American silk moth, Antheraea polyphemus Cramer 1775 (Lepidoptera Saturniidae), is one of the largest and most beautiful silk moths (Hall, Reference Hall2012) and has potential significance in sericulture for food consumption and silk production. Dutch merchant Pieter Cramer first described this species in 1775, and the name, polyphemus, is inspired by the Greek mythological giant cyclops, which boasts an enormous round eye in the centre of his forehead (Himmelman, Reference Himmelman2002), and resembles the big eyespots located in the centre of the hind wings of A. polyphemus. This species is widely distributed in the USA and Canada (Ferguson, Reference Ferguson, Dominick, Edwards, Ferguson, Franclemont, Hodges and Munroe1972). Except Arizona and Nevada, they can be found in all 48 states, from southern Canada to Mexico (Tuskes et al., Reference Tuskes, Collins and Tuttle1996). It can feed at least 50 plant hosts belonging to more than 30 genera and may be the most polyphagous of the North American silk moth (Ferguson, Reference Ferguson, Dominick, Edwards, Ferguson, Franclemont, Hodges and Munroe1972). Silk fibres produced by A. polyphemus have superior silk qualities such as good tensile strength and high elongation, similar to Bombyx mori L. (Reddy and Yang, Reference Reddy and Yang2010). A. polyphemus is frequently employed in many physiological studies, especially about the molecular mechanisms of sex pheromone action (Li et al., Reference Li, Wang and Zhang2020; Steinbrecht, Reference Steinbrecht1992), due to its huge size and simplicity of rearing (Villiard, Reference Villiard1975). For example, proteins such as the insect sensory neuron membrane protein-1 (SNMP1) and pheromone binding protein were first identified in A. polyphemus (Rogers et al., Reference Rogers, Sun, Lerner and Vogt1997; Vogt and Riddiford, Reference Vogt and Riddiford1981). Its chorion has also been studied as a model system (Hamodrakas et al., Reference Hamodrakas, Margaritis, Papassideri and Fowler1986). To date, approximately 100 nucleotide sequences of A. polyphemus are available in the NCBI database, among which 73 are derived from chorion protein and mitochondrial COI sequences (Table S1); the genetic information regarding this insect is severely limited.
The insect mitochondrial genome (mitogenome) is widely used for species identification and phylogenetic inference of insects (Cameron, Reference Cameron2014). Wild silk moths (Saturniidae) that contribute to wild silk production and insect food are distributed nearly all over the world, with 3,454 described species across 180 genera (Kitching et al., Reference Kitching, Rougerie, Zwick, Hamilton, St Laurent, Naumann, Ballesteros and Kawahara2018). However, only 45 mitogenomes from this family are accessible in GenBank (https://www.ncbi.nlm.nih.gov). As for 119 described species of the genus Antheraea, only seven mitogenomes can be available, including Antheraea pernyi Guérin-Méneville (Liu et al., Reference Liu, Li, Pan, Dai, Zhu, Lu and Xiang2008), Antheraea yamamai (Kim et al., Reference Kim, Kim, Hong, Kim, Kang, Hwang, Han, Jin and Kim2009), Antheraea frithi Bouvier (Shantibala et al., Reference Shantibala, Devi, Lokeshwari, Anju and Luikham2018), Antheraea assamensis Helfer (Singh et al., Reference Singh, Kabiraj, Sharma, Chetia, Mosahari, Neog and Bora2017), Antheraea proylei In 981 Jolly (Yang et al., Reference Yang, Zhang, Chen, Chen, Li and Liu2019), Antheraea formosana Sonan (Cheng et al., Reference Cheng, Huang, Cheng, Tseng, Wang, Lai and Hung2022) and Antheraea mylitta Drury (Balakrishnan et al., Reference Balakrishnan, Dubey, Debnath, Subrahmanyam, Arunkumar, Pandey, Ponnuvel, Sivaprasad and Sathyanarayana2024). All these seven Antheraea species are distributed in Asia, the origin centre of the genus Antheraea (Jolly, Reference Jolly1980).
