Introduction

The Bombyx mori is an important silk producer, which is believed to have been domesticated from the wild mulberry silkworm 5000–10,000 years ago1. The composition and quality of the silk produced is dependent on the raw material of the cocoon, which in turn greatly depends on the conditions during the silkworm rearing stages. Silkworms undergo complete metamorphosis during rearing. The four developmental stages of egg, larva, pupae, and adult are completely different in terms of morphology and physiology. After hatching, the larvae exclusively feed on mulberry leaves and develop into mature silkworms after approximately 20 days, prior to which they will pass through five instars. The first to third instar are referred to as the young larval stage, while the fourth and fifth instar compose the old larval stage2. The ideal rearing conditions and the demand for mulberry leaves vary between these stages. This complex life history can be analyzed through stable isotope technology. Stable isotope technology is widely used in feeding ecology to study the fractionation mechanisms of stable isotopes, in different organisms or within different tissues of the same organism, that can correlate with various ecological factors3,4,5. The fractionation and enrichment characteristics of 13C and 15N can be used to determine an animal's food source6, study its nutritional health7,8 and explore complex life history processes8. Stable isotope values differ between consumers and their food sources by a relatively definite discriminant value.

While most researchers using stable isotope technology in feeding ecology have focused on aquatic systems9,10,11, stable isotope technology has also been widely applied in geographical origin traceability of agricultural products. Regina et al.12 measured the stable isotope ratios for hydrogen, carbon, nitrogen, and oxygen of silk fabrics samples found in the underground palace of the Famen Temple, Shaanxi, China, to determine if stable isotope ratios can assist in provenance analysis of silk. Although data analyses were lacking in their report, the authors claimed that their results indicate that the silk raw material or silk cocoons of different provenance are clearly distinguishable from each other. They believed that through the combination of information regarding centers of sericulture during the Tang Dynasty, it may be possible to determine the region from which the silk fiber came.

Therefore, by exploring the isotope relationship in sericulture (production of silk fiber from the rearing of mulberry groves to the harvesting of cocoons), it is possible to trace the provenance of silk contained in both fabrics and cocoons. The first factor that must be accounted for when studying the origin of silk fabrics is that different breeds or rearing conditions may affect silkworm physiological development and cocooning. Secondly, the fractionation of stable isotopes is often associated with significant physiological shifts including changes in metabolic rate, energy expenditure, and fasting, depletion of fat reserves, and variation in hormones13. Previous studies provide compelling evidence that the magnitude of dietary fractionation differs among closely related species14 and is influenced by various exogenous factors, such as starvation15. Nearly all the research into the biochemical processes that govern incorporation of stable isotope ratio into consumer tissues stems from controlled feeding experiments in the lab16,17,18, while relatively little has been explored regarding the mulberry-silkworm-cocoon ecosystem. Importantly, given that the Bombyx mori exclusively feed on mulberry leaves, this provides the unique opportunity to investigate isotopic changes associated with metamorphosis and physiological processes, with little influence from diet. In theory, any isotopic offsets in mulberry-silkworm-cocoon ecological processes are the results of physiological effects associated with metamorphosis or any laboratory control conditions.

In this study, we used stable isotope analyses to determine the hydrogen, oxygen, carbon, and nitrogen isotope values of different silkworm tissues from different silkworm strains at different stages of development and measured the changes in stable isotope values of silkworms under starvation stress. Establishing a biochemical framework for the fractionation of stable isotopes using the mulberry-silkworm-cocoon ecosystem will expand our ability to resolve stable isotope anomalies in small regional-scale systems and potentially unlock new applications for stable isotope data in tracing the origin of silk textiles.

Materials and methods

Silkworm rearing and sample preparation

Three Bombyx mori strains, Jingsong haoyue (JSHY), Haoyue Jingsong (HYJS), and Hua Kang No.3 orthogonal (HK3), which were bred by Zhejiang Academy of Agricultural Science, were selected for experiments. Larvae were reared on fresh mulberry leaves in an environment with a temperature of 25 ± 2 °C and relative humidity of 75% ± 4%. The fresh mulberry leaves were collected from the China National Silk Museum with the permission of relevant staffs. Fresh mulberry leaves, newly hatched silkworms, silkworms at the fifth instar, silkworm excrement of fifth-instar larvae, and cocoons were collected and dried at 70 °C. To eliminate individual differences, all the samples were ground together into a fine powder.

Starvation experiment

Silkworms of JSHY and HYJS were chosen for a starvation experiment at the beginning of their third instar. For comparison, the larvae of each strain were divided into 7 groups and starved for 6, 12, 24, 36, 48, 60, and 72 h, respectively. Following starvation, around 20 larvae were collected from each group, dried, and ground into powder for testing, while the remaining larvae were left to resume feeding.

