Secretoglobin 3A2 peptides have therapeutic potential for allergic airway inflammation

Abstract

Three isoforms of secretoglobin (SCGB) 3A2, namely type A, B, and C, are endogenously produced through alternative splicing. SCGB3A2 type A, the correctly spliced major type, begins to be expressed from embryonic day 11.5 in mice and shows various physiological activities such as promoting lung maturation and bronchial branching, anti-inflammatory effects, and ameliorating induced pulmonary fibrosis. To investigate the potential of SCGB3A2 peptides as a therapeutic to treat respiratory diseases, in this study, serially overlapping nine peptides were synthesized to cover the entire type C isoform, and five and one peptides covering the C-terminal region of type A and B, respectively. To evaluate their biological activities, each peptide was subjected to cell proliferation and apoptosis analyses in vitro using mouse lung fibroblast-derived MLg cells, bronchial branching rate using ex vivo mouse fetal lung organ cultures, and in vivo allergic airway inflammation mouse model. Among type A and C peptides, those corresponding to the C-terminal region of the SCGB3A2 sequence exhibited its unique biological activities of promoting cell proliferation and bronchial branching, and/or inhibiting apoptosis. The type B peptide did not show any proliferative effect while inhibited apoptosis. In a mouse model of allergic airway inflammation, lung inflammation was improved by the administration of most of the C-terminal region-derived type A and type C peptides. The results suggest that the bioactivity resides towards the C-terminal region of SCGB3A2 sequence, and the peptides covering this region could be used as a therapeutic in treating lung inflammation.

Graphical Abstract

Introduction

Secretoglobin family 3A member 2 (SCGB3A2), belonging to the SCGB gene superfamily, is a small protein of approximately 10 kDa secreted from airway epithelial cells. In mice, SCGB3A2 begins to be expressed on embryonic day (E) 11.5 [1,2], promotes lung development and maturation [3], has anti-inflammatory effects [4,5], and inhibit pulmonary fibrosis [6,7], as well as being associated with chronic obstructive pulmonary disease (COPD) [8–10]. Because of the location of the expression, and the anti-inflammatory and immunomodulatory activities like those found in surfactant proteins A and D [11], SCGB3A2 could be considered as a surfactant protein. The activities that SCGB3A2 possesses suggest that SCGB3A2 may potentially be utilized as a drug to treat various lung diseases [12].

Recently, syndecan-1 (SDC1) was identified as a receptor for SCGB3A2 [13,14]. SCGB3A2 has a high affinity towards the endotoxin lipopolysaccharide (LPS), forming a SCGB3A2-LPS complex that is delivered into cells through its binding to SDC1, followed by endocytosis. The internalized LPS activates caspase-4 (human)/-11 (mouse), which in turn activates the non-canonical inflammasome pathway, leading to pyroptosis, the inflammatory programmed cell death [13,14]. The SCGB3A2/LPS-activated caspase-4/-11-dependent cell death appears as SCGB3A2 exhibiting a novel anticancer activity [13,14]. On the other hand, the high affinity of SCGB3A2 towards LPS naturally causes a large amount of LPS contamination when recombinant SCGB3A2 (rSCGB3A2) is biosynthesized in E. Coli system, and the step to remove LPS during purification of rSCGB3A2 dramatically reduces yield.

SCGB1A1, the most studied member of the SCGB gene superfamily [15] {also called uteroglobin, club cell secretory protein (CCSP), and club cell 10-kDa protein (CC10)}, is expressed mainly in airway club cells and regulates the immune system, exhibiting anti-inflammatory and anti-fibrotic effects [15–18]. The amino acid sequence M39-L48 of the mature secreted form of rabbit SCGB1A1 is commonly conserved in multiple species of SCGB1A1, and because the synthetic oligopeptide of this area shows anti-inflammatory effects in in vitro and in vivo studies, it was named antiflammin [19,20]. However, the antiflammin synthetic oligopeptide has not yet been developed into a peptide drug for respiratory diseases or for clinical applications [20].

In recent years, the development of biopharmaceuticals based on genetic recombination technology and middle-molecular drugs (MW 500–30,000) such as peptide drugs are on the rise. Peptide drugs have fewer side effects and better efficacy than small molecule protein drugs that have been the mainstay of drug discovery [21]. In addition, compared to biopharmaceuticals, which are very expensive to produce, peptide drugs can be produced by chemical synthesis at a lower cost [22]. On the other hand, peptide drugs have low absorption efficiency in the gastrointestinal tract and a short half-life in vivo [23]. The lungs are considered to be a suitable target organ for peptide drugs because of their large surface areas due to alveoli structure, lack of digestive enzymes, and low metabolic enzyme activity [23]. Various absorption enhancers have also been investigated to improve the absorption efficiency of peptide drugs in the lung [24]. In this study, we focused on chemically synthesized SCGB3A2 as a source for highly purified SCGB3A2 without LPS contamination that can be used to develop SCGB3A2 as a therapeutic drug for the treatment of respiratory diseases.

Materials and Methods

Animals

BALB/cA and C57BL/6N (7 to 18 week-old) wild-type (WT) mice were obtained from Clea Japan, Inc. (Tokyo). Mice were housed in plastic cages in a controlled indoor environment of 22–24°C, 40–60% humidity, and were provided with clean food (CE-2, Clea Japan) and water. Room lighting was automatically turned on and off every 12 hours. All animal experiments were performed with the approval of the Yamagata University Animal Experimentation Committee. The study was carried out in accordance with the Guidelines for Proper Conduct of Animal Experiments (Science Council of Japan) and international guiding principles for biomedical research involving animals.

