Collagen Peptides Derived from Sipunculus nudus Accelerate Wound Healing

Marine collagen peptides have high potential in promoting skin wound healing. This study aimed to investigate wound healing activity of collagen peptides derived from Sipunculus nudus (SNCP). The effects of SNCP on promoting healing were studied through a whole cortex wound model in mice. Results showed that SNCP consisted of peptides with a molecular weight less than 5 kDa accounted for 81.95%, rich in Gly and Arg. SNCP possessed outstanding capacity to induce human umbilical vein endothelial cells (HUVEC), human immortalized keratinocytes (HaCaT) and human skin fibroblasts (HSF) cells proliferation and migration in vitro. In vivo, SNCP could markedly improve the healing rate and shorten the scab removal time, possessing a scar-free healing effect. Compared with the negative control group, the expression level of tumor necrosis factor-α, interleukin-1β and transforming growth factor-β1 (TGF-β1) in the SNCP group was significantly down-regulated at 7 days post-wounding (p < 0.01). Moreover, the mRNA level of mothers against decapentaplegic homolog 7 (Smad7) in SNCP group was up-regulated (p < 0.01); in contrast, type II TGF-β receptors, collagen I and α-smooth muscle actin were significantly down-regulated at 28 days (p < 0.01). These results indicate that SNCP possessed excellent activity of accelerating wound healing and inhibiting scar formation, and its mechanism was closely related to reducing inflammation, improving collagen deposition and recombination and blockade of the TGF-β/Smads signal pathway. Therefore, SNCP may have promising clinical applications in skin wound repair and scar inhibition.


Introduction
Skin wounds are mainly caused by external forces, such as accidents, burns, traumas and genetic diseases, resulting in the destruction of skin tissue and cell arrangement, rupture of blood vessels and loss of blood [1]. Wounds are easily infected after suffering injury and in the absence of good care, these could easily develop into chronic wounds (such as diabetic ulcers). This can cause severe physical dysfunction and mental damage among patients. In total, approximately 2% of people in the United States suffer from kDa, respectively [26]. The different molecular weight of α chain in collagen might be due to the different class of Sipuncula Sedgwick. The patterns of the two α chains and one β dimer were similar to most collagens from fishes [27,28]. Type-I collagen could be applied in pharmaceutical, food and biomedical industries [29].

SNCP is Rich in Gly and Arg
As shown in Table 1, Gly accounted for around 36% of the total hydrolyzed amino acid compositions, consistent with the Gly-X-Y repeating sequence of collagen [37,38]. Previous studies showed that Arg and Pro possessed the potential effect of promoting wound healing [19,39]. In this work, SNCP were also rich in Glx (Glu + Gln), Ala, Arg, Thr, Ser and Pro, which might contribute to the potential wound healing effect of SNCP. Moreover, the majority of the amino acid residues were hydrophilic, such as Glx (Glu + Gln), Arg, Asx (Asp + Asn) and Pro. This was consistent with the good water solubility of SNCP. The major free amino acids in SNCP were Arg (0.81 g/100 g), Leu (0.51 g/100 g), Tyr (0.42 g/100 g) and Pro (0.34 g/100 g); among them, Arg, which accounted for 19.36% of the total free amino acids, generally reflects the wound healing activity.  The Fourier transform infrared (FTIR) spectra of SNC exhibited five characteristic bands of the amide structure, namely A, B, I, II and III, as indicated in Figure 1b. The amide A band was found at 3340 cm −1 , which indicates the stretching vibration of NH groups related to the hydrogen bonds. The amide B band found at 2931 cm −1 was the characteristic absorption of collagen different from other proteins, which was generally observed at 2930-2944 cm −1 , causing by the asymmetric stretching vibration of -CH 2 . The strong absorption peak of amide I band (1600-1700 cm −1 ) was observed at 1638 cm −1 attributing to the C=O stretching vibration involved in the secondary structure of the peptide chain [30]. The amide II band and amide III band were mainly observed at around 1556 cm −1 and 1241 cm −1, respectively, corresponding to the the N-H bending vibration and the in-phase C-N stretching vibration coupled with N-H bending vibration, which was responsible for collagen helical integrity [31,32]. The presence of the amide III band clearly demonstrated the existence of the triple helix structure in the SNC. Altogether, the FTIR spectra indicate a well-maintained secondary structure in SNC.
The differential scanning calorimeter (DSC) thermogram ( Figure 1c) indicate that the denaturation temperature of SNC was 32.8 • C, which was similar to the acid-soluble collagen from carp fish scale waste (32.2 • C) [27], fresh carp fish scale (32.9 • C), giant grouper (31.7 • C) and big eye snapper (31 • C) [33,34]. It is reported that the thermal denaturation temperature of collagen was mainly related to the amino acid composition, protein conformation, water content, temperature of habitat and extraction method [32,35,36].

