Characters and expression of the gene encoding DH, PBAN and other FXPRLamide family neuropeptides in Antheraea pernyi


Z.-J. Wei (corresponding author), Department of Biotechnology, Hefei University of Technology, Hefei 230009, China. E-mail:


The full-length cDNA encoding diapause hormone (DH) and pheromone biosynthesis activating neuropeptide (PBAN) in Antheraea pernyi (Anp-DH-PBAN) were cloned and sequenced by rapid amplification of cDNA ends methods. The Anp-DH-PBAN cDNA encodes a 196-amino acid (aa) prehormone that contains a 24-aa DH-like peptide, a 33-aa PBAN and three other neuropeptides, all of which share a common C-terminal pentapeptide motif FXPR/KL (X = G, T, S). The Anp-DH-PBAN shows highest homology (82%) to that of Samia cynthia ricini at amino acid level. Northern blots demonstrate the presence of a 0.8-kb transcript in the brain and suboesophageal ganglion (SG) complex. During the early pupal stages, the Anp-DH-PBAN mRNA contents increase consistently in both diapause- and non-diapause-destined pupae. From day 7, mRNA in diapause pupae dropped promptly and maintained a lower level than that in the non-diapause type. The Anp-DH-PBAN mRNA was expressed mainly in SG, at much lower levels in brain and thoracic ganglia (TG), but not in non-neural tissues. FXPRLamide (Phe-X-Pro-Arg-Leu) peptide immunoreactivity was detected in the SG, TG and terminal abdominal ganglion of A. pernyi by whole-mount immunocytochemistry. The titres of FXPRLamide peptides in the haemolymph in diapause type are consistently lower than in non-diapause insects. In non-diapause individuals, there are two peaks of FXPRLamide titres, day 2 in the wandering stage and day-5 in the pupal stage respectively.


Insect neuropeptides play important roles in the regulation of metabolism, development, reproduction, rhythm, diapause, behaviour and cognizance (Nassel 2002). The FXPRLamide family neuropeptides, with FXPRL (Phe-X-Pro-Arg-Leu) amide C terminal sequence have been characterized in recent years. Diapause hormone (DH) and pheromone biosynthesis activating neuropeptide (PBAN) are typical peptides of the FXPRLamide family; both are encoded by a cDNA of the DH-PBAN gene. PBAN showed similar physiological functions in different insect species, i.e. stimulation of pheromone biosynthesis (Choi et al. 1998, 2004; Iglesias et al. 2002; Wei et al. 2004; Lee and Boo 2005), whereas DH elicited contrasting effects in different diapause types, inducing the embryonic diapause in Bombyx mori (Sato et al. 1993), and terminating the pupal diapause in Heliothis virescens (Xu and Denlinger 2003), Helicoverpa armigera (Zhang et al. 2004b) and Helicoverpa assulta (Zhao et al. 2004). The Chinese oak silk moth Antheraea pernyi is close to B. mori in the taxonomic classification, but with pupal diapause, which is similar to H. armigera, H. virescens and H. assulta. Therefore, A. pernyi is a good model insect to illuminate the functions of FXPRLamide family neuropeptides. At the first step, it is important to characterize the molecular structure, location and developmental expression of FXPRLamide family neuropeptides in A. pernyi.

Using molecular cloning methods, the sequence of cDNA encoding the DH-PBAN pre-prohormone has been identified in various species including B. mori (Sato et al. 1993), Heliothis zea (Ma et al. 1994), H. assulta (Choi et al. 1998), Mamestra brassicae (Jacquin-Joly et al. 1998), Agrotis ipsilon (Duportets et al. 1999), Spodoptera littoralis (Iglesias et al. 2002), H. virescens (Xu and Denlinger 2003), Manduca sexta (Xu and Denlinger 2004), H. armigera (Zhang et al. 2004a), Adoxophyes sp (Choi et al. 2004), Samia cynthia ricini (Wei et al. 2004), Plutella xylostella (Lee and Boo 2005) and Spodoptera exigua (Xu et al. 2007). All DH-PBAN genes share similar structural characteristics and encode proteins that are cleaved into five FXPRLamide family neuropeptides: DH, PBAN, α-SG neuropeptide (SGNP), β-SGNP and γ-SGNP.