Generation and analysis of a mouse model of allergic airway inflammation

BALB/cA mice were administered 5 mg/kg of chicken ovalbumin (cOVA) intraperitoneally (i.p.) on days 0 and 14 and were intratracheally (i.t.) administered at 5 μg/kg daily on days 22–24 and 29–31 (see Supplemental Fig. S5A). On days 25–28 and 32–34, sensitized mice were i.t. administered 250 μg/kg of SCGB3A2 or SCGB3A2 peptides, or PBS as control. A non-sensitized control group of mice were given only PBS for both sensitization and challenge. SCGB3A2 was synthesized and purified using PUREfrex^® 2.0 (Gene Frontier Inc., Chiba, Japan), and endotoxin levels measured using the ToxinSensor^™ Chromogenic LAL Endotoxin Assay Kit (GenScript Corp., NJ, USA). Synthesis of SCGB3A2 and the endotoxin assays were performed according to the manufacturers’ protocols. Endotoxin levels in purified SCGB3A2 protein and most of synthetic SCGB3A2 peptides were <0.04 (lowest detection limit), except for some (0.29 EU/mL). To demonstrate the anti-inflammatory effect of a peptide in mice, the number of immune cells in bronchoalveolar lavage fluid (BALF) was counted and histological analysis of lung tissues was carried out. Briefly, mice were deeply anesthetized by i.p. administration of mixed anesthetic agents (medetomidine hydrochloride, midazolam, and butorphanol; for the details, please see [25]), and euthanized by drawing blood from the inferior vena cava. The trachea and bronchus were then exposed, and the right main bronchus was ligated with cotton thread. The left lung was gently lavaged three times with 0.5 mL of PBS through a tracheal cannula, and bronchoalveolar lavage fluid (BALF) was recovered. After BALF was collected, the threads of the right main bronchus were untied and the left main bronchus ligated. The left lung lobe was then resected, flash-frozen using liquid nitrogen, and stored at −80°C. Next, a 26G needle with a smooth tip was inserted into the right lung lobe, and 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer (PB, pH 7.4) was injected at a constant pressure and flow rate of 20-cm H₂O. The right lungs were then removed and immersed in 4% PFA in PB and fixed at 4°C for 16–24 hours. The blood samples were centrifuged at 4°C, 5,000 rpm for 5 min to prepare serum. Fifty μL of collected BALF was transferred to an eppendorf tube, to which an equal volume of Trypan Blue (Vector Laboratories, CA, USA) was added, and the total and live cell numbers were counted using a Thoma blood cell calculator (Sunlead Glass, Chiba, Japan). The remaining BALF was centrifuged at 4°C, 1,500 rpm for 15 min, and the cell pellet resuspended in 100 μL saline. Cell smears were prepared using Cytofuge 2 (Beckman Coulter, CA, USA), and cells were stained with Giemsa staining solution (Nacalai Tesque, Kyoto, Japan). A total of 200 cells per each glass slide were observed under a 3-axis microscope (Olympus, Tokyo, Japan) to calculate a fractional ratio of eosinophils, macrophages, neutrophils, and lymphocytes. For histological analysis, PFA fixed lungs were dehydrated in a series of dilution of ethanol, permeabilized with xylene (Nacalai Tesque), immersed in paraffin (Sakura Finetek Japan, Tokyo) at 60°C for replacement, and then embedded in paraffin. After embedding, serial sections of 4 μm thickness were prepared. Lung sections were subjected to Hematoxylin & Eosin (HE) and periodic acid-Schiff (PAS) staining, and were observed and photographed with an upright microscope (BX51, Olympus) attached to a digital camera (DP71, Olympus). The degree of lung inflammation and goblet cell hyperplasia was scored on a subjective scale of 0 to 4, as previously described [25]. Briefly, to score inflammatory cell infiltration in the luminal, alveolar, peribronchial, and perivascular areas, cell counts were carried out in blinded manner by two histology experts based on a 5-point rating system for the following characteristics: 0: normal, 1: low cell count, 2: inflammatory cells at a depth of 1 cell layer ring, 3: inflammatory cell ring at a depth of 2 to 4 cells, 4: inflammatory cell ring at a depth greater than 4 cells. A 5-point rating system was used to quantify goblet cells in bronchi and bronchioles, 0: PAS-positive cells <0.5%, 1: <25%, 2: 25–50%, 3: 50–75%, 4: >75%. Five fields of view were used for counting per slide and the average score was calculated from five animals; quantification of PAS-positive cup cells was expressed as the number of PAS-positive cells per 1 mm of basement membrane to correct for airway size.

ex vivo Organ Culture

Female C57BL/6N mice were mated, and mice whose vaginal plugs were identified in the morning following the day of mating were designated as day 0.5 of gestation and the fetuses were designated as E0.5. Pregnant female mice were euthanized at 11.5 days gestation by cervical dislocation, and E11.5 fetuses were removed from the uterus. Lungs were removed from the fetal mice and cultured using ex vivo organ culture system with RPMI-1640 medium (Roswell Park Memorial Institute medium, Nacalai Tesque) in a 37°C, 5% CO₂ incubator as previously described [3]. Fetal lungs ex vivo culture was added 10 μM of peptide (peptide group), 250 nM of SCGB3A2 (SCGB3A2 group), or medium only (control group) on Day 0, and the medium was replenished on Day 2. The number of bronchial branches was counted on Day 0 and 4 using bifurcation point where two bronchi diverge, starting the first point as 1. Bronchial branching ratio was calculated by dividing the number of bronchial branches on Day 4 by the number of bronchial branches on Day 0.

