Directory UMM :Journals:Journal of Insect Physiology:Vol46.Issue8.Aug2000:

(1)

www.elsevier.com/locate/jinsphys

Central processing of sex pheromone stimuli is differentially

regulated by juvenile hormone in a male moth

Christophe Gadenne

_{, Sylvia Anton}

b,*

a_{INRA, Unite´ de Recherches en Sante´ Ve´ge´tale, Centre de Recherches de Bordeaux, BP 81, 33883 Villenave d’Ornon cedex, France} b_{Department of Ecology, Ecology Building, Lund University, S-223 62, Lund, Sweden}

Received 5 November 1999; received in revised form 22 December 1999; accepted 5 January 2000

Abstract

In the male moth, Agrotis ipsilon, the neuronal basis for juvenile hormone (JH)-linked modulation of sex pheromone responsive-ness was investigated following stimulation of the antenna with i) an extract of female pheromone gland, ii) the synthetic pheromone blends from A. ipsilon and a closely related species, A. segetum, and iii) single components of the A. ipsilon blend. Response characteristics of olfactory interneurons were studied in the antennal lobe (AL) at different ages and with manipulated JH levels using intracellular recording techniques. Blend-specific, generalist and component-specific neurons were identified and described according to their response pattern. The proportion of low threshold AL interneurons increased significantly with age for all stimuli tested. Changes were, however, less pronounced for the minor single components. The proportion of low threshold AL interneurons in allatectomized (JH-deprived) mature males was significantly lower for all stimuli than in intact mature males. A large proportion of low threshold AL interneurons responding to the pheromone blends, but not as pronounced for single pheromone components, could be restored/induced by injecting JH either into JH-deprived mature males or into young immature males. The specificity for the species-specific blend compared to the A. segetum blend increased with age and JH injections. 2000 Elsevier Science Ltd. All rights reserved.

Keywords: Agrotis ipsilon; Juvenile hormone; Olfaction; Antennal lobe; Sex pheromone

1. Introduction

In insects, juvenile hormone (JH), produced in the corpora allata (CA), has been shown to control many different processes linked with development and repro-duction. However, very few studies have dealt with its effects on the nervous system of adult insects (see reviews by Wyatt and Davey, 1996; Strambi et al., 1999). Beyond the few known examples, JH was shown to play a role in central nervous regulation of phonotaxis (Walikonis et al., 1991) and in the neurogenesis of the adult house cricket (Cayre et al., 1994). In the honeybee, changes in the brain structures related to behavioural maturation and age polyethism seem to be induced by JH (Fahrbach and Robinson, 1996).

In the male moth, Agrotis ipsilon, JH was shown to

* Corresponding author. Tel:+46-46-2224-498; fax:+ 46-46-2224-716.

E-mail address: [email protected] (S. Anton).

be involved in both behavioural and physiological pro-cesses linked with mating events. The behavioural responsiveness towards the female sex pheromone increases with age and JH biosynthesis activity (Gadenne et al., 1993; Duportets et al., 1998). Moreover, removal (allatectomy) of the corpora allata, the retrocer-ebral gland producing JH, suppressed male responsive-ness, and JH was able to restore the sexual behaviour of the operated males (Gadenne et al., 1993; Duportets et al., 1996). Electroantennographic studies revealed that the peripheral olfactory system was fully functional in newly emerged and allatectomized males, thus suggest-ing that JH was more likely to act on the central inte-gration of the pheromone signal leading to sexual behav-iour (Gadenne et al., 1993).

In male moths, the female-produced sex pheromone is first detected by antennal receptor neurons and then integrated by interneurons in male-specific glomeruli called the macroglomerular complex (MGC) within the antennal lobe (AL) which is part of the deutocerebrum (Hansson, 1995). We have recently shown that JH is

(2)

involved in the central nervous processing of the sex pheromone blend in A. ipsilon by regulating the response threshold of the AL interneurons (Anton and Gadenne, 1999). The multi-component sex pheromone blend act-ing as a species-specific communication signal in A.