SNCP Consists of Peptides with Small Molecular Weight Less Than 5 kDa
The chromatogram of SNCP is shown in Figure 1d, respectively. With the retention time (t) as the horizontal axis and the lg Mw as the vertical axis, the regression equation lgMw = −0.2441t + 7.08 (R 2 = 0.99) was obtained. The components of less than 1, 3, 5 and 10 kDa accounted for approximately 34.36%, 67.13%, 81.95% and 95.13%, respectively. Compared with SNC, SNCP with smaller molecular weight might possess better water solubility and facilitated absorption.

SNCP is Rich in Gly and Arg
As shown in Table 1, Gly accounted for around 36% of the total hydrolyzed amino acid compositions, consistent with the Gly-X-Y repeating sequence of collagen [37,38]. Previous studies showed that Arg and Pro possessed the potential effect of promoting wound healing [19,39]. In this work, SNCP were also rich in Glx (Glu + Gln), Ala, Arg, Thr, Ser and Pro, which might contribute to the potential wound healing effect of SNCP. Moreover, the majority of the amino acid residues were hydrophilic, such as Glx (Glu + Gln), Arg, Asx (Asp + Asn) and Pro. This was consistent with the good water solubility of SNCP. The major free amino acids in SNCP were Arg (0.81 g/100 g), Leu (0.51 g/100 g), Tyr (0.42 g/100 g) and Pro (0.34 g/100 g); among them, Arg, which accounted for 19.36% of the total free amino acids, generally reflects the wound healing activity.

SNCP Possesses an Outstanding Capacity to Induce HUVEC, HaCaT and HSF Cells Proliferation and Migration
Endothelial cell, keratinocytes and fibroblasts are essential for effective angiogenesis, re-epithelialization and extracellular matrix (ECM) synthesis during the process of wound healing. In vitro scratch assay results showed that, compared with the negative control (NC) group, SNCP markedly promoted the wound closures by cell proliferation and migration in three cell strains, including human umbilical vein endothelial cells (HUVEC), human immortalized keratinocytes (HaCaT) and human skin fibroblasts (HSF) (Figure 2a,c,e). The wound closure percentages in HUVEC (97.54 ± 2.30%, in 12 h), HaCaT (96.13 ± 2.70%, in 36 h) and HSF cells (96.68 ± 5.08%, in 30 h), incubated with SNCP, were significantly higher than the ones of NC groups, respectively (p < 0.01) (Figure 2b,d,f). These results demonstrated that SNCP possessed outstanding capacity to induce HUVEC, HaCaT and HSF cells proliferation and migration.  Figure 3 shows the effect of SNCP on wound healing in vivo. There was no significant difference observed in the wound in 2 days. The significant promoting effect of wound healing on SNCP was observed at 4 days after modeling. The mice groups treated with the SNCP had decreased wound area compared to the negative control (NC) group ( Figure 3a). As shown in Figure 3b, a markedly higher wound closure rate was observed in SNCP group 2 days post-wounding, versus the NC group and positive control (PC) group, respectively. At 10 days post-wounding, wound closure rate of the SNCP group was (91.41 ± 5.77%), while for the NC group it was (58.48 ± 17.21%). The time of scab removal of each group is shown in Figure 3c. A markedly shorter time of scab removal   Figure 3b, a markedly higher wound closure rate was observed in SNCP group 2 days post-wounding, versus the NC group and positive control (PC) group, respectively. At 10 days post-wounding, wound closure rate of the SNCP group was (91.41 ± 5.77%), while for the NC group it was (58.48 ± 17.21%). The time of scab removal of each group is shown in Figure 3c. A markedly shorter time of scab removal was seen in the SNCP group (11.6 ± 0.6 d) versus in the PC group (14.5 ± 2.5 d) and in the NC group (18.0 ± 1.8 d). Importantly, SNCP group improved wound appearance. The surface of SNCP group in 28 days was smooth and flat without an obvious scar, whereas NC group developed noticeable large scars. surface of SNCP group in 28 days was smooth and flat without an obvious scar, whereas NC group developed noticeable large scars.