Synthesis of SCGB3A2 peptides

Serially overlapping SCGB3A2 type A (from aa 57–91), type B (aa 94–113), and type C peptides (aa 1–139) of approximately 20 amino acid residues each were synthesized by Fmoc solid-phase synthesis method (Table 1) [26]. The synthesized peptides were dissolved in 0.1% bovine serum albumin (BSA) (Fraction V, Wako 1^st Grad, Fujifilm Wako Pure Chemical Corporation, Tokyo, Japan) in PBS (−) (without CaCl₂ and MgCl₂, Nacalai Tesque).

Table 1.

SCGB3A2 peptides synthesized and used in this study

peptide	amino acid sequence
Type A	57           76
	SVEHLVTGLKKCVDELGPEA
	62           81
	VTGLKKCVDELGPEASEAVK
	67           86
	KCVDELGPEASEAVKKLLEA
	72           91
	LGPEASEAVKKLLEALSHLV
	54                      91
	LGISVEHLVTGLKKCVDELGPEASEAVKKLLEALSHV
Type B	94           113
	GRSLCYVNNLPSFEALSHLV
Type C	1           20
	MKLVSIFLLVTIGICGYSAT
	16           35
	GYSATALLINRLPVVDKLPV
	31           50
	DKLPVPLDDIIPSEDPLKML
	46           65
	PLKMLLKTLGISVEHLVTGL
	61           80
	LVTGLKKCVDELGPEASEAV
	76           95
	ASEAVKKLLVIIICSYFPGR
	91           110
	YFPGRSLCYVNNLPSFVSVL
	106           125
	FVSVLFLPMICAYPRDSKKQ
	121           139
	DSKKQTFAFIERVFEQSKL

Note: Peptides were synthesized comprehensively to cover all amino acid sequences of SCGB3A2 including the signal peptides. The amount of LPS in the SCGB3A2 and synthetic peptides used, with some exceptions, was below the detection limit of 0.04 EU/mL.

Quantitative Reverse Transcription (qRT)-PCR

Total RNAs were extracted from E11.5, E16.5, and E18.5 mouse fetal lungs. Specifically, fetal mouse lungs were immersed in 200 μL of TRIzol (Thermo Fisher Scientific, MA, USA) solution, crushed with a homogenizer pestle (AS ONE, Osaka, Japan), and then 800 μL of fresh TRIzol solution was added to make a suspension. For adult mouse lungs, a left lung was immersed in 2 mL of TRIzol solution and crushed with Physcotron (NS-52, Microtec Co., Ltd., Chiba, Japan) to make a suspension. Total RNAs were subjected to digestion of genomic DNAs and reverse transcription reaction using a TaKaRa PrimeScript reagent kit with gDNA Eraser (TaKaRa Bio, Shiga, Japan). Quantitative RT-PCR (qRT-PCR) was performed as previously reported [8]. All data of qRT-PCR were normalized to 18S rRNA. The primer sequences used in these experiments are listed in Supplementary Table S1.

Cell culture

Both mouse lung fibroblast cell line MLg and mouse monocyte macrophage-derived cell line RAW264 were obtained from American Type Culture Collection (ATCC, VA, USA). MLg cells were grown in DMEM (Dulbecco’s Modified Eagle’s Medium) (Merck KGaA, Darmstadt, Germany) and RAW264 was cultured in RPMI-1640 (Nacalai Tesque), both containing 10% fetal bovine serum (FBS) (Thermo Fisher Scientific) and 1% Penicillin-Streptomycin mixed solution (Nacalai Tesque). Cells were maintained in a 37°C, 5% CO₂ incubator (ASTEC, Fukuoka, Japan).

MTT cell proliferation assay

MLg cells (2.0 × 10³ cells/well) and RAW264 cells (5.0 × 10³ cells/well) were seeded on 96-well Cell Culture Plates (Greiner Bio-One, Frickenhausen, Germany). Cells were cultured for 3 hours, and were stimulated with 5 μM of each SCGB3A2 peptide or 59.5 nM of SCGB3A2. The control group did not have any additives. MLg cells were further cultured for 37 hours and RAW264 cells for 20 hours. After culturing, the medium was removed and 1 mg/mL of MTT solution (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide, C₁₈H₁₆BrN₅S; Dojindo Laboratories, Kumamoto, Japan) was added to each well and incubated at 37°C for 3 hours under 5% CO₂. The MTT solution was then removed and lysis Buffer [isopropanol (8.6 mL), 1N HCl (0.4 mL), SDS (0.1 g), distilled water (1 mL)] was added to completely elute the formazan from cells. The absorbance of formazan was measured at a wavelength of 595 nm using a microplate reader (Sunrise^™ Remote, Mӓnnedorf, Switzerland). The absorbance readings of the wells to which only DMEM was added instead of MTT solution were subtracted from each measurement as a background value.