ipsi-lon is a mixture of three main components at a 4:1:4

ratio: (Z)-7-dodecen-1-yl acetate (Z7-12:OAc), (Z)-9-tetradecen-1-yl acetate (Z9-14:OAc) and (Z)-11-hex-adecen-1-yl acetate (Z11-16:OAc) (Hill et al., 1979; Wakamura et al., 1986; Picimbon et al., 1997; Gemeno and Haynes, 1998). Moreover, under laboratory con-ditions, this noctuid moth was shown to hybridize with the sympatric species, Agrotis segetum (Gadenne et al., 1997). Both species share two pheromone components (Z7-12:OAc and Z9-14:OAc) (Lo¨fstedt et al., 1982; Wu et al., 1995) for which receptor neurons were found in

A. ipsilon males (Renou et al., 1996). Olfactory receptor

neurons responding to Z11-16:OAc have not yet been identified.

In order to elucidate whether JH acts both on the pro-cessing of the sex pheromone blend and on the discrimi-nation of relevant sex pheromone components in A.

ipsi-lon, we performed intracellular recordings from AL

interneurons in A. ipsilon males of different ages and hormonal status. We stimulated the antennae with vari-ous concentrations of both A. ipsilon and A. segetum synthetic sex pheromone blends as well as single compo-nents. The results indicate that the proportion of AL interneurons with different response thresholds for the pheromone in the male moth, A. ipsilon, is differentially controlled by JH.

2. Materials and methods

2.1. Insects

A. ipsilon males were taken from a laboratory colony

in Bordeaux, France. Moths were originally caught in southern France and adults from field catches were intro-duced into the colony each spring. Black cutworm larvae were reared on an artificial diet (Poitout and Bues, 1974) and maintained in individual plastic cups until pupation under a 16L:8D photoperiod and at 21±1°C. Pupae were sexed and newly emerged adults were removed from pupae containers every day. Adult males and females were held separately in plastic boxes and had free access to a 20% sucrose solution. For electrophysiological experiments, a male moth was mounted in a plastic pip-ette tip with the head protruding. The cuticle, tracheal sacs, and muscles were carefully removed from the front of the head to expose the brain. The AL was manually desheathed to facilitate microelectrode penetration. A flow system perfused the head cavity with saline (Christensen and Hildebrand, 1987).

2.2. Intracellular recording and staining

Standard intracellular recording and staining tech-niques were used (Christensen and Hildebrand, 1987). A glass microelectrode filled with 2 M KCl, or filled with 1% Neurobiotin (Vector labs) in 2 M KCl was used as the recording electrode. Pheromone-sensitive AL inter neurons were penetrated in the male-specific MGC (Gemeno et al., 1998). To confirm that recordings were made from MGC projection neurons, we attempted to stain neurons in 16 preparations. Neurobiotin was vis-ualized either with Lucifer Yellow Avidin (Molecular Probes) or with the ABC-DAB method (Vector Labs) on wholemounts. Neurons were reconstructed from 10 µm plastic sections.

2.3. Stimulation

The moth antennae were stimulated with single phero-mone components, synthetic pherophero-mone blends, a female gland extract or clean air as a blank. The single compo-nents used were the three single pheromone compocompo-nents found in the female A. ipsilon gland: Z7-12:OAc, Z9-14:OAc, Z11-16:OAc (two components: Hill et al., 1979 three components: Picimbon et al., 1997; Gemeno and Haynes, 1998). A synthetic 3-component A. ipsilon pher-omone blend was a mixture of Z7-12:OAc, Z9-14:OAc and Z11-16:OAc at a ratio of 4:1:4, as used in field trap-ping experiments (Causse et al., 1988). The synthetic A.

segetum pheromone blend was a mixture of Z5-10:OAc,

Z7-12:OAc, Z9-14:OAc, and Z5-12:OAc at a ratio of 1:5:2.5:0.1, as used in field trapping experiments in Sweden (Wu et al., 1995). One female equivalent of a gland extract of sexually mature 4-day-old A. ipsilon females was used for comparison with the synthetic pheromone blend. To separate plant-specific neurons from our study, a common green leaf volatile, (E)-2-hexenal, was also used as a stimulus.

