Species identification of Bombyx mori and Antheraea pernyi silk via immunology and proteomics.

In recent years, increasing attention has been paid to the origin, transmission and communication of silk. However, this is still an unsolved mystery in archaeology. The identification of silk-producing species, especially silk produced by Bombyx mori (B. mori) and Antheraea pernyi (A. pernyi), is of key significance to address this challenge. In this study, two innovative methods, i.e. immunology and proteomics, were proposed and successfully established for the species identification of silks. ELISAs result demonstrated that the two prepared antibodies exhibited high sensitivity and specificity in distinguishing B. mori and A. pernyi silk. No cross-reactivity with each other was observed. Moreover, biomarkers were obtained for Bombyx and Antheraea through proteomic analysis. It was also confirmed that the biomarkers were suitable for identifying the species that produced a given silk sample. Compared with conventional methods for distinguishing silk species, immunological and proteomics techniques used in tandem can provide intact information and have the potential to provide accurate and reliable information for species identification.

Introduction

Silk, a symbol of ancient Chinese culture, is the best-known textile material and was spread around the world along the Silk Road[1]. The ancestors of the Chinese spun raw silk as early as the Neolithic Age. Silk was derived from wild silkmoth species, rather than Bombyx mori (B. mori), until the wild silkworms were domesticated. Among wild silkworms, the cocoon of Antheraea pernyi (A. pernyi) is the most widely used and described. It is generally believed that silk produced by B. mori was transported from China to Western Asia and the Mediterranean after the opening of Silk Road in the Western Han Dynasty (206 B.C.-8 A.D.). However, the first example of wild Antheraea silk in the Indus civilization has been dated to 2450–2000 BC[2]. These findings have encouraged people to reconsider the origin and transmission of ancient silks. Thus, the identification of ancient traces of silk produced by A. pernyi and B. mori has become of great significance in the archaeological field.

A single silk glued with sericins consists of 70–80% fibroins[3], whose amino acid sequence reflects the diverse morphology and structure of silk[4,5]. Fibroin derived from different silk also imparts the specific desirable performance properties of silk, such as softness, lightness, and smoothness[6]. Thus, studying the fibroin differences between B. mori and A. pernyi silk is a basis for identifying their species. In the past several decades, various studies regarding silk characterization have been carried out using a variety of methods, such as scanning electron microscopy (SEM)[7], Fourier transform infrared spectroscopy (FTIR)[8-10] and amino acid analysis[11]. These methods can distinguish well-preserved silk from other fibers and identify what species produced the silk. However, silk is easily affected by light, radiation and temperature, which is particularly problematic for very old silk samples[12]. Thus, silk unearthed in graves usually suffers from physical and chemical changes, showing serious degradation and aging. Consequently, despite lots of meaningful information can be achieved by conventional methods, the efficacy is likely to be limited in the detection of ancient samples with contamination or severe degradation. Therefore, novel approaches for identifying the species that produced a given sample of silk should be developed.

Among the new rising technologies, immunological techniques and proteomics are preferred methods. More accurate information about ancient silks at the sequence level can be obtained using immunological techniques and proteomics. Both the two methods are particularly suitable for the identification of poorly preserved ancient silks (severely degraded or contaminated) or even the silk traces, as long as partial amino acid sequences are still present. This is also the key challenge facing researchers in archaeological sites.

Immunological techniques, including various efficient methods such as immunoprecipitation (IP), enzyme-linked immunosorbent assays (ELISAs), immuno-fluorescence microscopy and immunochromatography, present the advantages of high sensitivity, short processing times and low cost[13]. Therefore, immunological techniques have attracted increasing attention from professionals around the world. ELISA is the most commonly used of these methods and is widely applied in many fields[14-19]. In ELISA test, the antigen interacts with the primary antibody, and the antigen-antibody conjugates then combine with the secondary antibody. Thereafter, the reaction can be detected using a microplate reader. The preparation of the primary antibody is the most important step for immunoassays. The more specific the primary antibody, the more reliable the ELISA result. To design specific antibodies, the amino acid sequences of B. mori silk and A. pernyi silk were compared and the diagnostic sequences were selected for peptide synthesis. Immunospecific primary antibodies were produced by injecting rabbits with the synthesized peptide coupled with a carrier protein. Thus, it was feasible to use ELISA to identify the silk produced by B. mori or A. pernyi.