TUNEL

Apoptosis was detected by Terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick end labeling (TUNEL) method using the DeadEnd^™ Fluorometric TUNEL system (Promega, WI, USA). Each well of 8 well chamber slide (Thermo Fisher Scientific) was coated with 300 μL of 0.001% Poly-L-Lysine (molecular weight 300,000, Fujifilm Wako Pure Chemical Corporation) by incubating at 37°C, 5% CO₂ for 1 hour. After incubation, wells were washed three times with PBS (+) (0.90 mM of CaCl₂ and 0.49 mM of MgCl₂ in PBS (−)) for 5 min, and MLg cells (5.0 × 10⁴ cells/well) were seeded and incubated for 4 hours at 37°C and 5% CO₂. Cells were then stimulated with 15 μM acrolein (ACR, Tokyo Chemical Industry Co., Ltd., Tokyo, Japan) for 30 min. After medium change, cells were stimulated with 5 μM of each SCGB3A2 peptide or 59.5 nM of SCGB3A2 for 16 hours. The medium was removed, and 300 μL of 4% PFA in PBS (+) was added to each well, and cells were fixed in a chamber for 15 min. Cells were washed with PBS, and cell membrane was permeabilized with 0.2% Triton-X-100 in PBS (−). After removing the chamber slide walls and washing cells with PBS (−), cells were reacted in TdT incubation buffer in a dry incubator at 37°C for 1 hour in dark. Cells were immersed in 2× SSC (saline-sodium citrate) for 15 min to stop the TdT reaction. Cells were washed with PBS (−), and nuclei were stained with 0.1 μg/mL of DAPI (4’, 6-Diamidino-2-phenylindole, dihydrochloride) (Thermo Fisher Scientific). Cells were sealed using CC/Mount (Diagnostic, CA, USA) and observed with an upright fluorescence microscope BX51 (Olympus). The TUNEL-positive rate (%) was determined by calculating the TUNEL-positive cells per total nuclei of five randomly selected fields in each well.

Statistical Analysis

The results obtained from the experiments other than qRT-PCR are presented as mean ± standard error of the mean (SEM) or standard deviation (SD). Statistical evaluation of differences was performed between groups using one-way ANOVA with Dunnett’s post-hoc test where appropriate. Post-hoc tests were run only if F achieved p < 0.05. Differences were considered to be statistically significant when p < 0.05. The analysis of qRT-PCR results was carried out using the Kruskal–Wallis test. If there was a significant difference among subgroups, pairwise comparison with the Steel-Dwass test was performed. The significance level was set at p < 0.05.

Results

Scgb3a2 gene expression during mouse lung development

Three types of SCGB3A2 protein, type A, B, and C, are naturally produced in vivo by alternative splicing of the mouse Scgb3a2 gene transcript (Supplementary Fig. S1) [1]. Quantitative RT-PCR was performed to examine gene expression of each Scgb3a2 transcript during mouse lung development. The results revealed that the expression of Scgb3a2 type A and type B increased from E16.5 through 8w, and type A was always the most highly expressed Scgb3a2 throughout gestation to adult (Fig. 1). The expression of Scgb3a2 type C was highest at E18.5. These results suggest that SCGB3A2 type A may be the most important for physiological functions in adult, and type C may not be actively expressed in adult.

Figure 1.

Scgb3a2 gene expression during mouse lung development. Gene expression levels of each Scgb3a2 type were determined by qRT-PCR at various developmental stages. Relative expression levels were normalized to 18S rRNA levels. n=3–5, each measured in triplicate. Representative results are shown from more than 4 times repeated experiments. Values are the mean ± standard deviation. *p < 0.05 by the Steel-Dwass post hoc test. N.S.: No significant difference.

Promotion of bronchial branching in fetal mouse lungs by SCGB3A2 type C and type A peptides

Since SCGB3A2 protein was previously shown to promote bronchial branching in mouse fetal lungs using an organ culture systems [3], the same system was used to investigate the effects of SCGB3A2 peptides. First, serially overlapping peptides covering the entire type C isoform were synthesized (Table 1), and ex vivo organ culture of fetal mouse lungs was performed to investigate the effect of these type C peptides on bronchial branching. The E11.5 mouse lungs (Day 0) were approximately 1 mm in diameter, and the average number of bronchial branches on Day 4 was 6.40 ± 0.27 (Supplementary Fig. S2A and B). The bronchial branching ratio with the addition of type C peptides Y91-L110 and F106-Q125 increased bronchial branching rate by 1.75 ± 0.2 and 1.85 ± 0.3-fold, respectively as compared to control, while SCGB3A2 increased bronchial branching fraction by 1.74 ± 0.3-fold (Fig. 2A, Supplementary Fig. S2A). Since these peptides are located towards the C-terminal region of SCGB3A2 amino acid sequence, various overlapping peptides covering the C-terminal region of type A sequence (from L54 to V91) were next synthesized (see Table 1) and subjected to bronchial branching analysis. The bronchial branching ratio of type A L54-V91, S57-A76, V62-K81, K67-A86, and L72-V91 peptides was 2.15 ± 0.5, 1.64 ± 0.5, 1.76 ± 0.3, 1.83 ± 0.6 and 1.59 ± 0.2, respectively when the control value was set as 1 (Fig. 2B, Supplementary Fig. S2B). The bronchial branching ratio by addition of type B peptide was 0.84 ± 0.1 and that by SCGB3A2 was 2.24 ± 0.7 (Fig. 2B, Supplementary Fig. S2B). These results demonstrate that type A peptides covering L54 through V91 significantly increase bronchial branching.