Because of easy rearing, large body size, and the summer and winter diapause, A. pernyi has been extensively used as a model animal to analyse the photoperiodic time measurement and the neuroendocrine diapause regulation (Tohno et al. 2000; Tohno and Takeda 2001). Tohno et al. (2000) reported that two independent mechanisms operate to terminate diapause under the long-days photoperiod and during chilling, and Matsumoto and Takeda (2002) examined the significance of brain monoamines in diapause pupae. However, there have been no studies on the characterization, expression pattern or function of FXPRLamide family neuropeptides in A. pernyi, although zooblotting has recently indicated that the DH-PBAN-like gene may be present in A. pernyi genome (Wei et al. 2006).

In this study, we cloned DH-PBAN cDNA from the SG of A. pernyi by reverse transcriptase-polymerase chain reaction (RT-PCR) and rapid amplification of cDNA ends (RACE). Using Northern blot analysis and RT-PCR, we examined temporal differences in DH-PBAN gene expression between the non-diapausing and diapausing pupae of A. pernyi. Changes in the titre of FXPRL immunoreactivity in the haemolymph were monitored by competitive enzyme-linked immunosorbent assay (ELISA), and immunohistochemistry was used to identify neurosecretory cells containing FXPRLamide-like peptides.

Materials and Methods


The Qinghuang bivoltine strain of Chinese oak silkworm A. pernyi was used in the present experiment. Antheraea pernyi was reared on oak trees in the field under natural condition during May and June. To obtain summer diapausing pupae, larvae of A. pernyi were grown in the silkmoths factory of Xinyang, Henan provice, China. Non-diapause-destined pupae were grown in the Wild Silkmoth Research Center, Liaoning Province, China. When larvae began to spin, they were moved indoor and allowed to spin on the branches of oak trees as previously described (Tohno et al. 2000). Diapause-destined pupae (diapause rate was >90%) were maintained under a short-day photoperiod at 10 : 14 h [light : dark (L : D)] at 25°C (Tohno et al. 2000). Exclusively non-diapausing pupae were maintained under a long-day photoperiod at 14 : 10 h (L : D) and 25°C.

The SG, brain-SG complex and other tissues were dissected in insect saline containing 0.75% NaCl. The haemolymph was collected from wandering larvae, pupae and adults and stored at −70°C until use.

RNA extraction, RT-PCR and RACE amplification

The total RNA used for RT-PCR and RACE was isolated from the SG of A. pernyi pupae with the acid guanidinium thiocyanate–phenol–chloroform method (Chomczynski and Sacchi 1987). A homogenate of 20 SGs in solution D (4 m isothiocyanate guanidine, 25 mm sodium citrate (pH 7.0), 5 g/l sodium dodecyl sarcosine, 0.1 mβ-mercaptoethanol) was placed on ice for 5 min before sodium acetate and chloroform/isoamylalcohol (49 : 1) were added. The mixture was centrifuged at 10 000 g at 4°C for 20 min, the supernatant transferred into a new tube, mixed with isopropanol and centrifuged. RNA pellet was washed in 75% ethanol and dissolved in ddH2O. One microgram of total RNA was reverse transcribed at 42°C for 1 h in a 10-μl final volume reaction mixture containing reaction buffer, 10 mm 1,4-dithiothreitol (DTT), 0.5 mm dNTP, 0.5 μg oligo-dT18 and reverse transcriptase from avian myeloblastosis virus (AMV) (Takara, Japan).

Two degenerate primers DP1 and DP2 (fig. 1) synthesized by Invitrogen Company (Carlsbad, CA, USA) were used to amplify a fragment of Anp-DH-PBAN cDNA. PCR was performed under the following conditions: three cycles of 40 s at 94°C, 40 s at 45°C, 45 s at 72°C, then 30 cycles of 40 s at 94°C, 40 s at 47°C, 45 s at 72°C. The PCR product corresponding to 400 bp was gel purified (Takara), and ligated into T-vector (Takara), and the recombinant plasmid DNA was transformed into XL-1 blue competent bacteria. Positive clones were sequenced by the dideoxynucleotide chain termination method (Takara Co., Dalian, China).