Proteomics is a sensitive, precise and high-throughput tool for understanding complex mixtures at the protein level[20,21]. This approach has recently been broadly adopted in many fields[22,23]. In particular, researches regarding proteomics applying in archaeology have been reported sporadically. For example, Mai et al. analyzed red cosmetic sticks in early Bronze Age at Xiaohe Cemetery using proteomics[24]. Shevchenko et al. identified the composition and manufacturing recipe of the 2500-year old sourdough bread from Subeixi cemetery in China via proteomics[25]. Solazzo et al. also used proteomic technology to identify protein remains in archaeological potsherds[26]. The results of these studies indicated that proteomic analysis is suitable not only for biology but also for archaeology. Proteomic technique mainly consists of two analysis strategies: two-dimensional gel electrophoresis and mass spectrometry. In-gel digestion is a classic approach for MS analysis, which has frequently been used to detect proteins in these years[27-29]. In addition, a correlative method based on gel-free digestion of protein followed by liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS) analysis can now be used to achieve rapid development[30,31].

In the present study, we differentiate silk produced by B. mori and A. pernyi based on morphology and primary structure by combining traditional methods with immunological techniques and proteomics (Fig. 1). First, the silk was characterized via SEM and FTIR. Then, the amino acid composition of the silk was determined through amino acid analysis. Thereafter, an ELISA system was established to detect the corresponding fibroin in silk using diagnostic antibodies. To further distinguish B. mori silk from A. pernyi silk at the sequence level, an LC-MS/MS approach was utilized to elucidate the differences in their amino acid sequences and identify biomarkers for the silks. As we expected, the results revealed that the ELISA and proteomics approaches that we employed can provide accurate and reliable information. These findings probably have great potential for future applications.

Figure 1

Experimental design for identifying species of silks produced by B. mori and A. pernyi via immunology and proteomics.

Materials and Methods

Reagents and materials

HPR conjugated goat anti-Rabbit IgG (H + L) (100 µg at 2 mg/mL) and the TMB color system were purchased from Hangzhou Hua’an Biotechnology Co., Ltd. Bovine serum albumin (BSA), sodium deoxycholate and ammonium bicarbonate were purchased from Sigma-Aldrich. Copper hydroxide was supplied by Shanghai Macklin Biochemical Co., Ltd. Ethanediamine, diethyl ether, methanol, acetone and hydrochloric acid were supplied by Tianjin Gaojing Fine Chemical Co., Ltd. Dithiothreitol (DTT), iodoacetamide (IAA), trifluoroacetic acid (TFA), formic acid (FA), sodium deoxycholate (SDC) and acetonitrile (ACN) were purchased from Thermo Fisher Scientific. The water used in all experiments was purified using a TPM Ulrapure water system. The silk cocoons (B. mori and A. pernyi) and silk fabrics (Samples I and Sample II) were kindly provided by the China National Silk Museum.

Morphology

Images of B. mori silk cocoon, A. pernyi silk cocoon, and silk samples were obtained using a digital camera (Canon EOS700D). The cross-sectional morphology of the silk fibers was characterized with a polarizing microscope (AxioCam MRc 5). Samples were cut using a fiber slicer. The longitudinal surfaces of the samples were characterized via SEM (JEOL JSM-5610). The samples were adhered to conductive adhesives, sputter-coated with gold for 60 s and then examined via SEM at an extra-high tension of 1 kv.

Chemical structure

Attenuated total reflection-Fourier transform infrared (ATR-FTIR) spectroscopy analysis (Thermo Scientific) was employed to characterize the main chemical bond of the silk in the wavenumber range of 500–4000 cm−1. Origin 9.0 software was applied to process the data.