Figure 2.

Promotion of bronchial branching by SCGB3A2 peptides in fetal mouse lungs. A. Rate of increase in bronchial branching in mouse fetal lungs of SCGB3A2-treated (SCGB3A2) and various SCGB3A2 type C peptide-treated groups as indicated based on the control (cont) set as 1.0. Significantly increased bronchial branching rate was obtained with type C peptides Y91-L110 and, F106-Q125. N.S.: No statically significant difference vs. control. N=1–3 per experiment, and experiments were repeated 4 times. All data points are included in the final analysis. Values are the mean ± standard deviation. Dunnett’s post-hoc test, **p < 0.01, ***p < 0.001 vs. control.

B. Percentage increase in bronchial branching in mouse fetal lungs of SCGB3A2 (SCGB3A2), various SCGB3A2 type A peptides, and type B G94-V113-treated groups compared to control (cont). Significantly increased bronchial branching rate was obtained with type A peptides L54-V91, S57-A76, and K67-A86. Type B G94-V113 not significantly different from control (N.S.). N=1–3 per experiment, and experiments were repeated 6 times. All data points are included in the final analysis. Values are the mean ± standard deviation. Dunnett’s post-hoc test, *p < 0.05, **p < 0.01, ***p < 0.001 vs. control.

Effect of various SCGB3A2 type C and A peptides on cell proliferation

We previously demonstrated that SCGB3A2 binds to the cell surface of primary lung mesenchymal cells [3]. In the current study, mouse lung fibroblast-derived MLg cells were used to examine the effect of SCGB3A2 in various in vitro analyses. First, the effect of various type C and A peptides on the proliferation of MLg cells was examined using MTT assay.

The cell proliferation rates of MLg cells in the groups of type C peptides M1-T20 and F106-Q125 significantly increased the proliferation rates of MLg cells (126 ± 1.7% and 120 ± 1.9%, respectively), similar to the addition of SCGB3A2 (120 ± 6.5%) as compared with non-treatment group (cont as 100%) (Fig. 3A). The cell proliferation rates of type A L54-V91, S57-A76, V62-K81, K67-A86, and L72-V91 peptides were 107 ± 0.9%, 110 ± 1.7%, 111 ± 1.2%, 111 ± 1.3%, and 107 ± 0.5%, respectively, based on the proliferation rate of non-treatment group (cont) set as 100% (Fig. 3B). These proliferation rates showed a significant increase similar to the addition of SCGB3A2 (115 ± 1.0%). Type A peptide L72-V91 examined in macrophage-like RAW264 cells exhibited a 125 ± 8.8% proliferation rate compared to untreated RAW264 cells (cont as 100%) (Supplementary Fig. S3B). Although there was no statistical difference obtained, the results suggest that this activity may possibly be cell type independent (see Table 2), and that the C-terminal amino acid sequence of SCGB3A2 may be related to cell proliferation. However, when the type B G94-V113 peptide was subjected to cell proliferative assay using MLg and RAW264 cells, the proliferation rates of both were significantly decreased (Supplementary Fig. S3A, B). MLg cells treated with type B peptide G94-V113 showed fibrous morphology and appeared to be disintegrating (Supplementary Fig. S3C). Type C peptide M1-T20, even though exhibited cell proliferation activity, corresponds to the signal peptide and therefore was eliminated from further analysis.

Figure 3.

Effect of SCGB3A2 type C and type A peptides on cell proliferation of MLg cells. Significant differences were found in group with SCGB3A2 (SCGB3A2), type C peptides M1-T20 and F106-Q125 (A), and all five type A peptides (B) as compared to control (cont). Representative results are shown from 4–6 times repeated experiments (n=5–12 per group). Values are the mean ± standard error. Dunnett’s post-hoc test, *p < 0.05, **p < 0.01, ***p < 0.001 vs. control.

Table.2.

Summary of effect of SCGB3A2 and various SCGB3A2 type A and type C peptides

			bronchial branching rate (fold)*1	change	cells used	proliferation (%) *1	change	cells used: MLg	TUNEL positive rate (%)	change	anti-inflammatory effect in vivo
Control (no stimulation)			1		MLg/RAW 264	100		+ACR	11.0 ± 0.5*2		PBS +cOV A*4	inflammatory reaction
protein	SCGB3A2		1.93 ± 0.7*3	+	MLg RAW 264	115 ± 0.7*3	+	+ACR	3.56 ± 0.8*3	+	+cOVA	normal
						145 ± 12	+
peptides	type A	L54-V91	2.15 ± 0.5	+	MLg	107 ± 0.9	+	+ACR	6.45 ± 0.4	+	+cOVA	normal
		S57-A76	1.64 ± 0.5	+		110 ± 1.7	+		6.80 ± 0.5	+		normal
		V62-K81	1.76 ± 0.3	N.S.		111 ± 1.2	+		7.83 ± 0.3	+		almost normal*5
		K67-A86	1.83 ± 0.6	+		111 ± 1.3	+		8.60 ± 0.6	N.S.		almost normal*5
		L72-V91	1.59 ± 0.2	N.S.		107 ± 0.5	+		6.76 ± 0.5	+		normal
					RAW 264	126 ± 8.8	N.S.
	type B	G94-V113	0.84 ± 0.1	N.S.	MLg RAW 264	50.0 ± 3.1	-	+ACR	5.13 ± 1.9	+	N.A.	N.A.
						62.2 ± 8.0	-					N.A.
	type C	M1-T20	0.91± 0.1	N.S.	MLg	126 ± 1.7	+	+ACR	N.A			N.A.
		G16-V35	0.64 ± 0.1	N.S.		102 ± 2.0	N.S.		N.A			N.A.
		D31-L50	1.00 ± 0.2	N.S.		106 ± 5.5	N.S.		N.A		N.A.	N.A.
		P46-L65	0.75 ± 0.1	N.S.		104 ± 5.1	N.S.		N.A			N.A.
		L61-V80	0.85 ± 0.1	N.S.		108 ± 4.1	N.S.		N.A			N.A.
		A76-R95	1.27 ± 0.1	N.S.		105 ± 2.4	N.S.		7.86 ± 1.4	N.S.	+cOVA	normal
		Y91-L110	1.75 ± 0.2	+		97.1 ± 2.1	N.S.		10.6 ± 0.0	N.S.	N.A.	N.A.
		F106-Q125	1.85 ± 0.3	+		120 ± 1.9	+		4.73 ± 1.3	+	+cOVA	almost normal*5
		D121-L139	1.49 ± 0.1	N.S.		112 ± 6.8	N.S.		10.5 ± 2.8	N.S.	N.A.	N.A.