SNCP Reduces Excessive Inflammation Response
The HE stained section was used to determine the effect of SNCP on the re-epithelialization and contraction of the wound surface at 7 days post-wounding (Figure 4a). The distance of double arrows and red arrows represent the width of the wound surface and the length of newly-formed epidermal, respectively. During the course of wound contraction and re-epithelialization in the early stage of wound healing, the width of wound surface of SNCP group (1.99 ± 0.07 mm) was significantly smaller than that of the NC group (4.58 ± 0.48 mm) and the PC group (3.51 ± 0.21 mm) ( Figure 4b) and the lengths of the newly-formed epidermal of SNCP group and PC group were (3.00 ± 0.39 mm) and (2.47 ± 0.24 mm), respectively, which are significantly longer than that of the NC group (1.26 ± 0.42 mm) ( Figure 4b). As shown in Figure 4c, compared with the NC group, the SNCP group and the PC group reduced the mRNA levels of TGF-β1, TNF-α and IL-1β, and the mRNA level of TNF-α in the SNCP group was significantly lower than that in the PC group (p < 0.01). The decrease of these pro-inflammatory factors indicated that SNCP possessed an anti-inflammatory function, which was beneficial to rapid wound healing and the prevention of abnormal scar formation.

SNCP Reduces Excessive Inflammation Response
The HE stained section was used to determine the effect of SNCP on the re-epithelialization and contraction of the wound surface at 7 days post-wounding ( Figure 4a). The distance of double arrows and red arrows represent the width of the wound surface and the length of newly-formed epidermal, respectively. During the course of wound contraction and re-epithelialization in the early stage of wound healing, the width of wound surface of SNCP group (1.99 ± 0.07 mm) was significantly smaller than that of the NC group (4.58 ± 0.48 mm) and the PC group (3.51 ± 0.21 mm) ( Figure 4b) and the lengths of the newly-formed epidermal of SNCP group and PC group were (3.00 ± 0.39 mm) and (2.47 ± 0.24 mm), respectively, which are significantly longer than that of the NC group  Figure 4c, compared with the NC group, the SNCP group and the PC group reduced the mRNA levels of TGF-β1, TNF-α and IL-1β, and the mRNA level of TNF-α in the SNCP group was significantly lower than that in the PC group (p < 0.01). The decrease of these pro-inflammatory factors indicated that SNCP possessed an anti-inflammatory function, which was beneficial to rapid wound healing and the prevention of abnormal scar formation.

SNCP Improves the Healing Quality of Wound Epidermis
A thinner and flatter epidermis was observed in the SNCP group and PC group at 14 and 28 days post-wounding, while an uneven and thickened epidermis with tongue-like protrusions on the edges was observed in the NC group ( Figure 5a). The epidermal thickness index of the SNCP group and PC group were (1.23 ± 0.18) and (1.96 ± 0.44), which were significantly lower than in the NC group (3.82 ± 0.42) at 28 days post-wounding ( Figure 5b). An epidermal thickness index of more than 1 means that the epidermis was thicker than normal skin. The thickness ratio of the thinner and thicker epidermis (thin/thick ratio) was used to characterize whether the epidermis was recovered uniformly and flatly. At 28 days post-wounding, the thin/thick ratio of the epidermis in the SNCP group and PC group were (0.67 ± 0.04) and (0.41 ± 0.04), respectively, which were significantly higher than in the NC group (0.21 ± 0.10) (Figure 5c).

SNCP Improves the Healing Quality of Wound Epidermis
A thinner and flatter epidermis was observed in the SNCP group and PC group at 14 and 28 days post-wounding, while an uneven and thickened epidermis with tongue-like protrusions on the edges was observed in the NC group (Figure 5a). The epidermal thickness index of the SNCP group and PC group were (1.23 ± 0.18) and (1.96 ± 0.44), which were significantly lower than in the NC group (3.82 ± 0.42) at 28 days post-wounding ( Figure 5b). An epidermal thickness index of more than 1 means that the epidermis was thicker than normal skin. The thickness ratio of the thinner and thicker epidermis (thin/thick ratio) was used to characterize whether the epidermis was recovered uniformly and flatly. At 28 days post-wounding, the thin/thick ratio of the epidermis in the SNCP group and PC group were (0.67 ± 0.04) and (0.41 ± 0.04), respectively, which were significantly higher than in the NC group (0.21 ± 0.10) (Figure 5c).

SNCP Improves Collagen Deposition and Recombination at the Wound Dermis
As shown in Figure 6a, the formation of skin appendages (such as hair follicles, sebaceous glands and blood vessels) were not observed in the NC group at 14, 21 and 28 days, and the collagen was mainly arranged in parallel. However, blood vessel formation and collagen deposition were observed at the dermis of PC group at 21 days post-wounding, and a basket-like collagen distribution was formed at a later stage. Compared with the NC group and the PC group, sebaceous glands were produced on the dermis of the SNCP group at 14 days, and mature granulation tissue, including blood vessels, hair follicles and sebaceous glands, was produced at 21 days, which greatly improved collagen deposition and formed an orderly arrangement of a basket weave-like structure, and gradually showed the characteristics of normal dermis. The dermis of the SNCP group was similar to the normal healthy skin at 28 days (Figure 6a). In addition, the COL1A1 level in wound tissue of mice treated with SNCP was significantly higher than that of NC group on the 14th day after injury (p < 0.01), while the COL1A1 level was significantly down-regulated on the 21st and 28th day (p < 0.01) (Figure 6b).
To further observe the collagen fibers during tissue remodeling, the ultrastructure of collagen fibers in the wound was identified by transmission electron microscope (TEM) at 21 days post-wounding. The average diameter of collagen fibers in dermis of the NC group, the SNCP group, the PC group and normal skin were 108.7 nm, 58.9 nm, 52.3 nm and 59.5 nm, respectively. The SNCP and PC groups showed a similar appearance to normal skin, with relatively round collagen fibers and clear contours, while the NC group show obvious structural abnormalities (indicating by the red arrow in Figure 6c),