Figure 1.

 Nucleotide sequence and deduced amino acid sequence of the diapause hormone (DH) and pheromone biosynthesis activating neuropeptide (PBAN) in Antheraea pernyi (Anp-DH-PBAN) cDNA. The suggested start code ATG and stop code TAA are indicated in boxes. Putative polyadenylation signals are shown in bold letters. The five presumptive peptides are underlined, and endoproteolytic cleavage sites are shown by dotted line. Arrows under the nucleotide sequences represent the position of the different synthetic primers used in PCR. Degenerate primers are DP1 (5′-TGGTTCGG(T/C)CCC(C/A)G(G/A)CT-3′) and DP2 (5′-(G/C)AGCCT(T/G)GG(G/C)GAGAAGTA-3′). Specific primers for RACE and PCR are ADHF (5′-AACTGGGTAGAAGCATCG-3′) and ADHR (5′-AGACAG TCGTCTACCGAG-3′). Specific primer ADH (5′-TTCCTATTAGTTTATCC-3′) is for amplification of the probe.

Two specific primers ADHF and ADHR based on the partial cDNA sequences were synthesized for the 3′ and 5′ RACE performed according to the manufacturer’s protocol (SMARTTM Kit, Clontech, Mountain View, CA, USA). 5′-RACE amplification was done with 2.5 μl of 5′-ready-cDNA with Universal Primer Mix (UPM) (Clontech) and ADHR. 3′-RACE amplification was performed with 2.5 μl of 3′-ready-cDNA with UPM (Clontech) and ADHF. The reaction mixture was subjected to 32 thermal cycles that consisted of 94°C/30 s, 60°C/60 s, 72°C/45 s and 7 min at 72°C. The PCR products were sequenced after being inserted into T-vector (Takara Co.).

Based on the amino acids from the known DH-PBAN precursor, the neighbour-joining tree was constructed by CLUSTALX (Thompson et al. 1997).

Northern blot analysis

Northern blot analysis was performed according to Sambrook et al. (1989). Thirty microgram of total RNA from the brain and suboesophageal ganglion complexes (Br-SGs) was separated on a 1.2% agarose gel containing 0.22 m formaldehyde and ethidium bromide and subsequently transferred to a nylon membrane (Hybond N+, Amersham, Piscataway, NJ, USA). The Anp-DH-PBAN cDNA (corresponding to the fragment of 21–441 of Anp-DH-PBAN cDNA in fig. 1) was labelled with [α-32P]-dCTP with a random primed DNA labelling kit (Takara). Hybridization procedures and hybridization signals were visualized as described in Wei et al. (2004).

Developmental expression analysis of Anp-DH-PBAN mRNA

Developmental expression of Anp-DH-PBAN mRNA was measured with the combined methods of quantitative RT-PCR and Southern blots according to previously described procedures (Wei et al. 2004). Total RNA was extracted from 20 pupal Br-SG complexes. The first strand cDNA was synthesized from 1 μg total RNA with an AMV reverse transcript system kit (Takara) at 42°C for 1 h. The Anp-DH-PBAN cDNA fragment was amplified with the primers ADH and ADHR (fig. 1) for 18 cycles to assure that the reaction was in linear range, based on our preliminary experiment (data not shown). The PCR products were electrophoresed on a 1.2% agarose gel and transferred to a Hybond-N+ membrane. Using labelled Anp-DH-PBAN cDNA as a probe, Southern hybridization was performed using the same procedure as described for the Northern blots.

