Silk is one of the oldest fibers known to man and has been widely used in the textile industry. Recently, a variety of advantages for silk derived products have emerged. Silk protein has excellent mechanical properties, biocompatibility and biodegradability (Kundu et al., 2013). Due to these properties, various studies have been conducted on silk for biomedical and industrial applications (Okabayashi et al., 2009; Wang et al., 2008, 2010).

The silk fiber spinning process has the advantage of using water as a solvent at room temperature and atmospheric pressure (Jin and Kaplan, 2003). Silk fiber derived from the cocoon of Bombyx mori is composed of two proteins named fibroin and sericin. Sericin completely covers two strands of fibroin to form silk fibers. So far, the role of sericin during spinning process of silk fiber is unclear. Lee et.al proposed that sericin has a retardation effect on the crystallization of fibroin in the phase separation of the two proteins (Lee, 2004). In order to fully characterize the silk spinning process, the interactions between fibroin and sericin must be identified. However, it is difficult to observe the interface between the two proteins because both fibroin and sericin have transparent property in aqueous solutions.

A variety of dyes and dyeing processes have been developed for the use of silk in the textile industry (Tsukada et al., 1996). Recently, a large amount of research has focused on developing silk fibers with a desired color by introducing food dye to the silkworm breeding process (Tansil et al., 2011, 2012). Production of colored silk using this process is eco-friendly and efficient compared to conventional dyeing processes. However, a disadvantage of this type of dyeing process is that the concentration of dye transmitted to fibroin and sericin can be different depending on the dye. When a feeding process was employed using artificial feed laced with direct acid 80, almost all of the dye was transmitted to the fibroin layer. However, in the case of acid orange 80, most of the dye was transmitted to sericin layer (Nisal et al., 2014).

In this study, we prepared aqueous fibroin and sericin solutions laced with various dyes. We then observed the gelling behavior of fibroin that had been phase separated from sericin. Finally, an optimal dyeing system was determined for monitoring the interface of fibroin and sericin.

Bombyx mori silkworm cocoons were kindly provided by National Academy of Agricultural Science. All chemicals were purchased from Sigma-Aldrich (USA).

The silk cocoons were degummed twice with 0.2 % (w/v) sodium carbonate and 0.3 % (w/v) Marseille soap solution at 100°C for 30 min, and then rinsed with distilled water to remove any sericin. Following this, the fibroin was dried in an oven at 50°C.

Sericin was extracted by boiling 20 g of Bombyx mori cocoons in 500 mL of distilled water using an autoclave for 1 h at 120°C. The extractant was then filtered with a nonwoven filter (Miracloth, typical pore size: 22 ~ 25 μm, EMD Millipore, USA) in order to remove the remaining cocoons. After the filtering process, a 1 % (w/v) sericin solution was obtained.

The previously extracted fibroin was dissolved (10%, w/v) in a 9.3 M LiBr solution at 50°C for 4 h. The solution was then dialyzed in distilled water with a cellulose membrane (molecular weight cut off = 6000 – 8000, Spectrum Laboratories, Inc.) for two days. The final fibroin concentration was adjusted 5 % (w/v) with distilled water. To prepare the dyed aqueous fibroin and sericin solutions, 100 ppm of each dye was added to each solution.

The gelation experiments of fibroin were carried out as follows. First, 3 mL of aqueous fibroin was added to a 5 mL vial using a pipette. Then, 1.5 mL of aqueous sericin was added to the above fibroin solution. In order to maintain the interface between the two solutions, sericin was added carefully using a syringe pump (KD Scientific, USA). The experiments were carried out in a situation supporting a dye in either the fibroin or sericin solutions.

During the silk fiber spinning process, sericin envelopes two strands of fibroin. However, the role of sericin in covering fibroin is still unclear. Oh et al. reported that during the gelation of fibroin under phase separated conditions with sericin, sericin retards the gelation time of fibroin. Additionally, the interface between fibroin and sericin was maintained until gelation (Oh et al., 2009). However, because both fibroin and sericin are transparent in aqueous solutions prior to gelation, monitoring the interface between the two solutions is difficult. Table 1 shows the name, molecular weight, and chemical structures of the five dyes used in the experiments. All of the compounds are azo dyes except for cationic rhodamine B. When staining or labeling materials with fluorescence compounds or dyes, it is very important that there is no interaction between the target material and the dyes (Joosen et al., 2014). We found that there was no change in the gelation time of the fibroin and sericin solutions with the addition of the dyes (data not shown). We performed phase separated gelation experiments using aqueous fibroin solutions (5%, w/v) with a dye concentration of 100 ppm and sericin solutions (1 %, w/v). Fig. 1 shows the phase separation between the fibroin and sericin solutions before and after gelation. As shown in Fig. 1, before gelation, all of the fibroin solutions with various dyes maintained phase separation with the sericin solutions. Following the gelation of fibroin, it was confirmed that all of the color separation was maintained except for orange II sodium salt. Fig. 2 shows the relative height of the dyed sericin solution after gelation compared to its initial height. The relative height increase for orange II sodium salt was 300 %, suggesting that the dye diffused well into the sericin layer. This phenomenon is believed to be due to the molecular weight of dye. The molecular weight of orange II sodium salt is 350.32 g mol^-1, which is the lowest of the five dyes. Generally, the dyeing rate is proportional to the molecular weight of the dye. Because of the low molecular weight of orange II sodium salt, it could easily diffuse into the sericin layer. Comparing the remaining dyes, the relative height value for rhodamine B (159.1 %) was slightly higher than that of direct red 80 (118.2 %), Congo red (127.3 %) and mordant black 11 (130.3 %). This could be explained by the different dyeing mechanism of rhodamine B compared with the other dyes. Rhodamine B is a basic dye with a cationic group (Bhatnagar and Jain, 2005). This cationic group is attracted to and retained by the anionic sites of the fibers. The pH of fibroin and sericin solutions in the gelation experiments was 6.5 and 7.5, respectively which were higher than the isoelectric point of the two proteins. It means that aqueous fibroin and sericin solutions have negative net proton charge. This suggests that the abundance of negatively charged groups in fibroin and sericin could be easily attracted to rhodamine B. Therefore, the maintenance of phase separation between the two proteins with of the addition of rhodamine B is difficult.

To efficiently observe the interface between fibroin and sericin, we performed similar gelation experiments using dyed sericin aqueous solutions. The phase separation between the fibroin and sericin solutions before and after gelation and the relative height of the dyed solution after gelation compared to its initial height are shown in Figs. 3 and 4. Similar to the above experiment, when using direct red 80, Congo red and mordant black 11 dye, phase separation between sericin and fibroin solutions was well maintained. In the case of rhodamine B and orange II sodium salt, diffusion of dye from the sericin to the fibroin layer was observed. This could also be explained by the difference in molecular weights of the dyes and the cationic structure of rhodamine B. The interface between the two proteins is easier to observe when the dye is added to the sericin solution. The addition of dyes to the sericin layer did not show a large difference between the original and final location of interface. When the dyes were added to fibroin, movement of the dye into the sericin layer is relatively easy since sericin does not gel at 50°C. Conversely, when dyes were added to sericin, fibroin aqueous solution generates a three-dimensional polymer network during the gelation. This gel structure is to interfere with the penetration of the dye and facilitates the observation of the interface between fibroin and sericin.

In summary, the interface between fibroin and sericin could be monitored using azo dyes with a molecular weight high enough to prevent diffusion between the protein layers. It was also confirmed that when dyed sericin solutions were used, a more vivid interface was observed. This technique for monitoring the interface between fibroin and sericin using dyes could be helpful in understanding the interactions between these silk proteins.