key: cord-0754127-rvf3eoy4
authors: Chen, Zi-Shu; Han, Ning-Ning; Li, Jian-Hong; Huang, Guo-Hua; Wan, Hu
title: Selection of reference genes for expression analysis using RT-qPCR in the dissemination system of Heliothis virescens ascovirus 3 h (HvAV-3h)
date: 2017-08-01
journal: Sci Rep
DOI: 10.1038/s41598-017-07684-w
sha: e8be80485fdee7344e970941c62ca23fb3eed4da
doc_id: 754127
cord_uid: rvf3eoy4

Ascoviruses are double-stranded DNA viruses that mainly infect noctuid larvae, and are transmitted by the parasitoid wasp Microplitis similis Lyle. Ascovirus-parasitoids wasp-noctuid insects constitute the dissemination system. Selection of suitable reference genes for the dissemination system could play an important role in elucidating the pathogenic molecular mechanisms of ascovirus. Unfortunately, such studies on potential reference genes in the dissemination system of ascoviruses are lacking. In the present study, we evaluated 11 candidate reference genes: β-actin1 (ACT1), β-actin2 (ACT2), elongation factor 1 (EF1), elongation factor 2 (EF2), ribosomal protein L10 (L10), ribosomal protein L17A (L17A), superoxide dismutase (SOD), 28S ribosome (28S), Tubulin (TUB) and 18S ribosome (18S). The samples were originally from various virus concentrations and points-in-time of experimental treatments using RefFinder and four algorithms. The results showed that EF1 was the most stable internal gene in S. exigua and M. similis and that EF2 was the most stable in the IOZCAS-Spex-II-A cell line, and the stability of reference genes were confirmed via the expression levels of two inhibitor of apoptosis-like (iap-like) genes from Heliothis virescens ascovirus 3 h (HvAV-3h). This study provides a crucial basis for future research that explores the molecular mechanisms of the pathogenesis of ascoviruses.

RNA extraction and cDNA synthesis. Total RNA was isolated according to the Trizol RNA isolation kit manufacturer's protocol. A total of 50 μl DEPC water was used to dissolve the RNA sediment. The A260/A280 ratio and A260/A230 ratio of the RNA were determined using a UV-1800 Spectrophotometer (SHIMADZU) and DU 730 nucleic acid/protein analyzer (BECKMAN COULTER). RNA samples with an A260/A280 ratio ranging from 1.8 to 2.0 and an A260/A230 ratio >2.0 were used for cDNA synthesis 29 . The first-strand complementary DNA was synthesized from 1 μg of total RNA with a PrimeScript RT reagent kit with gDNA Eraser (TaKaRa, Dalian, China) in a total volume of 20 μl. According to the manufacturer's protocol, in the first step, 10 μl mixture was incubated for 2 min at 42 °C, and then 10 μl of master mix was added and incubated for 15 min at 37 °C and 5 s at 85 °C. The cDNA was preserved at −80 °C until further use.

Reference gene selection and primer design. According to a previous study 30 , the PCR primer sequences of S. exigua and the IOZCAS-Spex-II-A cell line were used for quantification of the expression of the genes encoding ACT1, ACT2, EF1, EF2, L10, L17A, SOD, TUB, 18S and 28S, as shown in Table S1 . The primers for M. similis (Table S2) were designed via Beacon Designer 8.0 software with the following settings: primer melting temperature, 60 ± 1 °C; primer GC content, 40-60%; primer length, 18-24 bp; and amplicon length, 100-200 bp. Other parameters were set by default. The lengths of the PCR-amplified specific products and not dimers of PCR products were assessed using gel electrophoresis. A 10-fold dilution series of cDNA was employed as a standard curve, and the reverse-transcription qPCR efficiency was determined for each gene and each treatment using the linear regression model 31 . According to the equation: E = (10 [−1/slope] −1) × 100, the corresponding qRT-PCR efficiencies (E) were calculated 32 . After detecting the efficiencies of the chosen primers, the primers that displayed a coefficient of correlation greater than 0.99 and efficiencies between 90% and 115% were selected for the next qRT-PCR (Table 1) .

Quantitative real-time PCR. Quantitative real-time PCR was performed using SsoFast TM EvaGreen ® Supermix (Bio-Rad, Singapore) via a MyiQ TM 2 Two Color Real-Time PCR Detection System (Bio-Rad). Each reaction was performed in a 20 μl total volume with 10 μl SsoFast TM EvaGreen ® Supermix, 1 μl of cDNA template, 1 μl of 10 μM of each primer and 7 μl of nuclease-free water in an iQ TM 96-well PCR plate (Bio-Rad). The program was set as follows: initial denaturation at 95 °C for 30 s, followed by 40 cycles of 95 °C for 5 s and 60 °C for 10 s. At the end of the reactions, a melting curve analysis from 65 °C to 95 °C was used to ensure amplified product consistency and specificity. All reactions were performed in triplicate.

A stable level of each reference gene was statistically analyzed with four software packages: geNorm 33 , NormFinder 34 , BestKeeper 35 , delta cycle threshold (Ct) method 36 , and Online software RefFinder (freely available at: http://fulxie.0fees.us/?type = reference). Application of the geNorm, NormFinder, and BestKeeper tools was based on the Microsoft Excel program. When geNorm and NormFinder tools performed a stable analysis of the data, the cycle threshold (Ct) was converted into a linear scale (the highest relative quantity for each gene was set to 1). The geNorm algorithm calculated an expression stable value (M) for each gene and then compared the pair-wise variation Vn/Vn + 1. The gene with the lowest M value represented the most stable expression. A ratio of Vn/Vn + 1 below 0.15 indicated that the use of an additional reference gene would not significantly improve normalization 33 . NormFinder combined the interclass variance and intraclass variance to calculate a stable value. The assessment of the reference gene stability was dependent on the size of the stable value 34 . The raw data of cycle threshold (Ct) values (CP values) and PCR efficiency (E) of the reference genes were determined as the best fitted standards by BestKeeper. The cardinal principle for identification of stably expressed reference genes by Bestkeeper was that the expression levels of suitable reference genes should be highly correlated. Therefore, the correlation between each candidate gene and the index was calculated, describing the relation between the index and the contributing candidate reference gene by the highest R value, lowest SD and CV values (<1) and the P value 35 . We also used the online software RefFinder, which integrates the above-mentioned four algorithms (geNorm, Normfinder, BestKeeper, and the delta Ct method) to compare and rank the examined candidate reference genes. According to the results of RefFinder, candidate genes with the top rankings were considered to be the most stably expressed under the tested experimental conditions and thus could be selected as optimal reference genes. Every gene was sorted by the five different statistical approaches separately.

Evaluation the stability of selected reference genes. The third instar larvae molted after 24 h and IOZCAS-Spex-II-A cells were infected with 10 2 -fold concentrations of hemolymph containing the virion of HvAV-3h, and then were harvested at 2 and 3 days p.i. The relative expression levels of inhibit apoptotic-like (iap-like) genes were measured using the most two stable reference genes and least two reference genes via Livak method 37 , respectively.

Selection of candidate reference genes. To investigate the eight commonly used reference genes ACT1, ACT2, EF1, EF2, L10, L17A, SOD, TUB, and 28S from S. exigua and the IOZCAS-Spex-II-A cell line and six reference genes, including 28S, 18S, EF1, TUB, SOD, and ACT, from M. similis (Table 1) , we determined the correlation coefficient (R 2 ) values of all candidates that varied from 0.990 to 0.998 across the cDNA diluted points and, concurrently, the PCR efficiency values of all pair-primers that varied between 90.3% and 114.5% (Table 1) . Expression profiles of candidate reference genes. It is well known that the threshold cycle (Ct) can reflect the expression level of candidate reference genes to a certain extent. In ascovirus-infected M. similis (Fig. 1A) , 18S RNA with a Ct value of 7.92 had the highest expression level and was more fluctuant than the other candidate reference genes. According to the original Ct value of S. exigua (Fig. 1B) , the highest expression reference gene was 28S with a Ct value of 12.44, and the maximal fluctuating amplitude was 6.75. ACT1 was the least variable compared to the other candidate reference genes. In ascoviruses-infected IOZCAS-Spex-II-A cell line samples, the Ct values of the candidate reference genes under the same threshold value for fluorescence ranged from 13.67 for 28S to 27.07 for EF1, which represented the highest and lowest expression levels, respectively. The fluctuation showed no significant difference with each gene (Fig. 1C ).

ranking of the most stable to least stable genes was EF1, SOD, TUB, 28S, ACT and 18S. All four programs identified EF1 as the most stable gene in ascoviruses-infected M. similis samples ( Table 2 ). Based on geNorm analysis, the four genes should not be used as reference genes for normalizing gene expression data for all samples (Fig. 2F ). From the point of view of different ascovirus concentrations, except for the 10 2 -fold treatment, EF1 was the most stable gene according to the geomean of ranking value (Table S4) . Analysis of gene expression stability in ascovirus-infected S. exigua larvae. The stability rankings generated by NormFinder were consistent with those generated by the Delta Ct method and geNorm. However, the gene stability rankings by BestKeeper analysis were different from the other three methods. In all programs except for BestKeeper, EF1, L17A, and EF2, showed the most stable genes (Table 3 ). According to the Geomean of Ranking value by Reffinder, the stability rankings from the most stable to the least stable gene in all ascovirus-infected S. exigua were EF1, L17A, EF2, ACT1, L10, SOD, TUB, ACT2, and 28S (Table 3) . Based on the geNorm algorithm (Fig. 3F) , the gene pair EF2/L17A was the most stably expressed in all samples. Moreover, the inclusion of additional reference genes did not lower the V value below the proposed 0.15 cut-off until the ninth gene was added at 10 0 -and 10 2 -fold dilutions in all samples (Fig. 3G ). From the point of different ascovirus concentrations, L10 was the most stable gene in 10 0 -and 10 8 -fold dilutions, and ACT2 and EF2 were the first positions in 10 2 -and 10 4 -fold dilutions, while L17A was the most stable in the 10 6 -fold dilution (Table S5 ).

The stability rankings generated by the Delta Ct method, NormFinder, and BestKeeper showed that EF2 and L17A were the most stable genes, and gene stability ranked by the Delta Ct method, BestKeeper, and NormFinder were different regarding the results generated by the geNorm method ( Table 4 ). As shown for M value and the optimal number for geNorm, all of the values were far below 1.5 (Fig. 4F) . Individually, the gene pairs ACT1/L10, ACT1/EF2, EF1/TUB and ACT1/EF1 were the most suitable genes in 10 2 -, 10 4 -, 10 6 -and 10 8 -fold dilutions, respectively. EF1/L10 was the best pair across all samples. According to the RefFinder results, the stability rankings from the most stable to the least stable gene in the ascovirus-infected IOZCAS-Spex-II-A cell line samples were as follows: EF2, L17A, ACT2, SOD, EF1, L10, TUB, ACT1 and 28S (Table 4 ). As for different ascovirus concentrations, ACT1 was the most stable gene in the 10 2 -fold dilution and SOD was the most stable in the 10 6 -fold dilution. EF2 was in the first position in the 10 4 -and 10 8 -fold dilutions (Table S6 ).

iap-like2 (Table S3 ) using the two most stable reference genes EF1 and LA17A in the S. exigua were shown in Fig. 5A ,B. Additionally, 28S and ACT2 predicted as the two least stable genes, were applied for normalization to further verify whether the use of unstable reference gene can lead to an inaccurate relative expression (Fig. 5C,D) . At the same time, the results of the relative expression analysis of iap-like1 and iap-like2 using the two most stable reference genes EF2, L17A and the two least stable reference genes 28S and ACT1 in the IOZCAS-Spex-II-A cell line were shown in Fig. 5E -H. In this two samples, the fold changes of the two iap-like genes normalized with stable reference gene showed consistent results.

Ascoviruses are insect-specific double-stranded circle DNA viruses that attack lepidopterans, most commonly species in the family Noctuidae 6 . HvAV-3h has been recently isolated from S. exigua 4 . Li found that the early instars of S. exigua were significantly easier to infect with HvAV-3h compared to the later instars, using 10-fold serial dilutions (0 to 7) of HvAV-3h-containing hemolymph to infect S. litura larvae. There were no significant differences in larval mortalities from 10 0 -to 10 3 -fold dilutions; however, significant declines were observed at the 10 4 -fold dilution and above 7 . Compared to the healthy larval population, the typical symptoms and survival times of the diseased larval population were considerably extended, food intake was significantly reduced, and the body weight remained fairly constant in the 3 rd and 4 th instar larvae, which happened after inoculation with the ascovirus. However, the corrected mortality rates for the 1 st through 5 th instar inoculated per os were very low 8 . Therefore, the ascovirus mode of dissemination, which relies on the parasitoid wasp M. similis, served as a vector when the female parasitoid wasp acquired the virus. The ovipositor with the virion viability was 4.1 ± 1.4 days, and infected host larvae were still acceptable for egg laying by parasitoids. The parasitoids thereafter transmitted the virus to healthy hosts 26 . It follows that ascoviruses, with this dissemination system, have probably been regarded as significant for long-term pest control. Moreover, there are few studies that have elucidated the selection of reference genes for ascovirus infection dissemination systems. Therefore, stable expression of reference gene (s) is important for understanding the molecular mechanism of rapid pathogenesis and chronic death. RT-qPCR is now the most sensitive method to study low-abundance mRNA from various tissue samples and experimental conditions. Thus, it is necessary to precisely determine normalization strategies 38 . Previous studies demonstrated that when reference genes were selected, the geometric mean of multiples should be used to ensure more accurate results 33 .

Spodoptera exigua would lead to several disordered phenomena due to the viral infection of S. exigua. There is no knowledge on the molecular mechanism. Thus, to gain a clear understanding of the pathogenic mechanisms, the screening of reference genes for S. exigua and different concentrations of ascovirus has provided a foundation for future research. At present, according to differences between physiological stages and different tissues in S. exigua, their suitability as reference genes was diverse with each treatment. In general, SOD, ACT2, ACT1, EF1 and GAPDH were stably expressed in all developmental stage sample sets. L10, EF2, L17A and EF1 were ranked highest in all tissue sample sets 30, 39 . The expression of these internal genes is considerably different in different experimental conditions. We therefore reassessed the stability before RT-qPCR testing. Our results indicated that EF1 ranked hi4ghest in all sample sets by RefFinder, Delta Ct and Normfinder, whereas the BestKeeper method ranked ACT1 as the best reference gene and EF1 ranked at the fifth position. The subtle differences in ranking among the top order reference genes could be imputed to differences in algorithms of the employed software programs and sensitivities towards the co-regulated reference genes.

A previous study showed that after the parasitoid M. similis possessed the virus, the virus just stayed in the ovipositor, and the virus could only be spread in a mechanical pathway 26 . Therefore, we inferred that the parasitoid possession of the virus should barely affect the transcription factors. In most studies of Hymenoptera, 18S or ACT has been commonly employed as the reference gene [40] [41] [42] . At the same time, none of the studies contributed to a comprehensive selection of internal control genes for Hymenoptera (Braconidae). In our study, although 18S RNA had the highest expression of all the candidate reference genes, it was not the most stable. When the parasitic wasp carried the virus, the most stable reference gene was EF1 in all sample sets by RefFinder, Delta Ct, geNorm and Normfinder. In this regard, the result was different from other studies that used 18S RNA as an internal gene.

HvAV-3e was replicated in three noctuid cell lines from Sf9 and Helicoverpa zea (BCIRL-Hz-AM1 and FB33). However, HvAV-3e did not replicate in the Pieris rapae (Pieridae) cell line, which was non-noctuid 43 . This means that ascoviruses were likely to impact the IOZCAS-Spex-II-A cell line, which was derived from the fat body of S. exigua. Consequently, stable reference genes can be helpful in further research on the cytopathic effect of ascoviruses. This study showed that the IOZCAS-Spex-II-A cell line was susceptible to infection by ascoviruses. AcMNPV could also easily infect the IOZCAS-Spex-II-A cell line 16, 44 . It has been reported that with the selection of reference genes in the Sf21 cell line infected by ascoviruses, DNA-free RNA was used as a template with a combination of the Sf21 cell line 28S gene-specific reverse primer and the oligo-dT primer for first strand cDNA synthesis. The results indicated that the Ct values were significantly higher and more variable during the course of viral infection when only the oligo-dT primer was used in the cDNA synthesis step than when the 28S-R primer in conjunction with the oligo-dT primer was used 16 , although the stability of the reference genes was not analyzed by the geNorm, NormFinder, BestKeeper, and delta cycle threshold (Ct) method and Online software RefFinder. In our study, we applied the conventional synthesis method of cDNA using RefFinder, Delta Ct, geNorm, Normfinder and BestKeeper algorithms to analyze the Ct values. As a consequence, the results indicated that 28S was the least suitable gene in the IOZCAS-Spex-II-A cell line across all samples, while the EF2 gene was expressed most stably in comparison with 8 other candidate internal genes. Generally, the most stable reference gene of the IOZCAS-Spex-II-A cell line and S. exigua should provide the same results. We acquired EF2 and L17A, which were relatively stable in screening of the reference genes for S. exigua and the IOZCAS-Spex -II-A cell line. However, in S. exigua, EF1 was the top-ranked by RefFinder, Delta Ct and NormFinder but was in the fifth positon by BestKeeper and the third position by geNorm. This phenomenon was probably due to the virus being able to directly impact the cell line; however, virus attack of insects may be influenced by other factors such as pH values, environmental temperature and host larvae with different instars.

Pairwise variation analysis with the geNorm applet suggested the use of two or more reference genes for attaining better accuracy in normalization for most of the experimental conditions 30 . The gene pairs EF2/L17A and EF1/L10 were considered the most suitable pairs of genes to normalize samples in S. exigua and the IOZCAS-Spex-II-A cell line, respectively, across all samples. However, conditions such as changes in M. similis would require normalization by three or more reference genes because the values of pairwise variations were above the cut-off range of 0.15. Thus, across all of the samples in S. exigua and the IOZCAS-Spex-II-A cell line, we advise the use of two reference genes under different experimental conditions.

To verify stability of candidate reference genes predicted by the RefFinder and four other algorithms, the most stable and least stable genes were applied for normalization the two IAP genes. Based on the sequence similarity to known IAPs, the iap-like genes have been chosen as the target genes. In the genomes of HvAV-3h, IAP genes have evolved mechanisms to reduce formation of apoptosis to guarantee the propagation of HvAV in host cells 45 . When the most stable reference genes (EF1, L17A and EF2) were employed to calibrated the data of gene expression in the S. exigua and IOZCAS-Spex-II-A cell line, the expression levels of the two iap-like genes revealed no significant changes. However, using the least stable reference genes 28S, ACT1and ACT2 to analyze the expression levels of the two iap-like genes, the results showed significant variation between two calculations which used the least stable reference genes. Therefore, it is necessary to use an appropriate stable reference gene for calibration of gene expression.

In summary, keeping in view the ecological control importance of the HvAV-3h-parasitic wasp (M. similis)-insect (S. exigua) dissemination system and pathogenic molecular mechanisms of ascoviruses, gene expression studies should continue to constitute a meaningful part of basic research with ascoviruses. Hence, establishing a best reference gene for RT-qPCR in dissemination systems will benefit researchers in this research arena. To the best of our knowledge, this is the first comprehensive report on the identification and validation of optimal candidate reference genes for accurate transcript normalization of gene expression in studies using RT-qPCR in the dissemination system of ascoviruses under various experimental conditions. We recommend the use of EF1 in S. exigua and M. similis and EF2 in the IOZCAS-Spex-II-A cell line. This study offers a way forward for the study of the pathogenic molecular mechanism of ascoviruses.

Evidence for the evolution of ascoviruses from iridoviruses

Sequence and organization of the Heliothis virescens ascovirus genome

Circular configuration of the genome of ascoviruses

Phylogenetic position and replication kinetics of Heliothis virescens ascovirus 3h (HvAV-3h) isolated from Spodoptera exigua

Genome analysis of Heliothis virescens ascovirus 3h isolated from China

A viral caspase contributes to modified apoptosis for virus transmission

A comparison of growth and development of three major agricultural insect pests infected with Heliothis virescens ascovirus 3h (HvAV-3h)

Characterization and Growing Development of Spodoptera exigua (Lepidoptera: Noctuidae) larvae Infected by Heliothis virescens ascovirus 3h (HvAV-3h)

The power of real-time PCR. Advances in physiology education

RNA integrity and the effect on the real-time qRT-PCR performance. Molecular aspects of medicine

Real-time RT-PCR normalisation; strategies and considerations

Selection of ovine housekeeping genes for normalisation by real-time RT-PCR; analysis of PrP gene expression and genetic susceptibility to scrapie

Validation of housekeeping genes for normalizing RNA expression in real-time PCR

Validation of housekeeping genes as internal control for studying gene expression in rice by quantitative real-time PCR

Reference gene selection for quantitative real-time PCR analysis in virus infected cells: SARS corona virus, Yellow fever virus, Human Herpesvirus-6, Camelpox virus and Cytomegalovirus infections

Verifying the stability of selected genes for normalization in q-PCR experiments of Spodoptera frugiperda cells during AcMNPV infection

Strategy of the use of 28S rRNA as a housekeeping gene in real-time quantitative PCR analysis of gene transcription in insect cells infected by viruses

Analysis of reference gene stability after Israeli acute paralysis virus infection in bumblebees Bombus terrestris

Guideline to reference gene selection for quantitative real-time PCR

Identification and expression pattern of putative odorant-binding proteins and chemosensory proteins in antennae of the Microplitis mediator (Hymenoptera: Braconidae)

IAP family proteins-suppressors of apoptosis

An apoptosis-inhibiting baculovirus gene with a zinc finger-like motif

Control of programmed cell death by the baculovirus genes p35 and iap

Influence of larval diets on development, fecundity and flight capacity of the beetarmyworm, Spodoptera exigua

Development of Microplitis similis (Hymenoptera: Braconidae) on two candidate host species, Spodoptera litura and Spodoptera exigua (Lepidoptera: Noctuidae)

Imperfection works: Survival, transmission and persistence in the system of Heliothis virescens ascovirus 3h (HvAV-3h), Microplitis similis and Spodoptera exigua

simple method of estimating fifty per cent endpoints

Transmission of ascovirus from Heliothis virescens (Lepidoptera: Noctuidae) by three parasitoids and effects of virus on survival of parasitoid Cardiochiles nigriceps (Hymenoptera: Braconidae)

Identification and Validation of Reference Genes for the Normalization of Gene Expression Data in qRT-PCR Analysis in Aphis gossypii (Hemiptera: Aphididae)

Selection and evaluation of reference genes for expression analysis using qRT-PCR in the beet armyworm Spodoptera exigua (Hübner) (Lepidoptera: Noctuidae)

A new mathematical model for relative quantification in real-time RT-PCR

Guideline to reference gene selection for quantitative real-time PCR

Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes

Normalization of real-time quantitative reverse transcription-PCR data: a model-based variance estimation approach to identify genes suited for normalization, applied to bladder and colon cancer data sets

Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper-Excel-based tool using pair-wise correlations

Selection of housekeeping genes for gene expression studies in human reticulocytes using real-time PCR

Analysis of relative gene expression data using real-time quantitative PCR and the 2 −ΔΔCT method

Absolute quantification of mRNA using real-time reverse transcription polymerase chain reaction assays

Validation of reference genes for quantitative expression analysis by real-time RT-PCR in four lepidopteran insects

Molecular characterization of the gene feminizer in the stingless bee Melipona interrupta (Hymenoptera: Apidae) reveals association to sex and caste development

Molecular characterization and expression profiles of olfactory receptor genes in the parasitic wasp, Microplitis mediator (Hymenoptera: Braconidae)

Transcript analysis and expression profiling of three heat shock protein 70 genes in the ectoparasitoid Habrobracon hebetor (Hymenoptera: Braconidae)

Replication of Heliothis virescens ascovirus in insect cell lines

A cell strain cloned from Spodoptera exigua cell line (IOZCAS-Spex-II) highly susceptible to S. exigua nucleopolyhedrovirus infection

Inhibition of apoptosis by Heliothis virescens ascovirus (HvAV-3e): characterization of orf28 with structural similarity to inhibitor of apoptosis proteins

Supplementary information accompanies this paper at doi:10.1038/s41598-017-07684-w Competing Interests: The authors declare that they have no competing interests.Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.