Whole-mount immunocytochemistry

The antiserum against H. armigera PBAN was prepared as described in the previous studies (Wei et al. 2004). The distribution of PBAN-like immunoreactivity in the central nervous system of A. pernyi was investigated by whole-mount immunocytochemistry as reported previously (Sun et al. 2003; Wei et al. 2004). Different tissues of A. pernyi were dissected and desheathed, and fixed for 3–6 h. After incubation with the primary antibodies (1 : 10 000) for 6 h, the tissues were soaked in peroxidase-labelled secondary antibodies (horseradish peroxidase-conjugated goat anti-rabbit IgG, Promega, Beijing, China) (1 : 2000) for 6 h. The tissues were washed in 1 × Phosphate Buffered Saline Tween-20 (PBST) (0.01 mol/l sodium phosphate, 0.15 mol/l sodium chloride, pH 7.25, 0.05% Tween-20) for 4–6 h and then washed in 50 mm Tris–HCl buffer (pH 7.6) for 30–60 min; colour development was obtained with a DAB Stock Stain Kit (Sino-American Biotechnology Co., Beijing, China). The tissues were then transferred to phosphate-buffered saline and dehydrated for 30 s each time with 50, 75, 95 and 100% ethanol; after clearing with dimethylbenzene and mounting with gum, stained tissues were observed under a microscope. The same procedures were followed for the controls, except the primary antibodies were replaced by pre-immunized rabbit serum.

Competitive ELISA

Titre of FXPRLamides was determined by competitive ELISA in 200 μl haemolymph collected from 10 pupae (20 μl per pupa). Five samples were tested at each stage. Competitive ELISA was performed at room temperature. The plates were coated with synthetic Anp-PBAN (2 pmol per well) overnight at 4°C, washed with PBST and blocked with 350 μl 2% non-fat dry milk before the samples and the primary antibody were added. The secondary antibody (horseradish peroxidase-conjugated goat anti-rabbit IgG, Promega) was added after 2 h. After the reaction with peroxidase substrate, the absorbance at 490 nm was measured with a microplate reader. The amount of FXPRLamides was determined by comparison with a standard curve (Sun et al. 2003; Wei et al. 2004).


Cloning of Anp-DH-PBAN cDNA

Based on known DH-PBAN cDNA sequences, two degenerate primers DP1 and DP2 were designed for amplification of partial fragment of Anp-DH-PBAN cDNA (fig. 1). A DNA fragment of expected size 400 bp was amplified by PCR, cloned and sequenced. The deduced amino acid sequence showed high homology to known DH-PBAN pre-prohormones. To obtain the full-length cDNA of Anp-DH-PBAN, two specific primers ADHF and ADHR (fig. 1) were designed for the 3′- and 5′-RACE. A 439-bp fragment from the 5′-RACE and a 445-bp fragment from the 3′-RACE were obtained, with an 89-bp overlapping sequence. The full-length Anp-DH-PBAN cDNA of 795 bp, with an open reading frame (ORF) of 597 nucleotides, was deposited in the GenBank under accession number AY445658. The 5′ untranslated region upstream of the transcription start code (ATG) is about 41 nucleotides. The ORF is terminated by a TAA stop code that is followed by a 163-bp 3′ untranslated region. A consensus polyadenylation signal (AATAAA) was found 15 bp upstream of the poly A tail. The Anp-DH-PBAN ORF encoded a pre-hormone of 196 amino acids containing six potential endoproteolytic cleavage sites (G48-R49-K50, K96-K97, G105-R106, G128-R129-R130, G164-R165 and G174-R175). After cleavage, five FXPRLamide family neuropeptides can be obtained: Anp-DH-like (N24-L47), Anp-PBAN (L131-L163), and three SG neuropeptides (α-SGNP: V98-L104, β-SGNP: S107-L127 and γ-SGNP: T166-L173).

Consistent with the taxonomic classification, the putative amino acid sequence of the Anp-DH-PBAN pre-prohormone showed 82% identity with that of S. cynthia ricini from the Saturniidae family, and only 53% with that of P. xylostella from Plutellidae. Comparison of five neurohormones derived from the DH-PBAN pre-prohormone of A. pernyi with their counterparts in other species showed different levels of similarity: Anp-α-SGNPs from the compared species were 86–100% identical, Anp-DHs 38–72%, Anp-PBANs 45–85%, Anp-β-SGNPs 33–76% and Anp-γ-SGNPs 50%–88% (table 1).