Amino acid analysis

The amino acid composition of the samples was characterized using an amino acid analyzer (AAA, Waters 2695). Samples were hydrolyzed with 6 M hydrochloric acid for 24 h at 110 °C to yield individual α-amino acids. After drying the hydrated solution under nitrogen, the hydrolyzate and the internal standard were dissolved in derivation liquid. The mass fraction and the molar fraction of the amino acids were calculated by AAA.

Extraction of silk fibroin

Considering that A. pernyi silk is very difficult to dissolve in traditional extraction solutions (calcium chloride aqueous solution or calcium chloride-water-ethanol solution), a new approach was adopted for extracting silk fibroin. The detailed procedures performed were as follows.

After peeling the husks from the cocoons, the cocoons were incubated in a 0.5% (w/w) Na2CO3 aqueous solution at 98 °C for 1 h at a bath ratio of 1:100. After repeating this process 4 times, the degummed cocoon layer was soaked in 1% (w/w) hydrochloric acid at 25 °C for 1 h, then washed with water and dried at 80 °C. Next, the desiccated silkworm layer was soaked in 70% (w/w) methanol at 60 °C for 2 h and then placed in absolute ethanol at 60 °C for 1 h. The sample was subsequently soaked in acetone at 40 °C for 1 h, followed by immersion in ether at 25 °C for 1 h. After drying, the silk was soaked in a 100-fold volume of distilled water at 98 °C for 1 h, which was repeated 4 times until no biuret reaction occurred[32].

Next, a solution was produced by dissolving 6.6 g of copper hydroxide in 50 mL of distilled water, to which 8.5 mL of ethylenediamine was added, and the mixture was stirred well to dissolve all solids before being brought to 100 mL distilled water. The silk sample was subsequently placed in the prepared copper ethylenediamine solution (1:50) and held at 60 °C for 1 h. After cooling to 25 °C, the pH of the solution was adjusted to 8.5 with 10% acetic acid on a magnetic stirrer. Then, the solution was purified using a dialysis membrane with a molecular weight cut off of 8 kDa. After dialyzing for 3 days, the resultant solutions were lyophilized to obtain silk fibroin powders of B. mori silk and A. pernyi silk, respectively (abbreviated as SF-BM and SF-AP).

0.1 g of silk fabrics were added to the extracting solution (copper ethylenediamine solution) at the bath of 1:50 and held at 60 °C for 1 h. The pH of the mixture was adjusted to 8.5 with 10% acetic acid on a magnetic stirrer and then centrifuged for 10 min at 4500 rpm. The supernatant was obtained for further identification.

Preparation of primary antibodies

The immunogens used to generate specific primary antibodies were well-designed and carefully prepared. First, the diagnostic sequence (hapten) was selected by comparing the preliminary sequence differences of silk fibroin derived from B. mori silk and A. pernyi silk. (Fragmented amino acid sequences of silks produced by different species were registered in NCBI public databases (http://www.ncbi.nlm.nih.gov/pubmed/)). The MQRKNKNHGILGKC sequence was selected for B. mori silk[33], while the CSHSHSYEASRISVH sequence was selected for A. pernyi silk. The selected peptides were synthesized using a peptide synthesizer (CS Bio). After coupling the synthetic peptide with keyhole limpet hemocyanin (Sangon, Shanghai), the immunogen (complete antigen) was used to induce specific primary antibodies via animal immunization. The process of animal immunization is described in detail in a previous report by our group[34]. The resulting anti-SF-BM and anti-SF-AP antibodies were further purified using a Protein A column and stored at −20 °C for use (anti-SF-BM antibody: 0.58 mg/mL; anti-SF-AP antibody: 0.33 mg/mL).

Indirect ELISA procedures

Samples were dissolved in carbonate buffer (CB) at pH 9.6, and the concentration of the silk fibroin solution was then adjusted to 10 µg/mL. In this indirect ELISA test, 96-well microplates were used and each column of the microplate was removable. In the microplates, the antigen was combined with the primary antibody, which was targeted by an enzyme-conjugated secondary antibody instead of being conjugated with the enzyme directly. The following section presents a detailed description of this procedure.