The single pheromone components and the phero-mone blends, all diluted in hexane, were applied at amounts between 1 pg and 100 ng on a filter paper in a Pasteur pipette. A 500 ms airpulse was sent through the pipette containing either clean air, or an odorant, by means of a stimulatory device (Syntech, Hilversum, The Netherlands) (Hansson et al., 1994). For screening, the pheromone blends were first used at an intermediate con-centration as stimuli, then the other stimuli were given randomly, separated by at least 10 s, or when spike fre-quency was back to pre-stimulation activity. The physio-logical data were stored on a Vetter video recorder and visualized on a Tektronix digital oscilloscope.

2.4. Data analysis

Spikes were counted manually from the storage oscilloscope, and responses were judged by observation

(3)

only. Net-spikes were calculated from the number of spikes during a 600 ms period after the stimulus minus the number of spikes counted during the preceding 600 ms (representing spontaneous activity). The 600 ms time interval was chosen to include the excitatory response in the majority of responses to the stimuli in AL neurons. The net number of spikes produced in response to the blank stimulus was subtracted from the net number of spikes produced in response to an odour stimulus to quantify the response to a specific stimulus (Anton et al., 1997). A neuron was classified as responding to a stimulus when the odour response exceeded the blank response by at least 10%. By comparing the results from pure observation and exact data analysis as described above for more than 400 responses, a discrepancy of less than 5% was observed. We therefore analysed the remaining data by simple observation.

For statistical treatment, four groups of neurons were created: one with non-responding neurons (no response), one with high threshold neurons (100 ng and 10 ng), one with medium threshold (1 ng and 0.1 ng) and another with low threshold (0.01 ng and 1 pg). Neurons responding exclusively to the plant odour were not taken into account. A G-test for heterogeneity (Sokal and Rohlf, 1995) was performed comparing the four groups of neu-rons pairwise for the different ages and treatments and for each stimulus. Statistical analysis was performed for each class of age and treatment group within each stimu-lus (Figs. 4–9) and for each stimustimu-lus within each class of age and treatment group (Table 2).

Regarding the tuning of AL neurons to both blends, we did not find any statistical difference (G=5.41; df=7; NS) in the percentage of neurons responding equally to both blends for all treatments (between 35% and 52% of the neurons for the different treatments). Therefore, a G-Test was performed by taking into account only neu-rons which showed a different response threshold between blends with the age or treatments.

2.5. Surgical and injection treatments

Surgical removal of the corpora allata was performed on 3-day-old males (Gadenne, 1993). 3-day-old allatec-tomized males were injected with 2µl olive oil, or with 10µg JH-III dissolved in 2µl olive oil (Duportets et al., 1996). Newly emerged males were injected with 10 µg JH dissolved in 2 µl olive oil. Intracellular recordings were performed one or two days after the treatment, respectively.

3. Results

3.1. General physiological and anatomical characteristics of antenna1 lobe neurons

Recordings from 506 AL neurons in 114 A. ipsilon males are included in this study. AL neurons were in

most cases spontaneously active at frequencies between 10 and 40 Hz. A large proportion of the AL neurons received input from mechanosensory receptor neurons and responded with an increase in action potential fre-quency to the clean air stimulus. Certain odour stimuli, however, elicited a stronger response than the clean air puff in all AL neurons included in this study. Neurons responding to an odour stimulus showed an increase in spike frequency (typically 70-to-90 Hz at the lowest above threshold stimulus), followed by an inhibitory per-iod, before spontaneous activity recovered (Fig. 1). No purely inhibitory responses, characterized by an IPSP in response to any odour, were found in the investigated AL neurons.

Different neuron types concerning response specificity and sensitivity were found independent of the age and treatment of the male moths. Neurons were divided into three specificity types: a) component specific neurons, b) generalist and medium specificity neurons, and c) blend specific neurons (Table 1). Within each specificity type, different sensitivity classes were found; low sensitivity neurons responding to 10 or 100 ng doses; medium sen-sitivity neurons responding to 0.1 or 1 ng doses; and high sensitivity neurons responding to 0.01 ng or 1 pg doses. Among the component-specific neurons, those responding only to Z7-12:OAc and to the pheromone blend, but not to any other single components, were found frequently within all three sensitivity classes (Table 1).