^*1+ : Significant increases in proliferative and bronchial branching promoting effects vs. control (no stimulation).

- : Significant decrease in proliferative and bronchial branching promoting effects vs. control (no stimulation).

N.S. : No significant difference.

^*2Average value obtained from the experiments carried out for various type A and C peptides without SCGB3A2 (see Fig. 4)

^*3Average value obtained from the experiments carried out for various type A and C peptides with SCGB3A2 (see Fig. 3)

^*4PBS refers to the group that received PBS after allergy induction with cOVA as inflammation positive control (see Fig. 5).

^*55 Lungs of some mice showed weak inflammation (see Fig. 5).

N.A.: not available

Inhibition of apoptosis by SCGB3A2 type C and type A peptides

To evaluate the inhibitory effect of SCGB3A2 C-, A-, and B-type peptides on apoptosis, TUNEL assay was performed using an acrolein (ACR)-induced apoptosis model generated by stimulating MLg cells with ACR, a substance that causes COPD [27]. First, the apoptosis inhibitory effect of type C peptides was examined. The positive rate of TUNEL in the ACR-induced apoptosis group without any addition (SCGB3A2-/ACR+) was 12.3 ± 0.9%, whereas the positive rate of TUNEL in the presence of A76-R95 and F106-Q125 was 7.86 ± 1.4% and 4.73 ± 1.3%, respectively. The latter rate was significantly lower and at the same level as when SCGB3A2 protein was added (4.73 ± 0.7%) (Fig. 4A and Supplementary Fig. S4A). Next, the apoptosis inhibitory effect of type A and B peptides was examined. The TUNEL positive rate by ACR was 10.3 ± 0.5% without any addition (SCGB3A2-/ACR+), and its positivity reduced to 5.29 ± 0.4% when SCGB3A2 was added (Fig. 4B). The type A peptides that statistically significantly reduced TUNEL positivity were L54-V91, S57-A76, V62-K81, and L72-V91 (Fig. 4B and Supplementary Fig. S4B). Type B peptide G94-V113 also significantly reduced the TUNEL positive rate (5.13 ± 1.9%).

Figure 4.

Anti-apoptotic effect of SCGB3A2 peptides. A. Statistically significant reduction of TUNEL-positive rate obtained in the presence of type C peptide F106-Q125, the latter at the same extent as SCGB3A2 (light gray column). Representative results are shown from more than 3 times repeated experiments (n=1–3 per group).

B. Significant reduction of TUNEL-positive cells by type A peptides L54-V91, S57-A76, V62-K81, and L72-V91 and type B peptide Q94-V113 vs. ACR (+) without SCGB3A2 (dark gray column). N=1–2 per experiment, and experiments were repeated 3 times. Both of figures, all data points are included in the final analysis. The apoptosis-positive rate was calculated by the number of TUNEL-positive cells to the number of total nuclei. Values are the mean ± standard error. Dunnett’s post-hoc test, **p<0.01, ***p<0.001 as compared to ACR (+) without SCGB3A2 (dark gray column). Black column: cont (control), Gray column: with ACR (+) and SCGB3A2. N.S.: no significant difference.

Improvement of allergic airway inflammation by SCGB3A2 peptides in a mouse model of allergic airway inflammation