SNCP Improves Collagen Deposition and Recombination at the Wound Dermis
As shown in Figure 6a, the formation of skin appendages (such as hair follicles, sebaceous glands and blood vessels) were not observed in the NC group at 14, 21 and 28 days, and the collagen was mainly arranged in parallel. However, blood vessel formation and collagen deposition were observed at the dermis of PC group at 21 days post-wounding, and a basket-like collagen distribution was formed at a later stage. Compared with the NC group and the PC group, sebaceous glands were produced on the dermis of the SNCP group at 14 days, and mature granulation tissue, including blood vessels, hair follicles and sebaceous glands, was produced at 21 days, which greatly improved collagen deposition and formed an orderly arrangement of a basket weave-like structure, and gradually showed the characteristics of normal dermis. The dermis of the SNCP group was similar to the normal healthy skin at 28 days (Figure 6a). In addition, the COL1A1 level in wound tissue of mice treated with SNCP was significantly higher than that of NC group on the 14th day after injury (p < 0.01), while the COL1A1 level was significantly down-regulated on the 21st and 28th day (p < 0.01) (Figure 6b).
To further observe the collagen fibers during tissue remodeling, the ultrastructure of collagen fibers in the wound was identified by transmission electron microscope (TEM) at 21 days post-wounding. The average diameter of collagen fibers in dermis of the NC group, the SNCP group, the PC group and normal skin were 108.7 nm, 58.9 nm, 52.3 nm and 59.5 nm, respectively. The SNCP and PC groups showed a similar appearance to normal skin, with relatively round collagen fibers and clear contours, while the NC group show obvious structural abnormalities (indicating by the red arrow in Figure 6c), including unclear contours between multiple collagens, tending to focal lateral fusion to form larger collagen fibers. As shown in Figure 6d, the fiber diameter distribution in skin of the SNCP, PC and normal skin groups were relatively concentrated, ranging from 30 mm to 80 nm, while the NC group had a wide distribution range (60-160 nm).
including unclear contours between multiple collagens, tending to focal lateral fusion to form larger collagen fibers. As shown in Figure 6d, the fiber diameter distribution in skin of the SNCP, PC and normal skin groups were relatively concentrated, ranging from 30 mm to 80 nm, while the NC group had a wide distribution range (60-160 nm).

SNCP Blockades the TGF-β/Smad Signaling Pathway to Achieve Scarless Healing
To investigate the effect of SNCP on TGF-β/Smad signaling pathway, the mRNA expression of transforming growth factor-β1 (TGF-β1), transforming growth factor-β receptor II (TGF-βRII), α-smooth muscle actin (α-SMA) and Smad7 in the wound at 28 days post-wounding, were investigated by qRT-PCR. As shown in Figure 7, compared with the NC group, the mRNA levels of Smad7 in both the SNCP group and PC group were up-regulated; in contrast, the mRNA levels of TGF-βRII, COL1A1 and α-SMA were down-regulated, with a statistically significant difference (p < 0.01), but there was no statistically significant difference between the SNCP group and PC group (p > 0.05). 500 nm; focal lateral fusion and unclear outline (red arrow) are indicated. (d) Diameter distribution of collagen fibers from wounds at 21 days post-wounding, n ≥ 1200.

SNCP Blockades the TGF-β/Smad Signaling Pathway to Achieve Scarless Healing
To investigate the effect of SNCP on TGF-β/Smad signaling pathway, the mRNA expression of transforming growth factor-β1 (TGF-β1), transforming growth factor-β receptor II (TGF-βRII), α-smooth muscle actin (α-SMA) and Smad7 in the wound at 28 days post-wounding, were investigated by qRT-PCR. As shown in Figure 7, compared with the NC group, the mRNA levels of Smad7 in both the SNCP group and PC group were up-regulated; in contrast, the mRNA levels of TGF-βRII, COL1A1 and α-SMA were down-regulated, with a statistically significant difference (p < 0.01), but there was no statistically significant difference between the SNCP group and PC group (p > 0.05).