Some Z9-14:OAc-specific neurons with medium and low sensitivity were present, while only few Zll-16:OAc-specific neurons were found showing only low sensi-tivity (Table 1). Generalist neurons with medium or no specificity, showing different threshold responses according to the stimulus were common (Table 1). Few generalist neurons, responding equally well to all single components and the blends, were found at all sensitivity levels (Table 1 and Fig. 1). Blend-specific neurons, responding to the A. ipsilon blend, but not to the single components present in the blend, were rare, had low sen-sitivity and in addition always responded to the A.

sege-tum blend (Table 1 and Fig. 2). Fourty percent of all AL

neurons responded equally well to the pheromone blends of their own species and of the related species A.

sege-tum. Another 40% of the AL neurons responded at a

lower dose to the A. ipsilon blend (e.g. in Fig. 2), and the remaining 20% responded at a lower dose to the A.

segetum blend.

The attempts to stain 16 AL neurons resulted in six successful complete visualisations of projection neurons, while only a cell body was stained in two preparations. All stained projection neurons had cell bodies in the medial cell group, arborized in 1 to 3 glomeruli within the MGC and sent their axons through the inner antenno-cerebral tract to the calyces of the mushroom bodies and to the lateral protocerebrum (Fig. 3).

(4)

Fig. 1. Response characteristics of a pheromone generalist antennal lobe neuron stimulated with 1 pg stimuli in an A. ipsilon male. The neuron had a rather high and variable spontaneous activity and responded to all stimuli with an increase in action potential frequency. Some smaller amplitude action potentials appear to originate from a second neuron, but the responses are restricted to one neuron only. Bar beneath registration indicates stimulation time. Vertical scale bar 10 mV, horizontal scale bar 500 ms.

3.2. Effects of age on the response thresholds of AL neurons

Responses of AL neurons to the extract of one 4-day-old female pheromone gland (1 female equivalent=1 FE) and to artificial pheromone blends of A. ipsilon and A.

segetum, as well as to the three single pheromone

components of A. ipsilon were analysed in 1-day, 3-day and 5-day-old male moths. The response thresholds for

single pheromone components and blends were com-pared statistically within each age group and for each stimulus between age groups. The number of neurons from young immature males that did not respond to any of the stimuli tested was much larger than in older males (16, 2 and 2 neurons from 1-day, 3-day and 5-day-old males respectively).

The proportion of AL neurons responding to one FE showed a significant increase with age between day 1 and day 5 (G=18.94; df=l; P#0.001). The increase was more pronounced between 3-day and 5-day-old males (G=5.24; df=1; P_#0.05) than between 1-day and 3-day-old males (G=3.67; df=1; NS) (Fig. 4). In parallel, the proportion of low threshold AL neurons increased for all other stimuli tested between day 1 and day 5. The proportion of low threshold AL neurons for all stimuli increased significantly from day 1 to day 3, except for Z11-16:OAc (G=4.61; df=3; NS). From day 3 to day 5 this proportion increased significantly only for the A. ipsilon blend (G=11.69; df=3;

P,0.001) and Z9-14:OAc (G=11.64; df=3; P,0.01), while it increased only slightly for the A. segetum blend and the two other single components (Fig. 5).

Most neurons were found to respond more strongly to the blends than to the single components. Within each age class of males, the proportion of low threshold neu-rons responding to the blends was significantly higher than of low threshold neurons responding to single components (Fig. 5 and Table 2). Moreover, the pro-portion of low threshold neurons responding to Z7-12:OAc was statistically higher than of those responding to the other two single components (Fig. 5 and Table 2). This difference in stimulus-dependent proportion of low threshold neurons did not vary with age.