First, 100 µL of the pretreated silk fibroin solution (10 µg/ml) was added to each well, followed by incubation at 4 °C overnight until the antigen was attached to the plastic. Then, the solution was removed, and the plate was washed three times with 200 µL of PBS. Next, 200 µL of blocking buffer was added to each well, followed by incubation at 37 °C for 2 h to bind non-specific binding sites. The blocking buffer was removed, and the plate was washed again with 200 µL of PBS three times. Then, 100 µL of the primary antibody was added to each well, followed by incubation at 37 °C for 1 h. Thereafter, the solution was removed, and the plate was washed with 200 µL of PBS three times. Next, 100 µL of the secondary antibody (HPR-conjugated goat anti-rabbit IgG (H + L)) was added to the well, followed by incubation at 37 °C for 1 h. After removing the secondary antibody and washing five times with PBS, 100 µL of substrate solution (TMB color system) was added to each well in a dark environment, where the plate was left at room temperature for 10 min. Finally, 100 µL of 1 mol/L H2SO4 was added to end the color reaction, and sample absorbance was measured using a microplate reader (Bio-Rad model 550) at λ = 450 nm.

To ensure the accuracy of the experiment, a battery of controls were set up. Silk fibroin solution was employed as a positive control, and the antigen was replaced by CB (pH = 9.6) as a negative control to ensure that the antibody combined only with antigen. The cut-off value was defined as the mean OD450 of CB (pH = 9.6) plus three standard deviations (mean + 3 SD).

Sample preparation for proteomics

To identify the silk fibroins of B. mori and A. pernyi at the primary structural level, a proteomic method was applied. Silk fibroins were first disintegrated in 100 µL of lysate (4% SDC, 50 mM Tris-HCl pH = 8.0) in an Eppendorf tube. The fibroin concentration was measured using the Bradford method[35]. Then, 25 µg of fibroins was added to 50 µL of 50 mM NH4HCO3, followed by adding 5.5 µL of 100 mM DTT and incubating at 37 °C for 1 h. Then, the sulfhydryl functional groups of silk fibroins were alkylated with 6 µL of 500 mM IAA via incubation in the dark for 30 min. Thereafter, the samples were centrifuged in an ultrafiltration tube (10 kDa) for 15 min at 14,000 g and washed twice with 50 mM NH4HCO3 (containing 0.4% SDC). Next, after transferring to a new Eppendorf tube, the silk fibroins were digested in a mixture solution of 50 µL of 50 mM NH4HCO3 (containing 0.4% SDC) and 2 µL of 0.5 µg/µL trypsin for 14 h at 37 °C. The resulting tryptic peptides were recovered via ultrafiltration at 14,000 g for 15 min and then washed with 50 mM NH4HCO3. Next, 5% TFA was added to the samples, and they were centrifuged for 2 min at 14,000 g to remove SDC. Thereafter, the supernatant was desalted using PierceTM C18 spin columns. Finally, the samples were eluted twice with 50 µL of 70% acetonitrile (ACN) and 0.1% FA in an ultrafiltration tube and lyophilized.

Sample analysis and database searches

Tryptic peptides were separated in a Thermo Fisher Scientific EASY-nLC 1000 UPLC system using a C18 trap column (5 µm, 100 Å, 2 cm × 100 µm) and a C18 analytical column (2 µm, 100 Å, 15 cm × 50 µm). The peptides were eluted using a gradient of solvent A (0.1% FA in water solution) from 97% to 5% and a corresponding gradient of solvent B (0.1% FA in acetonitrile solution) from 3% to 95%, at a flow rate of 250 nL/min over 90 min. The UPLC system was directly interfaced with a Thermo Scientific Q-exactive mass spectrometer which equipped with a nanoelectrospray source. Mass spectrometry analysis was performed in the data-dependent acquisition mode. There was a single full-scan mass spectrum in the orbitrap (300–2000 m/z, 70000 resolution) followed by 20 data-dependent MS/MS scans at 27% normalized collision energy.