Table 1.   Comparative analysis of the homology of the five FXPRLamide neuropeptides from different insects
 ORF (%)DH (%)α-SGNP (%)β-SGNP (%)PBAN (%)γ-SGNP (%)Accession number
  1. ORF, open reading frame; DH, diapause hormone; SGNP, suboesophageal ganglion neuropeptide; PBAN, pheromone biosynthesis activating neuropeptide.

  2. Only the partial sequence of cDNA of DH-PBAN gene from Mamestra brassicae and Lymantria dispar were cloned and sequenced. The hyphen in the line of Mamestra brassicae and Lymantria dispar represent the value that cannot be compared.

Antheraea pernyi100100100100100100AY445658
Samia cynthia ricini8292100768288AY286544
Manduca sexta759286688075AY172672
Bombyx mori677586437675D28810
Helicoverpa armigera6771100578575AY043222
Helicoverpa assulta6971100578575U96761
Helicoverpa zea6871100578575U08109
Heliothis virescens6571100577975AY173075
Agrotis ipsilon6875100527675AJ009674
Spodoptera littoralis6975100577375AF401480
Spodoptera exigua6975100577575AY628764
Adoxophyes sp5438100624550AF395670
Plutella xylostella5358100334575AY904342
Mamestra brassicae--100577975AF044079
Lymantria dispar----67-P43511

Amino acid sequence deduced from Anp-DH-PBAN cDNA was closely related to that found in the superfamily Bombycoidea, name in S. cynthia ricini (Saturniidae), M. sexta (Sphingidae) and B. mori (Bombycidae), and was distinct from the DH-PBAN pre-prohormones of Noctuidae, Tortricidae and Plutellidae families (fig. 2).

Figure 2.

 Phylogenetic analysis by neighbour-joining (NJ) trees based on the amino acid sequences of the ORFs of DH-PBAN. Antheraea pernyi (AY445658), Samia cynthia ricini (AY286544), Manduca sexta (AY172672), Bombyx mori (D28764), Helicoverpa armigera (AY043222), Heliothis zea (U08109), Helicoverpa assulta (U96761), Heliothis virescens (AY173075), Agrotis ipsilon (AJ009674), Spodoptera exigua (AY628764), Spodoptera littoralis (AF401480), Plutella xylostella (AY904342) and Adoxophyes sp (AF395670). Branch lengths are proportional to the number of amino acid substitutions.

Expression of mRNA of Anp-DH-PBAN gene

Northern blot hybridization resulted in a single band of 800 bp (fig. 3), which is consistent with the full-length cDNA (fig. 1). In a comparison of 9-day old pupae, the level of DH-PBAN mRNA was much lower in the diapausing pupae than in the non-diapausing insects (fig. 3). We also examined the developmental expression of the DH-PBAN mRNA using the combined methods of quantitative RT-PCR and Southern blotting as previously described (Wei et al. 2004). During the early pupal stage, Anp-DH-PBAN mRNA levels increased consistently in both diapause- and non-diapause-destined pupae (fig. 4). After day 7, Anp-DH-PBAN mRNA remained high until adult eclosion in non-diapause-destined pupae, whereas levels drop markedly and remain low in diapause-destined pupae.

Figure 3.

 Northern blot analysis of Anp-DH-PBAN mRNA. Total RNA (30 μg) from brain- suboesophageal ganglion complex of day-9 pupae was loaded on each lane and observed under UV light to verify equal loading. 32P-labelled Anp-DH-PBAN cDNA was used as a probe. ND, non-diapausing type; D, diapausing type.

Figure 4.

 Developmental expression of Anp-DH-PBAN mRNA in brain-suboesophageal ganglion (SG) complexes from non-diapausing and diapausing types. Total RNA was extracted from 20 brain and SGs complexes (brain-SG complexes) from pupae was reverse transcribed after RT-PCR amplification (18 cycles), and then Southern blotted with Anp-DH-PBAN cDNAs as probes. Numbers represent days of the pupal stage; A, adult.