Based on the results obtained in the above experiments, SCGB3A2 type A L54-V91, S57-A76, V62-K81, K67-A86, and L72-V91 peptides, and type C peptide A76-R95, F106-Q125, and SCGB3A2 protein as control were chosen for the following in vivo analysis. These peptides and the SCGB3A2 protein were intratracheally administered to the mouse model of allergic airway inflammation to verify their anti-inflammatory effect (Supplementary Fig. S5).The total cell counts, eosinophils, neutrophils, and lymphocytes in BALF of most of the SCGB3A2 peptides-treated groups were reduced to levels as comparable to the SCGB3A2-treated group (Fig. 5A). Histologically, infiltration of inflammatory cells was observed in the periphery of bronchi and blood vessels, and in lung parenchyma of cOVA-treated and PBS-administered group (Fig. 5B). In addition, bronchial epithelial cells of this group were hyperplastic and PAS-positive (Fig. 5C). No or almost no inflammation was observed in H&E-stained mouse lung tissues of SCGB3A2 type A peptides administered group. However, some bronchial epithelial cells from V62-K81 and K67-A86-treated mice showed hyperplasia-like structures, similar to those in the PBS treated group (Fig. 5B. white arrowheads), and they were PAS-positive (Fig. 5C. black arrowheads). Little inflammation was observed in the lungs of mice in the SCGB3A2 type C peptide A76-R95 treated-group. Some bronchial epithelial cells from type C peptide F106-Q125-treated mice showed hyperplasia-like structures (Figure 5B, white arrowheads) and PAS positive epithelial cells (Figure 5C, black arrowheads). Lung tissue inflammation and PAS scores showed that all SCGB3A2 type A and type C peptide treatment groups statistically significantly reduced inflammation (Fig. 5 D). These results suggest that three SCGB3A2 type A peptides (L54-V91, S57-A76, and L72-V91) and one type C peptide A76-R95 exhibit the same or similar suppresive effects to SCGB3A2 protein on allergic airway inflammation in the lungs caused by cOVA.

Figure 5.

Evaluation of anti-inflammatory effect of SCGB3A2 peptides using a mouse model of allergic airway inflammation. A. Measurement of cell counts in bronchoalveolar lavage fluid (BALF). Black column : no allergy induced by cOVA (Control), dark gray column: PBS administration under inflammation induced by cOVA, gray column: SCGB3A2 administration under inflammation induced by cOVA, white column: each peptide administration under inflammation induced by cOVA. Dots in bars: average value per mouse. Values are the mean ± standard deviation (total n=3 −10 per group). Dunnett’s post-hoc test, * p < 0.05, ** p < 0.01, *** p < 0.001 for PBS with cOVA (dark grey column) vs. SCGB3A2 or each peptide. N.S.: no significant difference.

B and C. Histological observation of cOVA-induced asthmatic mouse lungs (B: H&E, C: PAS staining). Each photograph is a representative image of each group. cont: control, no allergy induced, and under inflammation induced by cOVA with administration of PBS, SCGB3A2, or peptides A (type A): L54-L91, A:S57-A76, A:V62-K81, A:K67-A86, A:L72-V91, and C (type C): A76-R95 and F106-Q125. White arrowheads: Hyperplasia-like bronchial epithelial cells, Black arrowheads: PAS positive signal. Bars: 100 μm.

D. Histological scoring of inflammatory cell infiltration and goblet cells. Dunnett’s post-hoc test, * p < 0.05, ** p < 0.01 *** p < 0.001 for PBS with cOVA (dark grey column) vs. SCGB3A2 or each peptide. N.S.: no significant difference. This animal experiment was performed with 1–3 mice per group per an experiment and repeated at least three times. The F106-Q125 administration experiment was conducted once. All data points are included in the final analysis.

Discussion

One of the most important aspects of drug development is to achieve the recommended limit of 5 EU/kg for endotoxin contamination (https://www.fda.gov/drugs/pharmaceutical-quality-resources/questions-and-answers-quality-related-controlled-correspondence). In our previous studies, when rSCGB3A2 was biosynthesized using E. Coli and was subjected to standard endotoxin removal steps, the yield of rSCGB3A2 was drastically reduced. Recently, it was reported that SCGB3A2 is an LPS-binding protein and a SCGB3A2-LPS complex binds to the SCGB3A2 receptor SDC1, through which LPS is delivered into lung cancer cells and macrophages, resulting in pyroptosis, inflammatory cell death of these cells [13,14]. This mechanism might also play a role in the maintenance of lung homeostasis through LPS removal and pyroptosis in vivo [13]. In order to circumvent the endotoxin contamination problem that inevitably accompanies bacterially-purified proteins, in this study, SCGB3A2 peptides were chemically synthesized, and the bioactivities of SCGB3A2 peptides were examined to determine the potential use of SCGB3A2 peptides as a drug candidate to treat lung diseases.

The amino acid sequence from M1 to L84 perfectly matches between the A and C isotypes of SCGB3A2 that are derived from the shared sequence of exons 1 and 2 (Supplementary Fig. S1). SCGB1A1, which is highly homologous to SCGB3A2, is expected to be applied to neonatal distress respiratory syndrome [28–31]. Since SCGB3A2 type C is highly expressed during lung developmental stage, we investigated the physiological activities of SCGB3A2 type C as well as type A peptides to explore their potential as a drug to treat lung diseases that may include neonatal respiratory syndrome and other conditions related to fetal lung development. When a series of peptides derived from the type C and type A C-terminal amino acid sequence were subjected to cell proliferation assays of MLg cells, all type A peptides covering from L54 to V91 promoted cell proliferation rates, whereas type C peptides L61-V80 and A76-R95 did not enhance cell proliferation promotion activity even though they cover almost same areas of sequence as type A peptides. The reason for this discrepancy is not known. Nevertheless, the results obtained from the type A and C peptides suggest that the C-terminal portion of the SCGB3A2 amino acid sequences may play a role in its bioactivities.