So far, the phylogenetic relationship and divergence time of A. polyphemus and its Asian relatives remain unknown. To address these issues, this study determined two mitogenomes of A. polyphemus from the USA and Canada, respectively. Then, mitogenome-based phylogenetic analysis was performed to infer the placement of A. polyphemus within the family Saturniidae. Finally, mitogenome sequences were used to estimate the divergence time between A. polyphemus and its Asian sister relatives to infer when the dispersal of A. polyphemus from Asia into North America might occur. This study will improve our understanding the phylogenetic relationship and early evolution history of Saturniidae.
Materials and methods
Sample Collection
Pupal specimens of A. polyphemus from the USA were provided by Shady Oak Butterfly Farm (12876 SW County Rd 231 in Brooker, Florida 32622). Moth specimens from Canada were purchased on eBay. All individuals were preserved at −80℃ in the Laboratory of Insect Genetics and Genome, Shenyang Agricultural University. Total DNA was extracted from partial body using the traditional phenol/chloroform method.
Mitogenome Amplification and Sequencing
Two universal primer pairs, COI-F/COI-R and ND5-F/ND4-R (Simon et al., Reference Simon, Frati, Beckenbach, Crespi and Liu1994), and three specific primer pairs were used to amplify the whole mitogenome of A. polyphemus (table 1). LA Taq DNA Polymerase (TaKaRa Co., Dalian, China) was used in polymerase chain reaction (PCR) of 25 μL volume. PCR amplification procedure was as following: initial denaturation at 96℃ for 2 min, followed by 35 cycles of 10 s at 97℃ and 5 min at 55℃, and a subsequent 10 min final extension step at 72℃. After purification with a Gel Extraction Kit (Kangwei Century Biotechnology Co., LTD., Taizhou, China), these PCR fragments were sequenced with the Sanger method via primer walking.
Primers used for PCR amplification
Mitogenome Assembly, Annotation and Sequence Analyses
The resulting sequences were assembled using Clustal X 1.83 (Aiyar, Reference Aiyar2000), by alignment of neighbouring fragments. The mitogenome of A. polyphemus was annotated using MITOS Web Server (http://mitos.bioinf.uni-leipzig.de/index.py) (Donath et al., Reference Donath, Jühling, Al-Arab, Bernhart, Reinhardt, Stadler, Middendorf and Bernt2019). Finally, a mitogenome map was visualised with the online tool OrganellarGenomeDRAW (https://chlorobox.mpimp-golm.mpg.de/OGDraw.html, accessed on 23 September 2024) (Greiner et al., Reference Greiner, Lehwark and Bock2019). The structures of 22 tRNAs were predicted using MITOS2 v2.1.9 in Galaxy Europe (Donath et al., Reference Donath, Jühling, Al-Arab, Bernhart, Reinhardt, Stadler, Middendorf and Bernt2019; Galaxy Community, 2024) with default parameters, followed by manual correction.
Tandem Repeats Finder program (Benson, Reference Benson1999) was used to search for potential tandem repeats in the A+T-rich region. The mitogenome length, nucleotide composition, and composition skew were analysed using PhyloSuite v1.2.3 (Zhang et al., Reference Zhang, Gao, Jakovlić, Zou and Wang2020). The base composition bias was calculated based on previously known formulas (Perna and Kocher, Reference Perna and Kocher1995). Nucleotide diversity was calculated using DNASP 6.0 (Rozas et al., Reference Rozas, Ferrer-Mata, Sánchez-delbarrio, Guirao-Rico, Librado, Ramos-Onsins and Sánchez-Gracia2017).
Data Acquisition and Phylogenetic Analysis
The COI sequences of the genus Antheraea were downloaded from NCBI (https://www.ncbi.nlm.nih.gov/, accessed on 20 October 2024) and BOLD (www.barcodinglife.org, accessed on 20 October 2024) for phylogenetic analysis (Ratnasingham et al., Reference Ratnasingham, Wei, Chan, Agda, Agda, Ballesteros-Mejia, Boutou, Bastami, Ma and Manjunath2024). These COI sequences were used for species identification and phylogenetic inference within the genus Antheraea.
The nucleotide sequences of 45 mitogenomes of Saturniidae species available were downloaded from the NCBI database to illustrate their phylogenetic relationships (table 2). These 45 species belonged to the Agliinae, Salassinae, Saturniini, Attacini, Eochroini, Micragonini, Urotini, and Bunaeini tribes.