Tissue distribution of DH-PBAN mRNA was examined with RT-PCR (fig. 5). The expected 420-bp DH-PBAN cDNA was strongest in the SG, but a weak band could be detected in the brain and the thoracic ganglia (TG). No positive signals were found in the abdominal ganglia (AG) and the non-neural tissues, such as midgut, silk gland, fat body and epidermis. This distribution of Anp-DH-PBAN mRNA is similar to that noted in H. zea (Ma et al. 1998), H. virescens (Xu and Denlinger 2003), M. sexta (Xu and Denlinger 2004) and S. cynthia ricini (Wei et al. 2004).

Figure 5.

 Tissue distribution of Anp-DH-PBAN mRNA based on RT-PCR method. Total RNA was extracted from different tissues, and then subjected to 30 PCR cycles. SG, suboesophageal ganglion; Br, brain; TG, thoracic ganglion; AG, abdominal ganglion; Mg, midgut; SI, silk gland; FB, fat body; Ep, epidermis. The control (C) used the recombinant plasmid DNA as PCR template.

Location of cells expressing of FXPRLamide peptides in Antheraea pernyi

FXPRLamide-like immunoreactivity was detected in the whole-mounts of the nervous system. The labial, mandibular and maxillary neuromeres of SG possessed one, two and four pairs of immunopositive cells respectively (fig. 6a). No immunostaining was noted in the control experiment, when the primary antiserum was replaced with the pre-immunized rabbit serum (fig. 6b). A pair of immunopositive cells was also detected in the TG (fig. 6c) and in the terminal AG (fig. 6d) at day-11 pupae.

Figure 6.

 FXPRLamide peptides immunoreactivity against anti-Har-PBAN serum in whole-mount preparation of the pupal central nervous system. (a) Brain and suboesophageal ganglion complex (Br-SG). The SG is stained in three cell clusters: mandibular (Md), maxillary (Mx) and labial (Lb); (b) Br-SG, the primary antibodies were replaced by pre-immunized rabbit serum; (c) thoracic ganglion (TG); (d) terminal abdominal ganglion (TAG). Both TG and TAG contain one pair of cell clusters (arrow) and neuritis projections (arrowhead). Scale bars equal 40 μm.

Changes of FXPRLamide peptide titre in the haemolympha in Antheraea pernyi

Titre changes of FXPRLamide-like peptide were measured with competitive ELISA in the haemolymph of both diapause-destined and non-diapausing A. pernyi (fig. 7). Two peaks of FXPRLamides, on day 2 in the wandering stage and day 5 in the pupal stage, were found in the non-diapausing individuals. The first peak is correlated with the onset of larval–pupal, and the second with the onset of the pupal–adult transformation. FXPRLamide titre remained consistently low in the diapause-destined insects.

Figure 7.

 Developmental changes of FXPRLamide-like immunoreactivity in the hemolymph. The haemolymph samples were collected and analysed by competitive ELISA as described in the Materials and Methods. W, P and A represent wandering larvae, pupae, and adults respectively. The haemolymph from 10 individuals was mixed as a sample for competitive ELISA, and each point represents the mean ±SD of five repeats.


Neuropeptides of the FXPRLamide family play various biological roles in insects, such as stimulation of pheromone biosynthesis (Choi et al. 1998, 2004; Iglesias et al. 2002; Wei et al. 2004; Lee and Boo 2005), regulation of diapause (Sato et al. 1993), regulation of cuticular melanization (Matsumoto et al. 1992), acceleration of pupariation (Nachman et al. 1997; Zdarek et al. 1998) and promotion of pharate adult development (Sun et al. 2003; Zhang et al. 2004b). Diapause hormonr, one member of the FXPRLamide family, shows contrasting effects in different diapause types, inducing the embryonic diapause in B. mori (Sato et al. 1993), and terminating the pupal diapause in H. virescens (Xu and Denlinger 2003), H. armigera (Zhang et al. 2004b) and H. assulta (Zhao et al. 2004). Although A. pernyi is closely related to B. mori taxonomically, its life cycle includes a pupal diapause, whereas B. mori undergoes embryonic diapause. A comparison of these two organisms may illuminate the function of DH.