Discussion
Our previous studies had found that the Sipunculus nudus extract had been shown to have obvious activity of accelerating wound healing [6]. Sipuncula is rich in collagen [11,26], and collagen hydrolysate is considered to be an effective active substance for wound repair. To investigate functions of collagen peptides in Sipunculus nudus, the physical and chemical properties of collagen (SNC) and the effects of collagen peptides (SNCP) on promoting wound healing were studied. The results show that SNC was a typical type I collagen with a denaturation temperature of 32.8 °C (Figure 1), while SNCP possessed an outstanding capacity to induce HUVEC, HaCaT and HSF cells proliferation and migration (Figure 2). Through the use of the mouse skin full cortical wound model, we demonstrated that the SNCP had a potential activity of accelerating wound healing and preventing pathological scar formation (Figure 3), and its mechanism was related to regulation of inflammatory factors, collagen deposition, collagen fiber remodeling and blockade of the TGF-β/Smads signal pathway. Our findings provide a theoretical basis for the wound healing potential of SNCP.
Nutrition is an important part of wound healing, and maintaining good nutrition could be one of the effective strategies to promote a more rapid wound healing [3]. Derived from collagen of Sipunculus nudus, SNCP was consisted of collagen peptides and free amino acids, providing a nutrient matrix for wound healing. Collagen peptides exert a variety of biological activities. In the skin, collagen peptides act as false collagen degradation peptides, which send a false signal in the fibroblast cells to synthesize new collagen fibers. Moreover, they possess chemotactic properties to promote cell migration and proliferation, which is an important process in wound healing [7,15,17]. In this study,