Information about the mitogenomes used for phylogenetic analysis
Phylogenetic analysis was performed to infer the placement of A. polyphemus within the family Saturnidae. The mitogenome sequences were aligned using MAFFT v7 (Katoh and Standley, Reference Katoh and Standley2013). General Time-Reversible + G (gamma-distributed rate heterogeneity) + I (invariable sites) was identified as the suitable evolutionary model using MEGA-XI (Tamura et al., Reference Tamura, Stecher and Kumar2021). Bayesian inference (BI) and maximum-likelihood (ML) methods were performed using MrBayes ver. 3.2.7 (Ronquist et al., Reference Ronquist, Teslenko, van der Mark, Ayres, Darling, Hhna, Larget, Liu, Suchard and Huelsenbeck2012) and IQ-TREE ver. 2.3.6 (Minh et al., Reference Minh, Schmidt, Chernomor, Schrempf, Woodhams, Von Haeseler and Lanfear2019), respectively. Finally, the phylogenetic tree was visualised using ITOL v6 (Letunic and Bork, Reference Letunic and Bork2024).
Divergence Time Estimation
Divergence time was estimated using MCMCTree program and baseml in PAML v4.9 (Yang, Reference Yang2007). The reference divergence times for Drosophila melanogaster Meigen and B. mori (151.9–344.7 MYA), A. polyphemus and B. mori (39.8–95.1 MYA), and B. mori and Bombyx mandarina Moore (0.004–0.011 MYA) were obtained from the Time Tree database (Kumar et al., Reference Kumar, Stecher, Suleski and Blair2017). The effective sample size statistics was used to examine the convergence of the run parameters in Tracer v1.7 (Rambaut et al., Reference Rambaut, Drummond, Xie, Baele and Suchard2018). FigTree v1.4.4 (Rambaut, Reference Rambaut2018) was used to display the consensus tree.
Results
Species Identification and Genetic Diversity Analysis
The COI fragment of 658 bp that corresponds to the DNA barcode was used for species identification. NCBI BLAST search revealed that the 658-bp fragment had 100% identity with the A. polyphemus voucher BIOUG<CAN>: 04HBL005189 (GenBank accession no. GU093744) released by the International Barcode of Life, indicating that the samples used in this study belonged to A. polyphemus. This result was also confirmed by phylogenetic analysis (Fig. S1).
Seven COI sequences of A. polyphemus available in the NCBI and BOLD databases were further used to calculate haplotype diversity (Hd) and nucleotide diversity (π). The result suggested a high Hd (0.714 > 0.5) and low π (0.00405 < 0.005) for A. polyphemus (Grant and Bowen, Reference Grant and Bowen1998).
Mitochondrial Genome Structure and Composition
The two mitogenomes with NCBI accession numbers NC_072344 from the USA and OP870082 from Canada showed only two transitions and five indels. The NC_072344 mitogenome was selected for further analysis (fig. 1; table 3). The mitogenome of A. polyphemus from the USA was circular with 15,346 bp, and encoded typical mitochondrial 37 genes. The heavy (H) chain encoded 9 protein-coding genes (PCGs) and 14 transfer RNA genes (tRNAs), while the light (L) chain encoded 4 PCGs, 8 tRNAs, and 2 ribosomal RNA genes (rRNAs). This gene distribution pattern is identical to Saturniidae mitogenomes (Chen et al., Reference Chen, Zhang, Jin, Yang, Li and Liu2021; Kim et al., Reference Kim, Min, Jeong and Kim2018). Nineteen intergenic spacers with 181 bp in total were identified, among which the longest intergenic spacer was 50 bp between tRNAGln and NAD2. Three gene overlaps with 11 bp in total were identified, with the longest overlap (7 bp) observed between atp8 and atp6.
Circular map of A. polyphemus mitogenome. Genes outside and inside the outermost circle are transcribed in the clockwise and counterclockwise direction, respectively, with PCGs in yellow, rRNAs in red, and tRNAs in blue. The circle inside the GC content graph marks the 50% threshold.
Annotation of A. polyphemus mitogenome
OL indicates within-gene overlap. IGS represents intergenic spacer sequences.
In A. polyphemus mitogenome, the content of the AT base was 80.9%, which was significantly higher than that of the GC base (19.2%), similar to previously sequenced lepidopteran mitogenomes (Dai et al., Reference Dai, Zhu, Liu, Wei and Liu2013; Singh et al., Reference Singh, Kabiraj, Sharma, Chetia, Mosahari, Neog and Bora2017). The AT skew and GC skew values were −0.028 and −0.198, indicating a higher abundance of T compared to A and C compared to G.