Stable isotope analysis

Silkworm samples were transferred into a tin capsule for carbon and nitrogen isotope ratio analysis using a stable isotope ratio mass spectrometer (IRMS, MAT-235, Thermo Fisher Scientific Inc, USA) equipped with an elemental analyzer (Flash 2000 HT; Thermo Fisher Scientific Inc, USA). The samples were combusted at 960 °C in a combustion reactor and a column oven at 50 °C. A helium carrier gas with a purity of 99.999% was set to 100 mL/min. When measuring the stable isotopes of hydrogen and oxygen, the tin cup was replaced with a silver cup and the samples were analyzed in a reactor tube at 1380 °C.

The isotope ratios of H, O, C and N were expressed as parts per thousand (‰) against the international reference standards:

$$\delta = \frac{{R_{sample} - R_{standard} }}{{R_{standard} }} \times 1000 {{\permil}}$$

where Rsample represents the ratio of the heavy isotopes to the light isotopes of the element (C/N/H/O) in the sample, such as 13C/12C, 15N/14N, 2H/1H, 18O/16O; and Rstandard is the ratio of the heavy isotopes to the light isotopes in the internationally recognized standard, using Vienna standard mean ocean water for H and O, Vienna Peedee Belemnite for C and N. The analytical precision and reproducibility (n = 5; represented as mean ± standard deviation) based on the method validation analysis results were ≤ 0.1‰ for δ13C, ≤ 0.1‰ for δ13N, ≤ 0.4‰ for δ2H and ≤ 0.1‰ for δ18O, respectively. For the calibration, the certified reference materials IAEA-601 (δ18O = 23.3‰), IAEA-CH-7 (δ2H =  − 100.3‰; δ13C = 32.151‰) and IAEA-310 (δ15N = 47.2 ‰) were measured at intervals of 8 samples.

Data processing method

Simple data distribution and mapping were performed with Excel (Microsoft) and Origin (OriginLab Corporation) 2018. Linear Discriminant Analysis (LDA) of stable isotope characteristic values in the different samples was performed using SPSS v22.0 (IBM).

Results and discussion

The effects of strains on the isotope composition ratios during silkworm development

The results of 3 different strains Bombyx mori isotope value are shown in Fig. 1 and Table S1. The finding of a negligible difference in the value of δ13C among the newly-hatched silkworm and among fifth instar silkworms would indicate that silkworm strain has little effect on carbon stable isotope levels in silkworm rearing. In other words, the fractionation of δ13C by three strains Bombyx mori was consistent at the newly-hatched and fifth instar stages. However, our data showed a significant variation in δ13C values among silkworm pupae and cocoons, which might be driven by a differential selection between heavy and light elements during pupation and cocooning. The isotope analyses suggest that the molecular mechanisms underpinning the development of the silk glands lead to differences in nutrient allocation and ultimately results in variations in δ13C values between silkworm pupae and cocoons. Specifically, the process of silk protein production leads to the loss of protein, while lipids and carbohydrates are stored in silkworm pupae for metamorphosis19. Lipids and carbohydrates are known to have vastly different δ13C values compared to proteins20,21,22,23. Furthermore, there was also a difference in δ13C values among the 3 different strains silkworm pupae, which might be due to strain-specific differences in metabolic turnover rates.

Figure 1
Figure 1The alternative text for this image may have been generated using AI.
Full size image

Line chart of carbon stable isotope ratios in samples.

The fractionation of nitrogen, depicted in Fig. 2 and Table S2, is more complicated than that of hydrogen. The δ15N values from the 3 silkworm strains differed significantly in certain samples, including in newly-hatched larvae, but not in silkworm excrement or fifth-instar larvae samples. The significant differences of δ15N among the 3 newly-hatched silkworm strains might be attributed to the previous generation and indicates that mating and egg laying may result in an inconsistent kinetic nitrogen isotope fractionation among the different strains, which may selective release 14N into sperm and eggs24. The same results were found for δ2H and δ18O. However, the difference of δ15N value in the fifth instar silkworm was reduced to 0.1‰. The enrichment of δ15N in silkworms may be due to transamination during protein synthesis25,26,27. What is noteworthy is that almost all the heavy elements within the silkworm excrement were enriched, except for δ15N. Moreover, the δ15N values of silkworm cocoons varied greatly between strains. These changes may be attributed to the fact that the spinning process of a silkworm requires high energy input and involves complex biochemical reactions. At the same time, it is unclear how differential routing of amino acids might affect the fractionation of stable isotopes among the 3 silkworm strains. But what is clear, is that there are myriad opportunities during cocooning for isotopic discrimination to occur. Therefore, due to its instability, the δ15N value was deemed not credible to be used as a sole indicator of the origin of silk fabrics.