Discussion
Our previous studies had found that the Sipunculus nudus extract had been shown to have obvious activity of accelerating wound healing [6]. Sipuncula is rich in collagen [11,26], and collagen hydrolysate is considered to be an effective active substance for wound repair. To investigate functions of collagen peptides in Sipunculus nudus, the physical and chemical properties of collagen (SNC) and the effects of collagen peptides (SNCP) on promoting wound healing were studied. The results show that SNC was a typical type I collagen with a denaturation temperature of 32.8 • C (Figure 1), while SNCP possessed an outstanding capacity to induce HUVEC, HaCaT and HSF cells proliferation and migration ( Figure 2). Through the use of the mouse skin full cortical wound model, we demonstrated that the SNCP had a potential activity of accelerating wound healing and preventing pathological scar formation (Figure 3), and its mechanism was related to regulation of inflammatory factors, collagen deposition, collagen fiber remodeling and blockade of the TGF-β/Smads signal pathway. Our findings provide a theoretical basis for the wound healing potential of SNCP.
Nutrition is an important part of wound healing, and maintaining good nutrition could be one of the effective strategies to promote a more rapid wound healing [3]. Derived from collagen of Sipunculus nudus, SNCP was consisted of collagen peptides and free amino acids, providing a nutrient matrix for wound healing. Collagen peptides exert a variety of biological activities. In the skin, collagen peptides act as false collagen degradation peptides, which send a false signal in the fibroblast cells to synthesize new collagen fibers. Moreover, they possess chemotactic properties to promote cell migration and proliferation, which is an important process in wound healing [7,15,17]. In this study, SNCP was mixed collagen peptides with different molecular weights, containing a large number of low molecular weight peptides, of which components of less than 5 kDa accounted for approximately 81.95%. It has been proved that absorption of collagen peptide was improved by smaller size [3]. Therefore, the advantage of wound healing process by applying SNCP may also be due to its small peptides, which facilitate body absorption. Moreover, essential amino acids provide excellent nutritional support for wound healing [3]. L-arginine has been shown to enhance wound strength and collagen deposition in rodents and humans, as a supplementation accelerating the diabetic wound healing [19,39]. The SNCP is rich in Arg with 0.81 g/100 g (accounting for 19.36% of the total) (Table 1), which might contribute to the potential wound healing effect. However, the precise functional components of SNCP remained unclear. In the future, the separation and functional evaluation of collagen peptides with different MW size are needed to confirm the active components in SNCP.
The wound healing process takes place in distinct phases that at some times may overlap, proceeding by a preliminary hemostatic phase, following by the inflammatory phase, the proliferating phase and the maturation one [3,4]. In the inflammatory period, inflammation is necessary for resisting the invading environmental pollutants and removing tissue debris. However, excessive inflammation could cause damage to the normal tissues, which was the main cause of disease related to fibrosis [40]. It had been reported that both fetal trauma and oral mucosal trauma showed relatively low inflammation, and both of them could quickly seal the wound without scars [41,42]. Many experimental models of wound repair showed that excessive inflammation could delay healing and cause scar formation [43]. IL-1β and TNF-α are important inflammatory cytokines and indicators of inflammatory response. In this study, SNCP could significantly decrease the level of IL-1β and TNF-α in the wound surface at 7 days post-wounding, which indicates SNCP processed activity of reducing inflammation (Figure 4c). In addition, IL-1β and TNF-α could reduce collagen synthesis in a dose-dependent manner, among which TNF-α strongly promoted matrix metalloproteinase (MMP) secretion, degraded collagen and blocked the mRNA transcription of type I/III pro-collagen [44]. This might also be a cause of lower mRNA levels of COL1A1 in the NC group at 14 days (Figure 6b). TGF-β1 was an important driver of fibrosis process, which was related to the formation of abnormal scars [45]. In fetal wounds, the addition of exogenous TGF-β1 resulted in the same scarring effect as adults [46,47]. At 7 days post-wounding, the level of TGF-β1 mRNA in SNCP group was lower than that in NC group, which indicate that SNCP might have a better anti-scar activity. HE-stained sections confirm these views (Figures 4a and 5a). The epidermis of SNCP group was thin and flat, while the epidermis of NC group was fat, thickened and uneven, with a tongue-like edge [48].
In the proliferating phase, the structural and functional reconstruction of differentiated muscle fibroblasts in granulation perform repair processes. Scarless healing is largely dependent on the reconstruction of normal ECM. During this period, the mature collagen was remodeled to form an ordered basket weaving arrangement, with a structure and appearance similar to normal skin [49]. As shown in Figure 6a, mature granulation tissues, such as sebaceous glands, blood vessels and hair follicles, were observed in the dermal layer of the SNCP group. However, lack of normal healthy skin appendages was observed in the NC group and the collagen gathered in parallel to form a bundle, which was excessively accumulated and easily formed scars. In addition, the SNCP group could regulate the formation of collagen. The expression of COL1A1 mRNA was up-regulated at 14 days (Figure 6b), which indicates that SNCP could promote the synthesis of a large amount of collagen to form ECM during the proliferation period. However, during the remodeling period, COL1A1 mRNA maintains a relatively low level, which allowed the dermis to maintain a dynamic balance between collagen synthesis and decomposition. The tissue degraded excessive ECM including abnormal collagen while synthesizing new mature collagen to reshape ECM, preventing too much immature collagen from accumulating into large linear bundles to form scars or keloids [50]. In this study, the expression of COL1A1 mRNA in SNCP groups was down-regulation at 21 days post-wounding (Figure 6b). In addition, TEM analysis also confirmed that treatment with SNCP for 21 days show a similar appearance to normal skin, with relatively round and concentrated diameter distribution collagen fibers and clear contours (Figure 6c). TGF-β/Smad signal transduction pathway is an important way to affect post-traumatic scar formation and extracellular matrix deposition [51]. In the TGF-β/Smad pathway, TGF-β1 acts as a growth factorial signaling molecule in wound healing and scar formation. When TGF-β1 binds to TGF-β type I and type II receptors (TβRI and TβRII, respectively) on the cell membrane to form the active signaling complex, the TGF-β mediated signal transduction is activated [51], which results in the persistent presence of high levels of α-SMA, causing excessive collagen deposition (COL1A1) and inducing an abnormal fibroblast phenotype, a unique sign of pathological scars [52]. Smad7, a negative feedbackregulated inhibitory protein, could inhibit Smad2/Smad3 phosphorylation and block the TGF-β signaling pathway by strongly binding to TGF-βRI to prevent pathological effects caused by TGF-β and inhibit scar hyperplasia [53][54][55]. In our study, compared with the NC group, the expression levels of TGF-β1 and TGF-βRII in the SNCP group were significantly down-regulated (p < 0.01), while the expression levels of Smad7 were significantly up-regulated (p < 0.01) (Figure 7), which inhibited and blocked the TGF-β/Smads signal transduction pathway. Those might be the reasons why SNCP could regulate collagen formation and remodeling, thereby preventing pathological effects caused by TGF-β and inhibiting scar hyperplasia.