Raw MS files were analyzed using the Xcalibur 2.2 SPI system. The MS/MS spectra were searched against UniProt database (http://www.uniprot.org/uniprot/) using Proteome Discoverer software (Version PD1.4, Thermo Scientific, USA). Precursor ion masses between 350 Da and 8000 Da were searched, and the precursor mass tolerance was set to 10 ppm. The fragment mass tolerance was set to 0.02 Da. Peptide identification was filtered at 1% false discovery rate. The search included two modifications: oxidation and carbamidomethyl. The peptide length was set at 7–30 amino acids.

Results and Discussion

Figure 2 shows images of B. mori (left column) and A. pernyi (right column) cocoons. The differences between the silks from the two species regarding their morphology, cross-section and longitudinal surface can be observed by comparing the images.

Figure 2

Morphology of silks produced by B. mori (left column) and A. pernyi (right column). (A,B) Digital images of cocoons; (C,D) Cross-sections of degummed silk fibers; (E,F) SEM images of the longitudinal surface.

As shown in Fig. 2A,B, the cocoon of B. mori was oval, shiny and exhibited white color. The A. pernyi cocoon exhibited tawnier color and was more pointed at both ends. Compared with the B. mori cocoon, the cocoon of A. pernyi was more resistant to degumming. Cross-sections of B. mori silk and A. pernyi silk are displayed in Fig. 2C,D. The shape of B. mori silk was similar to an acute triangle, whereas A. pernyi silk was more rectangular, as verified by the SEM images shown in Fig. 2E,F. According to the magnified SEM images, the silk of B. mori was thinner and smoother than the silk of A. pernyi. In summary, B. mori silk and A. pernyi silk exhibit overt differences in morphology, which typically influences fiber mechanics.

ATR-FTIR analysis was used to identify the functional groups of the degumming silks[36]. The characteristic peaks of the silks are shown in Fig. 3. Both samples displayed a peak at 3288 cm−1, with the characteristic signal from N–H stretching vibration coming from the amide A and amide B zones. The peak at 2922 cm−1 was assigned to the –CH2– vibration region. The spectra also showed peaks corresponding to amide bonds: the peak approximately 1583–1670 cm−1 corresponded to the amide I bond; that at 1474–1583 cm−1 to the amide II bond; and that at 1192–1295 cm−1 to the amide III bond. Differences between B. mori silk and A. pernyi silk could be observed in the spectra: there was a peak that was present only in A. pernyi silk at 965 cm−1. This peak was attributed to the peptide structure alanine-alanine (Ala-Ala), which rarely existed in B. mori silk[37]. Thus, the characteristic peak could be used to identify the silk species. However, the results will become unreliable for contaminative silk because miscellaneous peaks in the spectra might lead to analysis errors.

Figure 3

FTIR spectra of silks produced by B. mori and A. pernyi.

Amino acid composition

The amino acid composition was analyzed, and the obtained mass fraction and molar fraction of the amino acids of the silk fibroins are shown in Table 1. The different silks exhibited different amino acid compositions. According to the results shown in Table 1, there were seventeen amino acids detected in the two silk fibroins. Consistent with the results of a previous study[38], histidine (His) was present mainly in A. pernyi silk and absent from B. mori. Similarly, B. mori silk lacked cysteine (Cys), which was present in A. pernyi silk. The mass fraction of aspartic acid (Asp) in A. pernyi silk was 8.44%, whereas in B. mori silk, the percentage was only 2.86%. Alanine (Ala), glycine (Gly), serine (Ser) and tyrosine (Tyr) were the main amino acids found in both B. mori and A. pernyi silk, with total mass fraction of 83.35% and 74.59%, respectively. The crystalline degree of silk fibroin was determined by the contents of Gly, Ala, Ser and threonine (Thr)[39]. The higher content of these amino acids in B. mori silk suggests that this silk has higher degree of crystallinity than A. pernyi silk.

Table 1

Comparison of the amino acid composition of B. mori and A. pernyi silk.