3.3. Effect of JH treatments on the response thresholds of AL neurons of mature males

Allatectomy and injection of either olive oil or JH were performed on 3-day-old males followed by intra-cellular recordings of AL neurons one or two days later. When the antenna was stimulated with 1 FE, the pro-portion of low threshold AL neurons in 5-day-old alla-tectomized males injected with olive oil decreased sig-nificantly as compared to that of 5-day-old control males (G=l9.34; df=1; P_#0.05) and reached the level of that of 1-day-old control males (G=0.001; df=1; NS) (Fig. 6). One FE stimulation of the antenna of operated males injected with JH, resulted in a significantly higher pro-portion of low threshold AL neurons compared to oper-ated males injected with olive oil (G=6.83; df=1;

P#0.05), reaching a level similar to that of 5-day-old (G=3.57; df=1; NS) control males (Fig. 6). When the response of the neurons to the stimuli was tested just one day after the injection of JH (2CA+JH, day 4) the recovery of the proportion of low threshold neurons in response to the FE was only partial, being intermediate

(5)

Table 1

Examples of neuron types encountered in A. ipsilon males. Thresholds for responses are given in ng pheromone on the filter paper in the stimu-lus pipette

Neuron type Stimulus A. segetum blend Z7-12:OAc Z9-14:OAc Z11-16:OAc Abundance

A. ipsilon blend a) component

specific

(7–12), high 0.001 0.001 0.01 – – common

sensitivity

(7–12), medium 1 0.1 0.1 – – common

sensitivity

(9–14), medium 0.1 0.1 – 0.1 – few

sensitivity

(7–12), low 10 10 10 – – common

sensitivity

(9–14), low 10 1 – 10 – few

sensitivity

(11–16), low 10 – – 100 few

sensitivity

b) generalist

varying sensitivity 10 0.1 100 100 – common

varying sensitivity 1 1 100 100 10 common

high sensitivity 0.001 0.001 0.01 0.001 0.001 few

medium sensitivity 1 0.01 1 0.1 10 few

low sensitivity 10 0.1 100 100 100 few

c) blend specific

low sensitivity 10 10 – – – few

Fig. 2. Response characteristics of a pheromone blend specific neuron in an A. ipsilon male (peri-stimulus-time histograms, created from manually counted spikes). The neuron responded to the pheromone blends, but not to the single components at the highest dose tested. The neuron had a lower response threshold to the A. segetum blend than to the A. ipsilon blend. Responses to low stimulus amounts did not show IPSPs. What might appear as an inhibition is within the normal variation of spontaneous activity. Bars beneath histograms indicate stimulation time (500 ms).

between allatectomized males injected with olive oil (2CA+oil) (G=0.86; df=1; NS) and operated males injected with JH and tested two days later (2CA+JH day 5) (G=1.58; df=1; NS) (Fig. 6).

The proportion of low threshold AL neurons in allatectomized/oil injected males was low for all single components and for the A. segetum blend, reaching the level found in 1-day-old control males (Figs. 5 and 7)

(G=4.02; G=6.56, G=2.80, G=4.57; df=3; NS for A.

seg-etum blend, Z7-12:OAc, Z9-14:OAc and Z11-16:OAc,

respectively). The proportion of low threshold AL neu-rons in allatectomized/oil injected males stimulated with the A. ipsilon blend was, however, significantly lower than in 5-day-old control males (G=44.44; df=3;

P,0.001) but still significantly higher than in l-day-old males (G=14.3; df=3; P,0.01).

(6)

Fig. 3. Structure of an antenna1 lobe projection neuron in an A.

ipsi-lon male. The neuron arborizes within the largest MGC

compart-ment — the cumulus — and sends its axon to the calyces of the mush-room bodies (cmb) and to the lateral protocerebrum (lp). The projection neuron responded to both pheromone blends at 10 ng and to the major pheromone component at 100 ng. Scale bar 100 µm. Reconstruction from frontal sections.

A statistical difference in the proportion of low thres-hold neurons was observed between operated/oil injected and operated/JH injected males for both blends and two single components. Only for Z11-16:OAc were the responses of the AL neurons not different between the two treatments (G=5.83; df=3; NS) (Fig. 7). As com-pared with 5-day-old mature control males, the recovery of the proportion of low threshold neurons following JH injection was different depending on the stimuli (Fig. 7). For the A. ipsilon blend, the JH injection allowed the proportion of low threshold neurons to increase to a level not different from that of 5-day-old males (G=6.78; df=3; NS). For all other stimuli, the recovery was partial as it did not reach that of 5-day-old males (G=11.76, df=3,

P,0.01; G=8.62, df=3, P,0.05; G=30.32, df=3,

P,0.001; G=12.10, df=3, P,0.01 following stimulation of A. segetum blend, Z7-12:OAc, Z9-14:OAc and Z11-16:OAc respectively between allatectomized/JH injected males and 5-day-old control males).