Extraction of the Collagen and Preparation of Collagen Peptides
To extract Sipunculus nudus collagen (SNC), all processes were conducted at 4 • C. The coelomic walls were fully cut and pasted, then soaked in 0.2 M NaOH for 24 h at a ratio of 1:10 (w/v) to remove non-collagenous proteins. The mixture was centrifuged at 10,000×g for 20 min. After removing the supernatant, the precipitation was washed repeatedly with distilled water, then soaked in 0.5 M acetic acid solution with 0.3% pepsin and homogenized at a ratio of 1:10 (w/v). After incubating for 48 h, the suspension was centrifuged at 10,000×g for 20 min (Heraeus Fresco Germany), and the supernatant was salted out by the addition of NaCl to a final concentration of 1 M. The mixture was allowed to stand overnight (at 4 • C), and the precipitates were collected (centrifugation, 12,000×g, 30 min), dissolved in 0.5 M acetic acid (10%), dialyzed (MD 34, USA) for 2 d and finally lyophilized (Labconco Freeze Dryer FreeZone 6 Liter, CA, USA). SNC was hydrolyzed by animal hydrolytic protease (3000 U/g) and flavor protease (3000 U/g) under the condition of pH 7.0, a solid/solution ratio of 1:300 (w/v) and at 55 • C for 5 h, and the mixture was heated to 100 • C for inactivation, followed by centrifugation. After lyophilization, Sipunculus nudus collagen peptide (SNCP) was obtained.

Sodium Dodecyl Sulphate Polyacrylamide-Gel Electrophoresis (SDS-PAGE)
One µL sample loading buffer (5×) (Beyotime Institute of Biotechnology, Jiangsu, China) was dissolved in 4 µL the SNC sample. The mixture was heated to 100 • C for 5 min, followed by centrifuging at 9000×g for 5 min. Subsequently, 5 µL supernatant was loaded onto an instant polyacrylamide-gel made of 10% running gel and 6% stacking gel. After electrophoresis, the gels were stained BeyoBlue™ Coomassie Blue Super Fast Staining Solution and destained with deionized water. Quantitative analysis of protein band intensity was performed using a Model GS-700 Imaging Densitometer (Bio-Red Laboratories, Hercules, CA, USA) with Molecular Analyst Software version 1.4 (image analysis systems).

Fourier Transform Infrared Spectroscopy (FTIR)
The chemical structure of SNC was studied by FTIR spectroscopy (Bruker Tensor 27 Instruments, Billerica, MA, Germany). The samples were prepared in agate mortar containing 3 mg collagen in approximately 150 mg potassium bromide (KBr). The spectra were obtained with wave numbers ranging from 4000 cm −1 to 400 cm −1 at a resolution of 2 cm −1 over a total of 32 scans. The background was subtracted by the Opus software (Bruker Instruments, Billerica, MA).

Differential Scanning Calorimetry (DSC)
The SNC (5 mg) was weighed accurately (0.1 mg) into aluminum pans and sealed. The samples were heated from 24 to 100 • C at a scanning rate of 1 K/min, with an empty sealed pan as a reference. The transition temperature (Tm) was recorded.

Determination of Molecular Weight Distribution of SNCP
The molecular weight distribution of the SNCP was determined by high performance liquid chromatography (Angilent 1200, Palo Alto, CA, USA) following the methods of other researchers [15]. The mixed standard samples, which consisted of cyyochrome (12,500 Da), aprotinin (6500 Da), bacitacin (1450 Da), ethyl amino acid-ethyl amino acidtyrosine-arginine (451 Da) and ethyl amino acid-ethyl amino acid-ethyl amino acid (189 Da), were loaded into the column. Acetonitrile/water (45:55) with 0.1% trifluoroacetic acid was adopted as mobile phase, the flow rate was 0.5 mL/min and the UV wavelength was 220 nm. A standard curve of retention time absorbance was plotted. The SNCP solution was filtered through a 0.45 µm filter and subjected in the same conditions. Then, the molecular weight was calculated according to the retention time using the curve equation.

Amino Acid Composition Measurement of SNCP
The amino acid of the SNCP was determined according to the previous study [56]. SNCP was hydrolyzed by dissolving it in 6 M HCl at 110 • C for 24 h. The filtrate and standard samples of amino acids solution was analyzed with an automatic amino acid analyzer (S-433D, Sykam, Bremen, Germany).

In Vitro Scratch Assay
The wound healing assay in vitro was performed by two-well wound-generating insert (Ibidi). After adding the wound-generating insert into a 24-well plate, 50 µL cell suspensions (5 × 10 5 cells/mL HaCaT, 5 × 10 5 cells/mL HSF, 1 × 10 5 cells/mL HUVEC) were seeded per well of the insert with Dulbecco's modified Eagle's medium (DMEM; Hyclone), containing 10% (v/v) fetal bovine serum (FBS; Hyclone), 1% (v/v) penicillin and streptomycin (Hyclone), respectively. After incubation overnight to form a monolayer, the inserts were removed, after which the debris was washed with PBS twice. The monolayer cells were incubated in complete medium containing 10% PBS (v/v) as the negative control (NC) and SNCP with a concentration of 500 µg/mL. The image was observed and analyzed by using an inverted fluorescence microscope (DMI4 000, Leica, Germany) and the wound area was calculated with the Image J software (National Institute of Heath, Bethesda, USA). The wound closure percentage was obtained by the following formula: where A 0 was the wound area at 0 h and A t was the remaining area at the designated time.