Figure 2
Figure 2The alternative text for this image may have been generated using AI.
Full size image

Line chart of nitrogen stable isotope ratios in samples.

The δ2H values displayed little variation across the three different silkworm strains (Fig. 3, Table S3). However, the δ2H value of silkworm excrement samples were significantly higher than that of the larvae, which is the opposite distribution compared to that seen in δ15N values, suggesting an enrichment of deuterium in silkworm excrement. The δ2H value decreased from fifth-instar larvae to pupae but was significantly enriched in the silkworm cocoons, which was consistent with δ13C enrichment.

Figure 3
Figure 3The alternative text for this image may have been generated using AI.
Full size image

Line chart of hydrogen stable isotope ratios in samples.

Finally, the δ18O values differed insignificantly among the three different silkworm strains and the variation across the silkworm samples was similar to that of the δ2H values (Fig. 4, Table S4). Indeed, the δ18O value of silkworm excrement samples was higher than that of the larvae, decreased from larva to pupae and was finally increased again in the cocoon samples, which was consistent with the variation trend of δ2H.

Figure 4
Figure 4The alternative text for this image may have been generated using AI.
Full size image

Line chart of oxygen stable isotope ratios in samples.

Stable isotope fractionation of hydrogen, oxygen, carbon and nitrogen

The scatter diagram for nitrogen and carbon isotope values (Fig. 5) shows that the enrichment among mulberry leaf, silkworm excrement, silkworm at fifth instar and newly-hatched silkworm samples is assumed to be low (<0.5‰). In contrast, δ13C was enriched in silkworm cocoons, a phenomenon likely linked to the discrimination of light and heavy isotopes derived from a combination of biochemical and physiological processes, such as metamorphosis28,29,30. By contrast, the δ15N levels change significantly with diet. Indeed, heavy nitrogen was enriched in newly-hatched silkworms relative to the mulberry leaves they consumed. The mulberry leaf, fifth-instar larva, and newly-hatched larva samples arranged on a nearly vertical line in Figure 5, suggesting fractionation of nitrogen stable isotopes during larval development but not of carbon.

Figure 5
Figure 5The alternative text for this image may have been generated using AI.
Full size image

Distribution map of carbon and nitrogen stable isotope composition in samples.

The scatter diagram for hydrogen and oxygen (Fig. 6) revealed a distinct separation in the clustering of different silkworm samples. Silkworm pupae had the lowest δ2H and δ18O values, suggesting the depletion of hydrogen and oxygen stable isotopes in silkworm pupae. The distributions of hydrogen and oxygen stable isotopes in newly-hatched larvae and silkworm at fifth instar were found to be small, compact and concentrated, suggesting the ratios are not helpful in distinguishing instars. Mulberry leaf and silkworm excrement samples both displayed the largest δ2H and δ18O values. Overall, these data reveal that the hydrogen and oxygen stable isotopes undergo significant fractionation during the lifetime of Bombyx mori.

Figure 6
Figure 6The alternative text for this image may have been generated using AI.
Full size image

Distribution map of hydrogen and oxygen stable isotope composition in samples.

The effect of starvation stress on the isotope composition ratios during silkworm development

The results of stable isotope values of Bombyx mori under different starvation conditions are shown in Fig. 7, Tables S5-1 and S5-2. It is obvious from the data that the δ13C values were significantly enriched in silkworms after starvation and there was a statistically significant positive correlation between the isotope values and starvation time. Moreover, with increasing starvation time, the δ13C values of both JSHY and HYJS silkworms increased by approximately 0.6‰, indicating that both stains of silkworms tend to use 13C under hunger stress. The different pattern of δ18O values for both strains with time of starvation showed an opposite trend to δ13C. On the other hand, the δ15N values of JSHY silkworms increased 0.8‰, while the δ15N values of HYJS silkworms decreased by 0.8‰ over 72 h. Furthermore, the changes in heavy nitrogen levels with increasing starvation time followed an opposite trend in the two different strains, with the highest variation observed at 24 h of starvation. This indicates that different silkworm stains of silkworm exhibit different nitrogen isotope usage under hunger stress.

Figure 7
Figure 7The alternative text for this image may have been generated using AI.
Full size image

Distribution of stable hydrogen, oxygen, carbon, and nitrogen isotope compositions under starvation conditions.