All PCGs exhibited a negative AT skew, suggesting that they were biased towards T, as observed in A. pernyi (Liu et al., Reference Liu, Li, Pan, Dai, Zhu, Lu and Xiang2008). All PCGs used typical ATN (ATG, ATT, and ATA) as the start codon, except COI that might use CGA, which was a common feature in Lepidoptera species (Wu et al., Reference Wu, Zhao, Su, Luo and Zhu2016). Nine PCGs used TAA as the stop codon, and four PCGs (COX1, COX2, NAD3, and NAD5) utilised T, which was commonly recognised across arthropods mitogenome (Ojala et al., Reference Ojala, Montoya and Attardi1981).
Among 22 tRNA genes identified in A. polyphemus mitogenome, the longest tRNA gene (72 bp) and smallest tRNA gene (63 bp) were tRNALys and tRNACys, respectively. Almost all tRNAs of A. polyphemus had typical clover-leaf secondary structures with four arms, except for the DHU arm of tRNASer(AGN) (Fig. S2), as observed in most metazoan mitogenomes (Garey and Wolstenholme, Reference Garey and Wolstenholme1989).
Two ribosomal RNA genes were located between tRNALeu(CUN) and A+T-rich region, with 775 and 1,365 bp, respectively, and separated by tRNAVal.
The A+T-rich region in A. polyphemus mitogenome was 360 bp in size, and might serve as the initiation site of mitochondrial replication (Cameron, Reference Cameron2014). By comparing with other lepidopteran mitogenomes (Liu et al., Reference Liu, Zhu, Dai, Wei and Liu2012; Shantibala et al., Reference Shantibala, Devi, Lokeshwari, Anju and Luikham2018; Li et al., Reference Li, Lin, Zhang, Hu and Ding2018), several conserved structures and repeats were identified, including the ATAGA motif (Saito et al., Reference Saito, Tamura and Aotsuka2005), 19-bp poly-T stretch (Saito et al., Reference Saito, Tamura and Aotsuka2005), microsatellite (TA)8 repeat sequence (Kim et al., Reference Kim, Park, Kim, Kang, Kim, Jin, Han and Kim2012), and poly-A region intermittently interrupted by T (Zhang and Hewitt, Reference Zhang and Hewitt1997) (fig. 2). Unlike A. pernyi and A. roylei that contained 38-bp tandem repeats (Arunkumar et al., Reference Arunkumar, Metta and Nagaraju2006; Liu et al., Reference Liu, Li, Pan, Dai, Zhu, Lu and Xiang2008), the A+T-rich region of A. polyphemus lacked tandem repeat components.
Features identified in the A+T-rich region of A. polyphemus, including the ATATG motif (red), poly-T stretch (yellow), ATTTA sequence (green), microsatellite TA repeat sequences (purple), and poly-A stretch (blue).
Phylogenetic Analysis and Divergence Time Estimation
To determine the phylogenetic placement of A. polyphemus, a wild silk moth native to North America, phylogenetic relationships based on ML and BI methods were constructed using 13 PCGs from 46 Saturniidae species (fig. 3 and S3). D. melanogaster (NC_024511) and two Bombycidae species B. mori (AB070264) and B. mandarina (AY301620) were used as outgroups. The topologies of ML and BI methods were highly similar, with slightly differences on node support value. In the phylogenetic tree, all species in the genus Antheraea used were grouped, as expected. Our phylogenetic analyses showed that A. polyphemus has a close relationship with A. formosana and A. assamensis, which are distributed into Asia, among eight Antheaea species used in this study. Interestingly, two clades were recovered for eight Antheraea species used: (A. polyphemus + (A. formosana + A. assamensis)) and ((A. mylitta + A. frithi) + (A. yamamai + (A. proylei + A. pernyi))).
Phylogenetic tree inferred from nucleotide sequences of 13 PCGs of 46 Saturniidae species using the maximum likelihood method.