In Vivo Wound Healing Assay
All animal programs in this study were approved by the Animal Ethics Committee of Guangdong Ocean University, and were conducted in accordance with the national regulations (the approval number assigned by the ethical committee for our animal experiment was GDOU-LAE-2019-0003). Kunming (KM) male mice aged 7 to 8 weeks were purchased from Sibefu Biotechnology Co., Ltd. (Beijing, China). The mice were anesthetized, and the area on the back was shaved and cleaned with ethanol before the surgery. Full-thickness excisional round holes with diameter of 8 mm were made on the exposed skin of the back using disposable biopsy punch (Rapid-Core, EMS, America). The group of mice with none-treatment wound was considered as negative control group (NC). Each 2 g SNCP (based on total contents of nitrogen) was added to 1 mL of sterile water to form an ointment for application to the wound in the SNCP group (SNCP), and the Yunnanbaiyao powder was applied to the wound in the positive control group (PC). In both cases were applied daily. Mice were placed in a single cage at a temperature of 25 • C and given enough food and water.

Macroscopic Evaluation of Wounded Tissues
For measurement of wound closure percentage, wounds were photographed every 2 days post-wounding, and the wound closure percentage was obtained by the following formula: Wound closure (%) = (S 0 − S t )/S 0 × 100, where S 0 was the wound area at 0 day and S t was the remaining area at the designated time.

Microscopic Evaluation of HE-Stained Sections
For histologic analysis, the surrounding skins of the wound were fixed in 4% formalin, embedded into paraffin, for the hematoxylin and eosin-(H&E; Bogu Biotechnology Co., Ltd. Shanghai, China) stained sections at day 7 post-wounding to measure the width of the wound surface and the length of the epithelial growth, for the H&E-stained sections at day 14 and 28 post-wounding to analysis the epithelial thickness index (ETI) and the thin/thick ratio of epithelial thickness and for the Masson's trichrome-stained sections at day 14, 21 and 28 post-wounding to observe the synthesis and deposition of collagen, angiogenesis and the regeneration of hair follicle and sweat gland in dermal. The ETI and the thin/thick ratio of epithelial thickness were obtained by the following formula: Epithelial thickness index (ETI) = δ 1 /δ 2 , Thin/thick ratio of epithelial thickness = δ 3 /δ 4 , where δ 1 , δ 2 , δ 3 and δ 4 refer to the average new-forming epidermal height, the average normal epidermal height, the new-forming thinner epithelial thickness and the newforming thicker epithelial thickness, respectively.

Quantitative Real-Time Polymerase Chain Reaction (qRT-PCR) Analysis
Total RNAs were extracted from the skins of wound using RNAIso Plus (Takara). The cDNAs were synthesized using Prime ScriptTM RT reagent Kit with gDNA Eraser (Takara) in accordance with the manufacturer's instructions. The qRT-PCR reaction conditions were 95 • C for 30 s, followed by 40 cycles of 95 • C for 5 s and 60 • C for 30 s. Expression levels of target genes were calculated using the comparative CT method with glyceraldehyde-3-phosphate-dehydrogenase (GAPDH) as an endogenous reference gene. To confirm amplification of specific transcripts, melting curve profiles were produced at the end of each reaction. The following primer sets were used ( Table 2). Table 2. Primers used for the qRT-PCR analysis.

TEM Analysis
The wounded tissues were fixed at 4 • C in 4% paraformaldehyde for 30 min, then in 2.5% glutaraldehyde for overnight, and subsequently in 1% osmium tetroxide for 2-3 h. Then, after washing, dehydrating, embedding and solidifying, transmission electron microscopy (TEM, Tecnai G2 20 TWIN; FEI) analyses were operated at an accelerating voltage of 200 kV and a magnification of 5000. Image J software (National Institute of Heath) was employed to analyze the diameter of each fibril in each image.

Statistics Analysis
All experimental values were presented as the means ± the standard deviations (SD). The comparison analysis between the groups was assessed via one-way analysis of variance (ANOVA) with the The John's Macintosh Product (Version 10, SAS, Cary, NC, USA). Statistical significance among groups was designated as follows: statistical significances of p < 0.01 and p < 0.05 between groups were presented via different capital letters and lowercase letters, respectively, while p > 0.05 was presented via the same lowercase letters among groups.

Conclusions
In this study, we found that SNCP could accelerate the wound healing and inhibit scar formation in a skin wound of mice, primarily by facilitating rapid proliferation of epithelial cells, reducing inflammation, improving collagen deposition and recombination and blockade of the TGF-β/Smads signal pathway. Therefore, SNCP may have promising clinical applications in skin wound repair and scar inhibition. However, further molecular mechanisms remain to be elucidated.