Using the Zooblotting method, Wei et al. (2006) demonstrated the presence of a DH-PBAN gene in the genomic DNA of A. pernyi. In the current study, we obtained the full length of Anp-DH-PBAN cDNA and confirmed similarity of the DH-PBAN pre-prohormones among Lepidoptera. Anp-DH-PBAN pre-prohormone and any of the five FXPRLamide peptides it contains are most closely homologous to S. cynthia ricini. Anp-DH is 75% identical with the DH of B. mori that has embryonic diapause, and 71% with those of H. armigera and H. virescens that have pupal diapause.

We investigated the developmental changes in the Anp-DH-PBAN mRNA in Br-SGs (fig. 4) and in the FXPRLamide-like peptides in the haemolymph (fig. 7). The content of DH-PBAN mRNA in Br-SGs was higher in the non-diapause pupae than in the diapause-destined insects (figs 3 and 4), suggesting that the expression was suppressed in diapause, similar to H. armigera (Zhang et al. 2004a) and H. virescens (Xu and Denlinger 2003). Similar levels of Anp-DH-PBAN mRNA between non-diapause- and diapause-destined pupae before day 7 may be attributed to specific features of the summer diapause in A. pernyi. H. armigera and H. virescens are induced to diapause with short-day photoperiod [10 : 14 h (L : D)] and adapted to survival at low temperature, (Xu and Denlinger 2003; Zhang et al. 2004b), whereas A. pernyi used in the present study entered diapause at higher summer temperature and were kept at 25°C and short-day photoperiod [10 : 14 h (L : D)]. Relative high level of DH-PBAN mRNA in diapausing A. pernyi may be related to high temperature. The drop of DH-PBAN mRNA in the diapause-destined pupae after day 7 suggests that diapause is firmly established at this time. There are indications that summer diapause is labile during the early pupal stage and may be broken by exposing A. pernyi pupae to a long-day photoperiod [16 : 8 h (L : D)]. It is easier to terminate summer pupal diapause in the early pupa than in the middle and latter pupal stages in A. pernyi (Tohno et al. 2000).

The difference in DH-PBAN mRNA levels between the diapause and non-diapause pupae was paralleled by differences in the titre of FXPRLamide-like peptides in the haemolymph. The titre was significantly lower in the diapausing than in the non-diapausing insects (fig. 7), similar to the situation in H. armigera (Sun et al. 2003) and H. assulta (Zhao et al. 2004). The titre peaks in wandering larvae and in about one-third of the pupal stage may be related to intensive morphogenesis that occurs at these times. Acceleration of puparium formation in the flesh flies (Zdarek et al. 1997, 1998) and stimulation of adult development in the pupae of H. virescens (Xu and Denlinger 2003) and H. armigera (Zhang et al. 2004b) also indicated a role of FXPRLamides in metamorphosis.

The DH-PBAN mRNA content in the Br-SG complex of the non-diapausing pupae seems to be maximal around day 9 (fig. 4), whereas FXPRLamide-like peptides in the haemolymph reached a peak 4 days earlier (fig. 7). A probable explanation of this discrepancy is that some of the FXPRLamides are derived from genes other than Anp-DH-PBAN mRNA. Interestingly, a similar difference between the expression of DH-PBAN mRNA (Zhang et al. 2004b) and the titre of FXPRLamides was found in H. armigera (Sun et al. 2003).

In this study, we used antibodies against H. armigera PBAN to characterize the developmental expression, tissue distribution and cell localization of neuropeptides from the FXPRLamide family in A. pernyi. The developmental titre and the cell location (figs 6 and 7) resemble those in H. armigera (Sun et al. 2003), but differ from Adoxophyes sp., in which Choi et al. (2004) found, in frontal brain region, two pairs of small cells reacting with the Hez-PBAN antibody. It is not excluded that this antibody was less specific than the Har-PBAN used in our study.


This study is supported by the National Natural Science Foundation of China (30500374) and the National Major Basic Research Project (‘‘973’’) of China (No. 2005CB121000). We appreciate helpful comments provided by the anonymous reviewers. We thank Dr. Joe Rinehart (the U.S. Department of Agriculture in Fargo North Dakota) for helping in revising of the manuscript; thanks to Mr. Bao Zhi-Yuan of the silkmoths factory of Xinyang, Henan provice, China, who helped rear diapause type Antheraea pernyi.