These mitogenome data were further used to estimate the divergence time of A. polyphemus from A. formosana and A. assamensis (fig. 4), which was estimated to be 18.18 MYA during the Miocene Epoch (23–5 MYA) (Jaramillo et al., Reference Jaramillo, Romero, D’Apolito, Bayona, Duarte, Louwye, Escobar, Luque, Carrillo-Briceño, Zapata, Mora, Schouten, Zavada, Harrington, Ortiz and Wesselingh2017).
Chronogram showing divergence time estimates for major lineages of Saturniidae. Scale: 1 = 100 million years.
Discussion
In this study, we provided the first mitogenome of the Antheraea genus from the Americas. The two mitogenomes of A. polyphemus from the USA and Canada exhibited highly identity to each other. Further comparison revealed a high level of identity with other known Saturniidae mitogenomes (Chen et al., Reference Chen, Zhang, Jin, Yang, Li and Liu2021; Kim et al., Reference Kim, Min, Jeong and Kim2018). Additionally, analysis of the COI sequences available suggested that the population of A. polyphemus might experience a rapid expansion after the bottleneck effect and show low genetic diversity (Grant and Bowen, Reference Grant and Bowen1998). These findings indicate that any one population of A. polyphemus can be used as a seed bank for further domestication for mass rearing.
Our mitogenome-based analysis confirmed the phylogenetic placement of A. polyphemus within the genus Antheraea. Although A. polyphemus was once classified under the genus Telea, its differences from the Old World species in the genus Antheraea were deemed insignificant (Hall, Reference Hall2012). Ferguson (Reference Ferguson, Dominick, Edwards, Ferguson, Franclemont, Hodges and Munroe1972) advised choosing the word Antheraea rather than Telea to avoid any misunderstanding because Antheraea has been more frequently referenced in the literature. The mitogenome-based phylogenetic analysis also showed that eight Antheraea species used could be separated into two clades: one included A. polyphemus, A. formosana, and A. assamensis, another contained A. mylitta, A. frithi, A. yamamai, A. proylei, and A. pernyi, as observed in a previous study (Cheng et al., Reference Cheng, Huang, Cheng, Tseng, Wang, Lai and Hung2022).
We found that the placements of Cricula trifenestrata Helfer, Cricula andrei Jordan, and Saturnia haraldi Schawerda in the mitogenome-based phylogenetic tree differed from those in the morphology-based taxonomology. In the mitogenome-based phylogenetic tree, S. haraldi was recovered as the most basal lineage of Saturniidae, as found in a previous study (He et al., Reference He, Wang and Chen2017); the tribe Saturniini was found to be a nonmonophyletic group, which was interrupted by Agliinae, Salassinae, Bunaeini, Urotini, Micragonini, Eochroini, and Attacini, as previously detected (Nethavhani et al., Reference Nethavhani, Straeuli, Hiscock, Veldtman, Morton, Oberprieler and van Asch2022). The inclusion of C. andrei or not in the mitogenome-based phylogenetic analysis can influence the topological structure of Saturniidae species. When C. andrei was removed, the phylogenetic relationship (Figs. S4 and S5) were highly similar to those of previous studies (Chen et al., Reference Chen, Zhang, Jin, Yang, Li and Liu2021; Singh et al., Reference Singh, Mosahari, Sharma, Neog and Bora2024), where C. trifenestrata showed a close relationship with the genera Eriogyna, Saturnia, and Rinaca. When C. andrei was included in the phylogenetic tree (Figs. 3 and S3), the genus Cricula containing C. andrei and C. trifenestrata used in this study was recovered as the basal lineage of Saturniidae. The variable topological structures of C. trifenestrata, C. andrei, and S. haraldi were likely due to low sequence quality (Langley et al., Reference Langley, Van der Westhuizen, Morland and van Asch2020).
Our work found that the phylogenetic positions of Agliinae, Salassinae, and Eochroini were not clearly distinguished in the BI tree, although the resulting BI and ML trees exhibited highly similar topologies (Figs. 3 and S3). These taxa were distinctly separated in the ML tree, however this distinction lacked well support. The phylogenetic relationships based on ML and BI methods were also constructed using complete mitogenome sequences from 46 Saturniidae species, and the results showed differences with the BI and ML trees constructed based on 13 PCGs (Figs. S6, S7, S8 and S9). Previous study has indicated that different phylogenetic reconstruction methods (i.e. BI vs ML) may influence tree topology, especially with regards to the order of deeper nodes. For example, a study of the high-level phylogeny of the Coleoptera inferred with mitogenome sequences showed that different inference models (ML and BI) can yield inconsistent topologies for the same data (Yuan et al., Reference Yuan, Zhang, Zhang, Guo, Liu, Shen and Shao2016). The incomplete sampling may hinder a well-supported recovery of the tribal structure within this family (Yuan et al., Reference Yuan, Zhang, Zhang, Guo, Liu, Shen and Shao2016).