Amino acid composition	B. morisilk		A. pernyi silk
	wta/%	Xb/%	wta/%	Xb/%
Asp	2.86	2.05	8.44	6.35
Ser	13.36	12.12	12.34	11.76
Glu	2.47	1.6	2.17	1.48
Gly	34.24	43.48	20.69	27.61
His	0	0	2.46	1.59
Arg	1.29	0.7	5.65	3.25
Thr	1.35	1.08	1.63	1.37
Ala	25.23	27	31.57	35.87
Pro	0.81	0.67	0.75	0.65
Cys	0	0	0.17	0.14
Tyr	10.52	5.54	9.99	6.86
Val	3.63	2.95	1.1	0.94
Met	0.02	0.01	0.32	0.22
Lys	0.7	0.45	0.95	0.65
IIe	1.19	0.86	0.61	0.47
Leu	0.9	0.65	0.61	0.47
Phe	1.43	0.83	0.54	0.32

awt: The mass fraction.

bX: The molar fraction.

Species identification of silks via immunology

Optimum antibody dilution

It was necessary to establish optimal antibody dilutions to identify the silks. Figure 4 shows the indirect ELISA curves of the secondary antibody against the primary antibody when applied at different dilutions. As shown in Fig. 4A,B, the optical density (OD450) values decreased with an increasing dilution of the primary antibody. The anti-SF-BM antibody was diluted 1:500, 1:1000, 1:2000, 1:4000, 1:8000 and 1:16,000, while the anti-SF-AP antibody was diluted 1:1500, 1:3000, 1:6000, 1:12,000, 1:24,000 and 1:48,000 with the same blocking solution. Besides, the secondary antibody was diluted 1:500, 1:1000, 1:2000, 1:4000 and 1:8000 with blocking solution accordingly. To obtain a high signal-to-background ratio and to avoid false-positive results, the antibody dilutions giving a favorable OD450 (approximately 1.0) were chosen. Considering the cost and the precision of the result, an anti-SF-BM antibody dilution ratio of 1:1000 and a secondary antibody dilution of 1:1000 were selected as the optimal antibody dilutions. Analogously, dilution ratios of 1:5000 for the anti-SF-AP antibody and 1:2000 for the secondary antibody were chosen. Under the optimal conditions, the OD450 value of positive control was approximately 1.0, while the negative control was present in low background levels, ensuring the sensitivity of the following ELISA tests.

Figure 4

Indirect ELISA curves of secondary antibodies against primary antibodies at different dilutions. (A) ELISA curves of the secondary antibody against the anti-SF-BM antibody. (B) ELISA curves of the secondary antibody against the anti-SF-AP antibody.

Cross-reactivity

Figure 5 presents the cross-reactivity results for the B. mori and A. pernyi silks. Silk fibroin was detected by the anti-SF-BM antibody and the anti-SF-AP antibody under the optimal antibody dilutions. As shown in Fig. 5, the OD450 value for the B. mori silk detected using the anti-SF-BM antibody was much higher than the cut-off, but the application of the anti-SF-BM antibody to A. pernyi silk yielded a negative result in the immunological test. Similarly, the anti-SF-AP antibody detected only A. pernyi silk and did not detect B. mori silk. Strikingly, there was no cross-reactivity between B. mori and A. pernyi. These results indicated that the antibodies were prepared well and exhibited high specificity for the two different silk fibroins. Therefore, the developed ELISAs can successfully identify the species that produce each silk.

Figure 5

Cross-reactivity of the anti-SF-BM and anti-SF-AP antibodies at optimized antibody dilutions. The two dashed lines indicate the cut-off values.

Species identification of silk samples via ELISA

The species that produced the silk samples were subsequently identified via ELISA using the highly specific antibodies under the optimal conditions. Figure 6A,B show images of Sample I and Sample II. Sample I, which was unearthed from a Chu tomb (approximately 2300 years old) in Anji county of Zhejiang province, appeared brown and exhibited a rotted morphology. Sample II which exhibited few impurities and a well-woven construction, was kindly provided by Prof. Feng Zhao in China National Silk Museum.

Figure 6

Digital images of Sample I (A) and Sample II (B). (C) ELISA results for Sample I and Sample II. The two dashed lines indicate the cut-off values.

To evaluate the accuracy of the established ELISA method for given silk samples, the two silks were identified according to the procedures described above. The ELISA results for Sample I and Sample II are shown in Fig. 6C. The anti-SF-BM antibody produced a positive result for Sample I but a negative result for Sample II. In contrast, the OD450 value for Sample II with the anti-SF-AP antibody was higher than the cut-off, which indicated that Sample II was produced by A. pernyi. Hence, the developed ELISA method with well-designed primary antibodies is practical and valid for identifying which species produced a given silk.

Species identification of silks via proteomics

Biomarkers for B. mori and A. pernyi silk

The LC-MS/MS chromatograms of the digested fibroins are shown in Fig. 7. The chromatograms represent the most intense peaks of the peptides fragment at every retention time. In Fig. 7, the differences in the two digested fibroins can be easily observed under representative retention times from 10 to 80 min. The prominent peak of the B. mori silk appeared at approximately 19.25 min. However, the prominent peak of the A. pernyi silk appeared earlier, starting at 11 min and increasing to a maximum at 11.89 min. The two digested fibroins exhibited similar peaks at approximately 68 min, with obvious differences in the peaks occurring at retention times of 10 min to 25 min. The MS/MS chromatograms within this time range were subjected to search using Proteome Discoverer software.

Figure 7

Mass chromatograms of the silk fibroins extracted from B. mori silk (A) and A. pernyi silk (B).

There were thirteen proteins identified and screened in B. mori silk, but only four of them showed high score and PSMs. In the A. pernyi silk, fifteen proteins were identified, among which only two were credible proteins relevant to fibroin with high score and PSMs. The other proteins in the B. mori and A. pernyi silk displayed low signals and were not reliably identified. Table 2 shows the seven proteins detected in B. mori and A. pernyi silk. Fibroin heavy chain and fibroin light chain numbered as P05790 and P21828, respectively, were identified as components of fibroin derived from B. mori silk. Fibroin heavy chain fragment (Q99050) and fibrohexamerin (Q9BLL8) originated from Bombyx mandarina (B. mandarina) were also detected, which could be explained because both B. mori and modern B. mandarina were originated from ancient B. mandarina[40-42]. No proteins ascribed to A. pernyi silk fibroin were detected in the B. mori silk. Next, A. pernyi silk was also evaluated, and three proteins were identified: fibroin (O76786), fibroin heavy chain (A0A0K0KR73) and fibroin fragment (Q8ISB3). The fibroin (O76786) displayed high score and PSMs, which credibly verified the silk from A. pernyi. Interestingly, fibroin heavy chain (A0A0K0KR73) ascribed to Antheraea assama and fibroin fragment (Q8ISB3) ascribed to Antheraea mylitta were also observed, which may be attributed to the hybridization of A. pernyi with Antheraea assama and Antheraea mylitta genome. And, withal, none of the proteins from B. mori silk were present in A. pernyi silk. The unique peptides of proteins greater than or equal to 2 are generally considered to be more reliable. Therefore, the four proteins detected in B. mori silk and two proteins (O76786, Q8ISB3) detected in A. pernyi silk could potentially be used as biomarkers for identifying Bombyx and Antheraea genera.

Table 2

Proteins bearing different species-diagnostic peptides between B. mori and A. Pernyi silk.

Sample	Accession numbera	Protein name	Scoreb	Coverage	Matched peptides	Unique Peptides	PSMsc	Species
B. mori silk	P05790	Fibroin heavy chain	798.51	1.86	14	12	240	B. mori
	Q99050	Fibroin heavy chain (fragment)	334.15	24.16	5	3	107	B. mandarina
	P21828	Fibroin light chain	229.25	25.95	7	7	110	B. mori
	Q9BLL8	Fibrohexamerin	51.76	23.64	7	7	26	B. mandarina
A. pernyi silk	O76786	Fibroin	2280.07	28.38	26	18	577	A. Pernyi
	A0A0K0KR73	Fibroin heavy chain	457.70	3.31	6	1	118	A. assama
	Q8ISB3	Fibroin (fragment)	303.42	14.20	5	2	112	A. mylitta

aAccession number corresponds to the UniProt accession number.

bScore obtained from LC-MS/MS identification.

cThe peptide spectrum matches.

To gain a better understanding of the amino acid sequences of B. mori and A. pernyi silk, peptides supporting the identification of fibroin in B. mori and A. pernyi silk are shown in Tables S-1 and S-2, respectively. Altogether, 74 peptide fragments were detected, and eleven peptides fragments were post-translationally modified by the oxidation of methionine or carbamidomethylation of cysteine. A total of 37 peptide fragments were observed in B. mori silk, and the remaining peptides were detected in A. pernyi silk. Modifications mainly occurred in B. mori silk, and there were no modifications detected in A. pernyi silk. All peptide fragments digested with trypsin end in arginine or lysine, which indicated the cutting sites of trypsin.

Species identification of ancient silks via proteomics

To explore the practicality of using the selected biomarkers in silk samples, the two silks were identified via proteomics. Table 3 shows the proteomic results for Sample I and Sample II. Protein accession numbered Q99050, one of the biomarkers found in B. mori silk, was detected in Sample I. The peptide DASGAVIEEEITTK at m/z 1462.73 corresponded to the detected peptide at m/z 1462.73 in B. mori silk. Therefore, it is speculated that Sample I was produced by Bombyx. The identified peptides of Q99050 in Sample I and B. mori silk are shown in Figure S-1. Moreover, proteins accession numbered Q8ISB3 and O76786 were detected in Sample II. Peptides GDGGYGSDSAAAAAAAAAAAAGSGAGGR, at m/z 2208.99, and NAATRpHLSGNER, at m/z 1422.72, of Q8ISB3 were common peptides that were detectable in both Sample II and A. pernyi silk. All the ten peptides from O76786 were also identified in A. pernyi silk. Figures S-2 and S-3 show the identified peptides of Q8ISB3 and O76786, respectively. Thus, the results revealed that Sample II came from Antheraea, which could also be verified via ELISA. Consequently, the biomarkers for Bombyx and Antheraea obtained through proteomics were shown to be effective in an archaeological application. Furthermore, the proteomics method is useful for identifying silk-producing species and could provide positive proof at the sequence level.

Table 3

Proteins bearing species-diagnostic peptides identified in Sample I and Sample II.

Sample	Accessio number	Protein name	Score	Unique peptides	PSMs	MW [kDa]
Sample I	Q99050	Fibroin heavy chain (fragment)	2.65	1	1	18.3
Sample II	O76786	Fibroin	134.04	9	51	215.9
	Q8ISB3	Fibroin (fragment)	277.88	7	90	45.4

Conclusion

In contrast to traditional techniques, the present study provides a series of novel methods for identifying silks at the species and genus levels. In particular, B. mori silk and A. pernyi silk were identified and distinguished from each other by combining traditional methods with immunological techniques and proteomics. Significant differences between B. mori silk and A. pernyi silk were found, ranging from morphology to primary structure. As traditional methods are not applicable for identifying silks with contamination or severe degradation, ELISA and proteomics were applied to distinguish silk samples. The ELISA results showed that the antibodies exhibited high specificity for combining with the corresponding antigen and did not cross-react with the other antigen. Through proteomic analysis, biomarkers for Bombyx and Antheraea were identified. The subsequent results revealed that the species that produced a given silk sample can be identified using these biomarkers. Thus, ELISA and proteomics are suitable and efficient tools for identifying the species that produced a sample of silk. The strategy that we adopted in this study will lead to new research avenues for archaeological samples.

Supporting Information for: Species Identification of Bombyx mori and Antheraea pernyi silk via Immunology and Proteomics