The stable hydrogen, oxygen, carbon, and nitrogen isotope ratios were compared before and after three days of starvation and after three days of re-feeding (Fig. 8, Tables S6-1 and S6-2). After resumption of a normal diet, the δ13C values of the silkworms continued to be enriched, indicating that the fractionation of δ13C is mainly influenced by foreign nutrients and metabolism rather than starvation pressure. The same results were seen for δ2H. However, the δ18O values of both silkworm strains were decreased after 72 h of starvation but returned to original levels after resumption of diet. These results suggest a statistically significant diet-dependent change in the δ18O values. Conversely, the analysis of nitrogen stable isotope levels across the two strains revealed significant strain-specific differences after three days of re-feeding. This suggests that there is a differential selection of 15N by JSHY and HYJS strains under starvation stress.

Figure 8
Figure 8The alternative text for this image may have been generated using AI.
Full size image

Changes of stable isotopes of hydrogen, oxygen, carbon, and nitrogen before and after starvation.

Discriminant analysis of hydrogen, oxygen, carbon and nitrogen stable isotopes

LDA, first proposed by Fisher31, is a supervised pattern recognition method32 carried out in an m-space (m = number of variables). It calculates an m-1 dimensional surface that separates two established categories as far as possible. According to the criteria of minimizing intra-class variance and maximizing inter-class variance, a discriminant function is obtained and then used to classify the discriminant samples33. The mulberry leaf, newly-hatched larva, fifth-instar larva, excrement, pupae and cocoon samples were used for discriminant analysis. The model was based on four independent variables including δ2H, δ18O, δ13C, and δ15N and the results were shown in Fig. 9. The discriminant score analysis (Table 1, Fig. 10) revealed the respective classification accuracy of the back generation and the cross-examination of δ2H, δ18O, δ13C and δ15N of samples to be 100.0% and 100.0% respectively, indicating that each of the seven sample types had a distinct stable isotope "fingerprint."

Figure 9
Figure 9The alternative text for this image may have been generated using AI.
Full size image

Distribution of stable isotopes of hydrogen, oxygen, carbon and nitrogen before and after starvation.

Table 1 Discrimination results of samples.
Full size table
Figure 10
Figure 10The alternative text for this image may have been generated using AI.
Full size image

Discriminant analysis of samples.

Conclusion

Silkworm strain had little effect on the carbon stable isotope ratios in the majority of the silkworm-related samples. A large difference in δ15N was observed within newly-hatched silkworm samples in JSHY and HK3 strains, but the 15N levels in the fifth instar silkworms and their excrement was relatively stable. The difference of δ15N among the three newly-hatched silkworm strains indicated that mating and egg laying may result in an inconsistent kinetic nitrogen isotope fractionation. In contrast to δ15N, the δ2H and δ13C values were enriched in silkworm cocoon samples. Moreover, the δ13C signatures of silkworm samples displayed significant starvation-time effects, whereas the δ15N values displayed a diametrically opposite trend between the JSHY and HYJS strains. This indicates that different silkworm strains exhibit different nitrogen isotope usage under starvation stress. The carbon stable isotope ratios in silkworm excrement, fifth-instar larvae, and pupae samples demonstrated significant fractionation, while the newly-hatched larvae, fifth instar larvae, and mulberry leaf samples exhibited mostly nitrogen fractionation. Mulberry leaf and silkworm excrement were found to be highly enriched for 2H and 18O.

Starvation experiments revealed that δ13C levels increased due to metabolic requirements as indicated by continual enrichment of δ13C after diet resumption for 72 h. Similarly, heavy hydrogen was found to be gradually enriched under starvation and increased further upon diet resumption. However, 18O was depleted after starving and enriched after re-feeding, which indicated that starvation greatly impacts δ18O, but not δ2H. Under starvation conditions, orthogonal (JSHY) and anti-crossing (HYJS) newly-hatched silkworms accumulated heavy nitrogen differently, showing that the different hybridization methods had a great impact on the larval feeding habits. Meanwhile, isotopic fractionation occurred in the silkworm ecological process.

Finally, LDA was able to discriminate the mulberry leaf, newly-hatched larvae, silkworm excrement, fifth-instar larvae, silkworm pupae, and silkworm cocoons with a classification accuracy rate of 100.0%, indicating that the specific "fingerprint" information carried by stable isotopes can effectively be utilized to identify and distinguish each of these life stages and substances. Due to the high classification accuracy rate and the appropriate discrimination elements, we propose that IRMS can be used to help trace nutrient transmission during silkworm development and silk production. As a result, the combination of LDA and isotope measurements may provide the foundation for new techniques and strategies for the traceability of textiles.