Our study answered the question of when and how the dispersal of A. polyphemus from Asia into North America might occur. South Asia is the origin centre of the genus Antheraea (Jolly, Reference Jolly1980). The closer relationship between the American species A. polyphemus and the Asian species A. formosana and A. assamensis revealed in our study led us to estimate when the dispersal of A. polyphemus from Asia into North America might occur. In our study, the divergence time between Bombycidae lineage from Saturniidae was estimated to be 44.97 MYA, consistent with previous studies (Balakrishnan et al., Reference Balakrishnan, Dubey, Debnath, Subrahmanyam, Arunkumar, Pandey, Ponnuvel, Sivaprasad and Sathyanarayana2024; Dubey et al., Reference Dubey, Pradeep, Neog, Debnath, Aneesha, Shah, Kamatchi, Ponnuvel, Ramesha, Vijayan, Nongthomba, Bora, Vankadara, VijayaKumari and Arunkumar2024, Kumar et al., Reference Kumar, Stecher, Suleski and Blair2017). The divergence time between A. mylitta and A. assamensis was 23 MYA, and that between A. mylitta, A. yamamai, and A. proylei was 15.7 MYA, which was also consistent with the results by Balakrishnan et al. (Reference Balakrishnan, Dubey, Debnath, Subrahmanyam, Arunkumar, Pandey, Ponnuvel, Sivaprasad and Sathyanarayana2024). The Asian species A. formosana and A. assamensis are endemic to Taiwan and the North-Eastern part of India, respectively, and their divergence time was predicted to be 4.57 MYA (Cheng et al., Reference Cheng, Huang, Cheng, Tseng, Wang, Lai and Hung2022; Singh et al., Reference Singh, Kabiraj, Sharma, Chetia, Mosahari, Neog and Bora2017). The divergence time of A. polyphemus from A. formosana and A. assamensis was estimated to be 18.18 MYA (Fig. 4) during the Miocene Epoch (23–5 MYA) (Jaramillo et al., Reference Jaramillo, Romero, D’Apolito, Bayona, Duarte, Louwye, Escobar, Luque, Carrillo-Briceño, Zapata, Mora, Schouten, Zavada, Harrington, Ortiz and Wesselingh2017). During the Middle Cretaceous to the Eocene (80 to 40 MYA), North America separated from the Eurasian plate (Geological Dictionary Office Ministry of Geology and Mineral Resources, 2005). A previous study has proposed that biotic interchange between Eurasia and North America might have occurred across the Bering land bridge, with three peaks at the end of the Eocene (40–34 MYA), the late Oligocene warming (26–24 MYA), and the early to middle Miocene (16–14 MYA) (Jiang et al., Reference Jiang, Klaus, Zhang, Hillis and Li2019). Hence, we inferred that the dispersal of A. polyphemus ancestry from Asia to North America might occur during the Miocene Epoch (18.18 MYA) across the Berling land bridge.
Conclusion
In this study, the mitogenome of A. polyphemus was determined, which is native to the America and has potential significance for silk production and food consumption. The mitogenome structure of A. polyphemus is described. Our phylogenetic analyses based on the mitogenome confirmed the placement of A. polyphemus within the genus Antheraea and its close relationship with the Asian species A. formosana and A. assamensis. We inferred that the dispersal of A. polyphemus ancestry from Asia to North America might occur during the Miocene Epoch across the Berling land bridge. This study will help us understand the phylogenetic relationships and early evolution history of Saturniidae. Additional mitogenomes are needed to enhance our understanding of the phylogenetic relationships within Saturniidae in the future.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0007485325100278.
Acknowledgements
This work was supported by the National Natural Science Foundation of China grant no: 31972617), the Applied Basic Research Program of Liaoning Province (grant no: 2022JH2/101300178) and the China Agriculture Research System (CARS-18).
Competing interests
The authors declare that no potential conflicts of interest exist.