SDC1 has been reported to be expressed in many normal tissues [32]. In the lung, its expression is mainly reported in epithelial tissues [13,32,33], however it is also expressed in human fibroblast cell lines [34]. We have observed SDC1 expression in the mouse lung tissues, not only in bronchial and bronchiolar epithelial cells, but also in vascular endothelial cells, muscle layer, alveoli, and MLg cells (Supplementary Fig. S6). Furthermore, the binding of SCGB3A2 peptide (type A L54-V91) to SDC1 was confirmed by immunocytochemistry and co-immunoprecipitation using MLg cells (Supplementary Fig. S7). These results suggest that the observed bioactivities of SCGB3A2 peptides might be due to their binding to SDC1. The widespread expression of SDC1 may provide the possibility that SCGB3A2 and its peptides could act in tissues and cells other than lungs if given systematically.

On the other hand, when SCGB3A2 type B peptide G94-V113 was used to examine its effect on cell proliferation, it was reduced to about 50% without apparent reduction in cell numbers as observed under microscopy. MTT assay calculates cell proliferation rate by measuring the relative cell numbers based on the reducing ability of mitochondrial dehydrogenase [35]. The decrease in cell proliferation rate by type B peptide G94-V113 without cell number changes raises a possibility that type B peptide G94-V113 may decrease mitochondrial function. Alternatively, elongated cells with fibrous projections are indeed dead cells. The type B results support the specificity of each peptide. Due to unknown reasons, SCGB3A2 type B peptide was easily degraded and unstable (data not shown), and the synthesis of other peptides derived from SCGB3A2 type B-specific amino acid sequences was unsuccessful.

The amino acid residues M60-L69 (corresponding to M39-L48 in mature secreted form) of the rabbit UG (SCGB1A1) are known to be well conserved in all animal species including mice [15]. Peptides in this region, called antiflammin, were reported to exhibit anti-inflammatory and immunomodulatory activities by inhibiting phospholipase A2 (PLA2) activity [15,16,18,19,36,37]. It was reported that human and mouse SCGB3A2 also exhibit PLA2 inhibitory activity [38]. SCGB3A2 type A peptide L72-V91 and type C peptide A76-R95 suppressed mouse lung inflammation in animal experiments at similar extent to SCGB3A2 protein as shown in this study, however, these peptide sequences are outside the T63-L72 mouse antiflammin sequence. The results indicate that peptides other than the rabbit amino acid sequence of M60-L69 may exhibit bioactivity and anti-inflammatory activity. The predicted conformation of SCGB3A2 revealed that SCGB3A2 has four α-helices similar to SCGB1A1 to form a dimer, and the type A peptide L72-V91 was predicted to include the fourth α-helix [13,19,39,40]. Furthermore, CD spectral analysis confirmed that the type A peptides L54-V91 and L72-V91 form an α-helix, and the L54-V91 peptide appears to have the most stable structure (Supplementary Fig. S8). In general, the three-dimensional structure of a protein is important for its function and thus must be considered for the application of peptides as drugs. How these peptides exert their anti-inflammatory function remains to be fully understood.

Protein glycosylation is a post-translational modification taking place in vivo that has a significant impact on protein function. It was reported that glycosylation improves membrane permeability and metabolic stability of peptides in vivo [41,42]. The status of glycosylation, especially within the C-terminal region of SCGB3A2 protein is currently unknown. If indeed it is glycosylated, this could enhance the bioactivity and therapeutic efficacy of the peptide. Moreover, in general, peptide drugs are susceptible to degradation in vivo [23]. In the current study, SCGB3A2 peptides were delivered directly to the target lungs by intratracheal administration, which may have been protected from digestion and/or degradation. Specialized inhalers are already used for bronchial asthma and chronic obstructive pulmonary disease (COPD) [43], and the use of such inhalers is also possible for delivery of SCGB3A2 peptides for human use.

The in vivo activity of SCGB3A2 peptides was validated for mouse type A peptide L54-V91, S57-A76, V62-K81, K67-A86, L72-V91, and SCGB3A2 type C peptide A76-R95 and F106-Q125. Although SCGB3A2 type C is not expressed in adult mice, we think that if the type C peptide shows an anti-inflammatory effect when administered, it could be valuable as a new drug especially for diseases that occur during fetal and neonatal stages of lungs [28–31]. However, the SCGB3A2 type C peptide F106-Q125, even though exhibiting promotion of bronchial branching and cell proliferation, and apoptosis inhibition, its in vivo anti-inflammatory effect was weaker than that of other peptides.The number of mice analyzed might need to be increased and experiments repeated to obtain more accurate and valid information. Further, type A L54-V91 and L72-V91 showed effects at similar degree to SCGB3A2 protein in terms of inflammatory cells count in BALF, suggesting that these peptides may be large enough to maintain the functionality. When amino acid sequences of mouse type A peptide L54-V91 are aligned with the corresponding part of the human SCGB3A2 sequence, they demonstrated 92.1% identity with only three amino acid residues difference (see Fig. S1; mT63 => hE65, mK66 => hR68, mD70 => hN72). The amino acid sequence of mouse type A peptide L72-V91 is 100% identical to human type A, suggesting the high probability of this peptide or L54-V91, or the human equivalent of mouse L54-V91 being functional as an anti-inflammatory agent in treating lung diseases in humans. Based on the results obtained, it appears likely that this peptide or other SCGB3A2 peptides may also improve pulmonary fibrosis and/or COPD. Future experiments are required to address these questions. The current study presented the potential of SCGB3A2 type A L72-V91 and L54-V91 as effective peptide drugs for the treatment of allergic airway inflammation.

Data Sharing Statement

All data are provided in the manuscript and Supporting Information, or are available from the corresponding authors upon reasonable request.