rabbits demonstrated improvement in oxygenation and compliance with all three mixtures, with less improvement in rabbits on CPAP (likely due to less of the surfactant mixture getting into the lungs). Based on the stability data and the responses in the rabbits, the DP-2L mixture was chosen to be tested preterm sheep. The addition of a second dose of aerosolized surfactant at 1 hour provided additional benefit to the lambs.
The lack of the Curosurf or phospholipid only groups in the animal studies makes placing the response to the aerosolized mixtures more difficult. The authors addressed this concern in the discussion. Evaluation of their previous publication demonstrates no response to phospholipid alone over time of ventilation, and a similar response to the IT instillation of the SMB compound and Curosurf. Based on this data, the conclusions they draw within the manuscript are correct and the aerosolized dry surfactant is nearly as effective a commercially available surfactant given via an endotracheal tube.

Minor suggestions:
Abstract: The methods section of the abstract about the animal studies is a little confusing. Reading the methods in the manuscript, it is clearer that some of the animals were intubated and the dry powder aerosolized to assure it made it to the lungs of the animals. The word "active" in the conclusion sentence of abstract could be removed. Introduction: Since BPD can occur in infants without mechanical ventilation and is likely a combination of multiple factors, the phrase "is the result of" in the introductory paragraph could be altered. Table 1: There are two C after 5 degrees. Figure 2: The figure is confusing as the diagrams for the 3 spectral plots look like they have difference degrees of absorption with A absorbing more than B and more than C. The graphs could be superimposed on one another with a different line type for each mixture. Then a legend could be put on the graph that demonstrates the name of the mixture. The information could also be placed in Table 2 alongside the secondary data structure. Figure 3: Curosurf data does not have any error bars, whereas other types of treatment have error bars.
Were the values identical on repeat measures? Figure 4: For ease of reader, the authors could consider changing DP-C to Control or Negative Control in the figure. Figure 5/6: Statistics between the group would be useful on the figures to assist readers. They are available in the figure legend and the manuscript text. Figure 5: The straight lines for the CPAP animals could be removed and the markers left at the 120 minute position since it is unlikely the compliance improved in a linear fashion between the two points of assessment of compliance in these spontaneously breathing animals. 6. Figure 4: For ease of reader, the authors could consider changing DP-C to Control or Negative Control in the figure.
: We considered it, but this might lead to confusion when readers consult reference 16 Response for the data on intratracheal administered Curosurf and lipid controls. 7. Figure 5/6: Statistics between the group would be useful on the figures to assist readers. They are available in the figure legend and the manuscript text.
: To avoid overcrowding, we described the data and report statistics in the legends of Response figures 5 and 6.
8. Figure 5: The straight lines for the CPAP animals could be removed and the markers left at the 120 minute position since it is unlikely the compliance improved in a linear fashion between the two points of assessment of compliance in these spontaneously breathing animals.
: Figure 5 was changed accordingly. Response

No competing interests
Competing Interests: