Expressional divergence of the fatty acid-amino acid conjugate-hydrolyzing aminoacylase 1 (L-ACY-1) in Helicoverpa armigera and Helicoverpa assulta

How FACs-producing generalist and specialist herbivores regulate their FACs-hydrolyzing enzyme L-ACY-1 to balance FACs’ beneficial vs. detrimental effects remains unknown. To address this question, we compared L-ACY-1 expression in Helicoverpa armigera and Helicoverpa assulta, a pair of closely related sibling species differing mainly in their host range, by the same sets of hostplants, protein to digestible carbohydrate (P:C) ratios, or allelochemical. L-ACY-1 expression remained low/unchanged in H. armigera, but was induced by hot pepper fruits and repressed by cotton bolls in H. assulta. The representative allelochemicals of the tested hostplants significantly (capsaicin) or insignificantly (gossypol and nicotine) induced L-ACY-1 expression in H. armigera, but insignificantly inhibited (capsaicin and gossypol) or induced (nicotine) it in H. assulta. L-ACY-1 expression remained low/unaltered on balanced (P50:C50 and P53:C47) or protein-biased diets and induced on carbohydrate-biased diets in H. armigera, but was at the highest level on balanced diets and reduced on either protein- or carbohydrate-biased diets in H. assulta. Furthermore, L-ACY-1 expression was significantly higher in H. assulta than in H. armigera for most of feeding treatments. Such expressional divergences suggest that FACs are utilized mainly for removal of excessive nitrogen in generalists but for nitrogen assimilation in specialists.


Results
L-ACY-1 cDNA in H. assulta. The full-length cDNA sequence of L-ACY-1 from H. assulta (NCBI accession number KY348705) was obtained by RT-PCR cloning of its middle fragment (1199 bp) with a pair of primers (Table 1) designed based on H. armigera L-ACY-1, followed by cloning of its 5′ and 3′ ends by 5′ and 3′ RACE (random amplification of cDNA ends), and cloning of its full-length sequence with the two full-length primers ( Table 1) designed based on its 5′ and 3′ ends. The H. assulta L-ACY-1 is 1758 bp long, containing a 30 bp 5′ UTR (5′ untranslated region), an ORF (open reading frame) of 1314 bp (from 31 bp to 1344 bp) encoding 437 amino acids, and a 414 bp 3′ UTR (Fig. 1). The deduced H. assulta L-ACY-1 protein shares 95% amino acid identity with H. armigera L-ACY-1, 96% with H. zea L-ACY-1, 88% with H. virescens L-ACY-1, 69% with S. frugiperda L-ACY-1, 56% with Plutella xylostella L-ACY-1, 63% with Amyelois transitella L-ACY-1. Phylogenetic analysis reveals that H. assulta L-ACY-1 belongs to the Noctuidae lineage and is the immediate common ancestor of H. armigera and H. zea L-ACY-1 (Fig. 2), consistent with the evolutionary relationship among the 3 Helicoverpa species 23,28,29 . Like other aminoacylases, H. assulta L-ACY-1 has a dinuclear zinc-binding site, which is composed of His96 Asp130, Glu164 Glu165 Asp207 and His392 (circled amino acids in Fig. 1). RT-PCR gel analysis showed that when the larvae of both species were reared on the same wheat germ-containing artificial diets [i.e. the control diet, 53 protein (P): 47 carbohydrate (C)] 30 , L-ACY-1 transcripts were about 3.89 fold higher in the specialist H. assulta than its generalist counterpart H. armigera (Fig. 3), consistent with the qRT-PCR data shown in Fig. 4 (see control). When the newly-molted final instar larvae of both species were transferred onto detached tobacco leaves (common host plants for both species), cotton bolls (cut in half; favored host plant for H. armigera but poor for H. assulta), or hot pepper fruits (favored host plant for H. assulta but poor for H. armigera) and reared for 48 h, L-ACY-1 expression remained unchanged relative to the larvae fed on the control diets in H. armigera (Fig. 4). By contrast, L-ACY-1 expression in H. assulta was very sensitive to hostplant species; cotton bolls significantly repressed L-ACY-1 expression, whereas pepper fruits significantly induced it (Fig. 4); tobacco leaves did not cause significant changes in L-ACY-1 expression. Cross-species comparison revealed that L-ACY-1 expression was significantly higher in H. assulta than in H. armigera in both the control diets and the three tested hostplants, with a H. assulta / H. armigera ratio ranged from 1.77 (cotton bolls) to 5.21 (tobacco leaves), 3.90 (control diets) or 5.62 fold (hot pepper fruits) (Fig. 4).   Effects of plant allelochemicals on L-ACY-1 expression in H. armigera and H. assulta. To determine whether allelochemicals present in their common (capsaicin in hot pepper and nicotine in tobacco) or unique (gossypol in cotton for H. armigera) natural hostplants account for their differential regulation of L-ACY-1 expression in response to the three hostplants, we subjected the newly-molted final instar larvae of both species to control diets or diets supplemented with 0.1% capsaicin, nicotine, or gossypol for 48 h and then analyzed the expression of L-ACY-1 in the midguts of these larvae by qRT-PCR. Two-way ANOVA showed significant impacts of insect species (Df = 1, F = 61.7, P = 0.00), allelochemicals (Df = 3, F = 4.4, P = 0.02) and their interactions (Df = 3, F = 12.1, P = 0.00) on L-ACY-1 expression. In H. armigera, capsaicin (3.58-flod) significantly induced the expression of L-ACY-1, gossypol (1.85-flod) and nicotine (1.28 fold) slightly but not significantly increased the expression of L-ACY-1 (Fig. 5). In H. assulta, nicotine (1.44-flod), slightly induced the expression of L-ACY-1, whereas capsaicin (0.90-flod) and gossypol (0.84-flod) slightly inhibited the expression of L-ACY-1 (Fig. 5). Cross-species comparison revealed that L-ACY-1 expression was significantly higher in H. assulta than in H. armigera when their larvae were fed on control diets (3.90 fold) or nicotine-containing diets (4.35-flod). When their larvae were fed on diets supplemented with capsaicin (p = 0.823) or gossypol (p = 0.141), the expression of L-ACY-1 in H. assulta was similar to or insignificantly higher than that in H. armigera (Fig. 5).   Values sharing the same letter [small case letters for H. armigera, capital letters for H. assulta] are not significantly different at p < 0.05 (Tukey's HSD tests). Significant differences at each of the four treatments between the two species are indicated by one star (*p < 0.05, independent t-test) and two stars (**p < 0.01, independent t-test), respectively.

Discussion
FACs play a beneficial role in nitrogen assimilation 12 of lepidopteran herbivores while harming them by eliciting direct and indirect plant defenses [1][2][3][4][5] . Excretion of FACs into the oral secretions and frass of these herbivores 13 indicate that FACs also help them to remove overeaten nitrogen, which is toxic to insects and thus must be excreted. Because generalist and specialist herbivores differ in their capacity to acquire macronutrients such as proteins and carbohydrates and to counteract plant defenses, they are expected to differentially regulate the Values sharing the same letter [small case letters for H. armigera, capital letters for H. assulta] are not significantly different at p < 0.05 (Tukey's HSD tests). Significant differences at each of the four treatments between the two species are indicated by one star (*p < 0.05, independent t-test) and two stars (**p < 0.01, independent t-test), respectively.

Figure 6.
Effects of protein to digestible carbohydrate (P:C) ratios on the expression of L-ACY-1 in H. armigera and H. assulta. The data and error bars represent the means and standard deviations of three biological replicates of three technical repeats each. Values sharing the same letter [small case letters for H. armigera, capital letters for H. assulta] are not significantly different at p < 0.05 (Tukey's HSD tests). Significant differences at each of the five treatments between the two species are indicated by one star (*p < 0.05, independent t-test) and two stars (**p < 0.01, independent t-test), respectively. expression of L-ACY-1 to adjust FACs amount in their oral secretions and frass and thus balance the above three conflicting roles (nitrogen assimilation, nitrogen removal, and triggering plant defenses) of FACs.
To test this general postulation, we compared the expression profiles of the FACs-decomposing L-ACY-1 in H. armigera and H. assulta final instar larvae fed with hostplant tissues, diets with different allelochemicals, or diets with different P:C ratios. Consistent with the reverse relationship between L-ACY-1 expression and host range from studies of H. virescens, H. subflexa and H. zea maintained on a common artificial diet 13,20 , L-ACY-1 transcription was significantly greater in the specialist H. assulta than in the generalist H. armigera when they were fed on the common near-balanced control diet (Figs 3 and 4). Such a reverse relationship held true (Fig. 4) even we offered their larvae with tissues of cotton (bolls), a preferred hostplant for H. armigera but a non-host for H. assulta, tobacco (leaves), a hostplant commonly utilized by both species, and hot pepper (fruits), a common host for H. assulta but an occasional host for H. armigera 25,[31][32][33] . That being said, the between-species L-ACY-1 transcriptional difference (H. assulta / H. armigera) increased in the larvae fed with hot pepper fruits, but markedly reduced in those fed with cotton bolls, relative to that of the larvae on the control diet (Fig. 4). This host-dependent variation was largely resulted from significant induction and suppression of L-ACY-1 transcription in the specialist H. assulta but not in the generalist H. armigera by hot pepper fruits and cotton bolls, respectively.
Plant tissues such as tobacco leaves, hot pepper fruits and cotton bolls used in this study are usually inferior to artificial diets because they are not only nutritionally less adequate and balanced than the later 34-37 , but also contain plant defense end products such as allelochemicals 2, 31, 38 . The plant-dependent significant induction or repression of L-ACY-1 in H. assulta but lower and stable expression of L-ACY-1 in H. armigera in response to the same set of plant tissues (Fig. 4) implies L-ACY-1 is differentially regulated by nutrient imbalance and plant defense end products in the two species. In agreement with this inference, gossypol and nicotine, the two representative allelochemicals found in cotton and tobacco respectively 31,39,40 , the preferred or commonly-used host plants of the generalist H. armigera, insignificantly induced L-ACY-1 expression in H. armigera but marginally repressed (gossypol) or elevated (nicotine) L-ACY-1 expression in H. assulta (Fig. 5). Capsaicin, a representative allelochemical found in H. armigera's poor host hot pepper, significantly induced L-ACY-1 expression in H. armigera but marginally inhibited it in H. assulta. The aforementioned species-specific opposite effects of capsaicin and gossypol rendered L-ACY-1 expression level in H. assulta equal to that in H. armigera.
Overall, L-ACY-1 transcription was more sensitive to allelochemicals in H. armigera than in H. assulta. The insignificant (gossypol and nicotine) or significant (capsaicin) induction of L-ACY-1 in the generalist H. armigera is an intuitively expected self-protection response by reducing the FACs amount in its oral secretion and frass and thus minimizing elicitation of plant defenses. Induction of L-ACY-1 may also increase use of FACs for nitrogen assimilation and thus protect this species via reduced allelochemical exposure by eating less food. By contrast, the lack of L-ACY-1 induction in the specialist H. assulta by capsaicin is probably because H. assulta uses capsaicin as a feeding stimulant 41 and is much more tolerant to capsaicin than H. armigera 42 . That H. assulta does not naturally encounter gossypol in its restricted range of hostplants may explain why this allelochemical failed to induce L-ACY-1 expression in this specialist. The overall lower sensitivity of H. assulta L-ACY-1 expression to allelochemicals is probably because specialist detoxification enzymes are usually less inducible by allelochemicals and degrade the allelochemicals present in the hostplants of a given specialist at a much higher efficiency than do the corresponding generalist homologs 22,43 .
Contrary to the allelochemical induction profiles of L-ACY-1 in the two Helicoverpa species, L-ACY-1 transcription was overall more sensitive to macronutrient (protein and carbohydrates) imbalance in the specialist H. assulta than in the generalist H. armigera (Fig. 6). On the one hand, L-ACY-1 expression level in H. armigera remained low when P:C ratio ranged from the relatively balanced (50:50 and 53:47) to high P:C ratio diets (80:20 and 90:10), and significantly increased only when undereating proteins on a low P:C (35:65) ratio diet (Fig. 6). On the other hand, L-ACY-1 expression in H. assulta was at the highest level on the two balanced diets (50:50 and 53:47), but significantly repressed when overeating either low (35:65) or high (70:20 and 90:10) P:C diets (Fig. 6). The fact that the plant tissue-dependent expression patterns of L-ACY-1 in the two species are more similar to those elicited by the P:C ratio diets-lower and less sensitive in H. armigera than H. assulta-suggests that dietary imbalance in the two major macronutrients contributes more to the cross-species difference in the plant tissue-elicited L-ACY-1 expression than allelochemicals.
Our data also suggest that nutritional imbalance plays a more important role than do allelochemicals in determining the inverse relationship between host-range and L-ACY-1 expression on artificial diets and hostplants. Although the self-selected balanced P:C ratio (intake target) of lepidopteran herbivores ranges from 0.8-1.6 (1.1 to 1.6 for generalists and 0.8 to 1.1 for specialists) 44,45 , herbivores seldom encounter a plant species or tissue that is perfectly balanced nutritionally. Instead, they are frequently forced to overingest protein on a high-protein hostplant or carbohydrate on a high-carbohydrate plant tissue. Correlated with their host range difference, generalists are capable of overeating larger nutrient excesses (especially protein) than specialists since the former naturally experience a greater degree of nutritional heterogeneity than do specialists [46][47][48] . While overconsumed carbohydrates can be converted into lipids and stored inside the body, the nitrogen excesses from overingested proteins must be excreted because it is toxic to insect herbivores. Because of their expected and documented greater ability to overeat surplus quantities of excessive protein nutrients [46][47][48] , generalists may often utilize FACs for removal of excessive nitrogen, rather than nitrogen assimilation. By contrast, it is probably not easy for specialists to get enough protein nutrients from their narrow range of hostplants. Thus, specialists may use FACs mainly for nitrogen assimilation, which requires hydrolysis of FACs by L-ACY-1.
The divergences in the relative importance of the three conflicting functions (nitrogen assimilation, nitrogen removal, and eliciting plant defenses) of FACs and thus of L-ACY-1 in specialists and generalists may account for the inverse relationship between L-ACY-1 expression level and host range 13,20 (Figs 3-5) and the differential non-linear expression patterns of L-ACY-1 caused by P:C ratios (Fig. 6). For generalists such as H. armigera, SCIEntIfIC REpoRts | 7: 8721 | DOI:10.1038/s41598-017-09185-2 which are readily to obtain enough or even surplus nutrients from a wider host range [46][47][48] and have a more powerful detoxification system to overcome the diversity and unpredictability of plant defenses 22 , nitrogen removal is expected to be the most important function of FACs/L-ACY-1, followed by nitrogen assimilation, and FACs decomposition to avoid elicitation of plant defenses. The lower L-ACY-1 expression of H. armigera larvae on relatively balanced (50:50 and 53:47) to high P:C ratio diets (80:20 and 90:10) (Fig. 6) allows this species to excrete at least part of the overingested nitrogen in the form of FACs in its frass and oral secretions 13 . The significantly enhanced L-ACY-1 expression when undereating proteins on a low P:C (35:65) diet (Fig. 6), on the other hand, increases this species' ability to assimilate nitrogen by breaking down FACs into amino acids and lipids 12 .
By contrast, because of a narrower host range (thus greater difficulty to get enough nutrients) and a weaker detoxification system (thus less capacity to afford FACs elicitation of plant defenses), nitrogen assimilation becomes the most important function for specialists such as H. assulta, followed by FACs decomposition to avoid elicitation of plant defenses, and nitrogen removal. This explains why L-ACY-1 expression remains at the highest level in the near-optimal 55:47 and 50:50 P:C diets, marginally reduce at the low P:C ratio diet (hydrolyzing FACs to avoid elicitation of plant defense and nitrogen assimilation), and significantly decrease at the high P:C ratio diets (nitrogen removal). Another possible reason for the marginal reduction of L-ACY-1 expression in the low protein diet would be that plant tissues with lower P:C ratio often contain certain allelochemicals that are hard for specialists to degrade and thus H. assulta must lower L-ACY-1 expression to avoid triggering production of the corresponding allelochemicals by plants. That cotton bolls, which have a P:C ratio of 0.41-1.17 45 and contain gossypol, significantly repressed L-ACY-1 expression in H. assulta but not in H. armigera (Fig. 4), supports this speculation.

Materials and Methods
Insect. The  Feeding experiments and preparation of midguts. Three different feeding experiments were conducted for both species with 3 biological replicates of 10 newly-molted final instar larvae each for all the treatments. In Experiment 1, we fed 30 newly-molted final instar larvae of each species with control diets, cotton bolls (cut in half; variety 'Shiyuan 321'), leaves of 15-25 cm-tall tobacco plants (variety 'Yabuli') or fruits of hot pepper (variety 'Xiaoxin 9') collected from the corresponding plants grew in the open field in Langfang experimental station. In Experiment 2, we transferred 30 newly-molted larvae of each species from the control diet onto control diets or diets supplemented with 0.1% (W/W) of gossypol, capsaicin (Aimeida Company, Beijing) or nicotine (Sigma, USA). In Experiment 3, we transferred 30 newly-molted larvae of each species from the control diet (53 P:47 C) onto control diets or diets with a P:C ratio of 35:65, 50:50, 80:20 or 90:10. After 48 h of the feeding treatments mentioned above, we cooled the 30 larvae of each treatment on ice for about 10 mins, cut off the front and back ends of each larva between 2 nd and 3 rd thorax and immediately posterior to the last proleg segment, respectively, pulled out their midguts, open them to remove midgut contents, washed them in cold ddH 2 O, pooled them into 3 biological replicates of 10 midguts each, flash-frozen with liquid nitrogen, and stored in −80 °C for subsequent RNA extraction and analyses of L-ACY-1 transcripts.
RNA extraction and cDNA synthesis. Each replicate of 10 midguts were grinded to fine power in liquid nitrogen and transferred to a 1.5 ml centrifuge tube containing 600 µL of guanidinium-HCl RNA extraction buffer. Total RNA was extracted from the powder using the procedure described in Li et al. 49 , treated with DNase I (Promega, USA) and RNase inhibitor (Thermo, USA) for 40 min to remove potential genomic DNA (gDNA), purified by phenol/chloroform extraction and ethanol precipitation, and dissolved in 500-800 µL DEPC H 2 O. Total RNA samples were stored in −80 °C for subsequent RT-PCR cloning of H. assulta L-ACY-1 gene and/or RT-PCR and qRT-PCR analyses of L-ACY-1 expression.

RT-PCR cloning and sequence analysis of H. assulta L-ACY-1 gene.
Cloning of the full-length cDNA sequence of H. assulta L-ACY-1 began with RT-PCR amplification of its central region (1199 bp) with a pair of partial cloning primers (Table 1) designed based on H. armigera L-ACY-1 (GenBank: JF922297.1), followed by 3′ rapid amplification of cDNA end (RACE) with the universal primer mix (UPM in the Clontech SMARTer ® RACE 5′/3′ Kit) and the gene-specific primer H. ass-ACY-5′ to amplify its 3′ end, 5′ RACE with UPM and the gene-specific primer H. ass-ACY-3′ (Table 1) to amplify its 5′ end, and amplification of its full-length cDNA with a pair of full-length cloning primers ( Table 1). The initial partial PCR reaction (20 µL) was composed of 0.1 µL diluted H. assulta cDNA prepared above, 2 µL 10 × Ex Taq buffer, 0.1 µL Ex Taq (TaKaRa, Japan), 1.6 µL dNTP (2.5 nM), 200 pmol of the forward and reverse primer. The partial PCR reaction was initiated with 5 min denaturation at 95 °C, followed by 40 cycles of 30 s denaturation at 95 °C, 30 s annealing at 55 °C and 1.5 min extension at 72 °C, and a final 10 min extension at 72 °C on an Eppendorf Master Cycler.
Two µg of H. assulta total RNA, together with 1 µl 5′-RACE CDS Primer A and 1 µl SMARTer IIA oligo A or 1 µl 3′-RACE CDS Primer A were used to synthesize the 5′ or 3′ RACE-ready cDNAs with the SMART RACE cDNA Amplification kit (Clontech, USA) according to the manufacturer manual. The resultant 5′ or 3′ RACE-ready cDNA was used as the template to amplify the 5′ or 3′ end of H. assulta L-ACY-1 with UPM and the gene-specific primer H. ass-ACY-5′ (5′ RACE) or H. ass-ACY-3′ (3′RACE) following the setup and touchdown cycling procedure described in the manufacturer manual.

RT-PCR and qRT-PCR.
Two µg of each total RNA sample were reverse transcribed into cDNA at 42 °C for 1 h in a 20 µL reaction with 4 µL primer Mix (Tiangen, China), 4 µL dNTP (TaKaRa, Japan), 1 µL M-MuLV reverse transcriptase (New England Biolab) and 1.5 µL RNase inhibitor (Thermo, USA). The cDNAs were diluted 10-fold and used as the templates for RT-PCR gel analyses of L-ACY-1 and EF-1a (as an internal reference gene) and qRT-PCR analyses of L-ACY-1, Ribosomal protein L-13 (RPL-13), ribosomal protein L-32 (RPL-32) and β-Tubulin in the two Helicoverpa species. The primer pairs for qRT-PCR analyses of the three reference genes (RPL-13, RPL-32 and β-Tubulin) were from Songdou et al. 50 (Table 2). The primer pairs for RT-PCR gel analyses of L-ACY-1 and EF-1a (Table 3) and for qRT-PCR analyses of L-ACY-1 ( Fig. 7 and Table 2) were designed with Primer Premier 5. All the primers satisfy the standard of MIQE.
SCIEntIfIC REpoRts | 7: 8721 | DOI:10.1038/s41598-017-09185-2 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA). The cycling program was consisted of an initial denaturation at 95 °C for 15 min, followed by 40 cycles of denaturation at 95 °C for 15 s, annealing at 60 °C for 30 s and extension at 72 °C for 32 s, during which data collection and real-time analysis enabled. Melting curve analysis was performed from 65 °C to 95 °C for L-ACY-1 and the three reference genes to ensure no interference of junk products. Each biological replicate was qRT-PCR-analyzed in triplicate. Amplification efficiency (E) of each gene was determined from the slope of the log template concentration (x-axis) -Ct value (y-axis) line, using the formula E = 10 −1/slope − 1 51 . The expression levels of L-ACY-1 and the three reference genes in the larval midguts from different feeding treatments of the two Helicoverpa species were calculated with their mean Ct and amplification efficiency (E) (Equation 1). The expression levels of H. armigera and H. assulta L-ACY-1 were further normalized with the geometric mean of the expression of the three reference genes (RPL-13, RPL-32 and β-Tubulin) (Equation 2).
( Statistical analysis. Two-way ANOVA was performed to examine the impacts of species, feeding treatment and species × feeding treatment interaction on the expressions of L-ACY-1. Independent t-tests were conducted to test the significance of differences in the expression of L-ACY-1 for all treatments between the two Helicoverpa species. Tukey's HSD tests were performed to test the significance of differences in the expression levels of L-ACY-1 among host plants, diets with different P:C ratio or plant allelochemicals within each species. All of the statistical tests were performed by SPSS version 19.0 software.