JH injection one day before testing AL neuron responses (2CA+JH, day 4) resulted in significant

Fig. 4. Effect of age on pheromone responses of antennal lobe (AL) neurons. Percentage of tested AL neurons responding to 1 female gland extract equivalent in 1-day-old (n=72 neurons in 14 males), 3-day-old (n=56 neurons in 12 males), and 5-day old (n=65 neurons in 18 males)

A. ipsilon males. Bars with same letter are not statistically different

(G-Test, P,0.05).

changes of the proportion of low threshold neurons for the blends and for Z7-12:OAc compared to AL neurons in allatectomized/oil injected males, but not for Z9-14:OAc and Z11-16:OAc (Fig. 7).

The response of the AL neurons in both operated/oil injected and operated/JH injected males differed for the different stimuli (Fig. 7, Table 2). Statistical differences in the responses of neurons to the stimuli were observed within each treatment. The responses to the two blends were similar for allatectomized/oil injected males (G=6.58; df=3; NS) and the single components induced different responses in the AL neurons, Z7-12:OAc elicit-ing lower threshold responses than the two other compo-nents (Fig. 7 and Table 2). Responses varied even more in response to the different stimuli in allatectomized/JH injected males (Fig. 7 and Table 2).

3.4. Effect of JH treatments on the response thresholds of AL neurons of young immature males

JH was injected into 0-day-old males and the response of the AL neurons to the various stimuli was tested one or two days later. When submitted to a female equivalent of sex pheromone, the AL neurons of young injected males showed an increase in the proportion of low thres-hold neurons two days after the treatment which was different from 1-day-old males (G=6.34;df=l; P#0.05) and not different from 3-day- and 5-day-old control males, (G=0.21;df=l; NS and G=3.72; df=1; NS respectively) (Fig. 8). This JH-induced increase of the proportion of low threshold neurons in young immature

(7)

Fig. 5. Effect of age on response thresholds of antenna lobe (AL) neurons. Percentage of tested AL neurons responding to the pheromone blends and single components at different thresholds in 1-day-old (14 males), 3-day-old (12 males), and 5-day old (18 males) A. ipsilon males. Part of the original data for A. ipsilon blend stimulation from Anton and Gadenne (1999). For each column of histograms, those with the same letter are not statistically different (G-Test, P,0.05).

Table 2

Statistical analysis of the effects of the stimuli on the response of AL neurons as a function of age and JH treatmenta

Stimuli

Males A. ipsilon blend A. segetum blend Z7-12:OAc Z9-14:OAc Z11-16:OAc

Day-1 a ab b c c

Day-3 a a b b c

Day-5 a ab b c c

2CA+oil a a b c c

2CA+JH, day-4 a b b bc c

2CA+JH, day-5 a b c d e

+JH, day-l a a b c c

+JH, day-2 a a b c c

a_{For each line, columns with same letters are not statistically different (G-test, P}

,0.05). See Figs. 5, 7 and 9.

males was already detectable one day post-treatment (G=8.43, df=1, P,0.01; G=0.97, df=1, NS; G=1.54; df=1, NS between injected and 1-, 3- and 5-day-old males respectively). When JH was injected into young day-0 immature males, the two blends and the single components induced different responses of AL neurons in injected males when tested after two days (Fig. 9). A high proportion of low threshold neurons was observed for both blends, different from 1-day-old control males

(G=40.37 and G=18.64; df=3; P.0.001 for A. ipsilon and A. segetum blends respectively) but not different from 5-day-old males (G=4.9 and G=0.89 df=3; NS for

A. ipsilon and A. segetum respectively). Two types of

responses were observed after stimulation with the single components of JH injected males: the proportion of low threshold AL neurons in response to Z7-12:OAc, although not different from 1-day-old males (G=7.22, df=3, NS), reached the level of 5-day-old males (G=7.64,

(8)

Fig. 6. Effect of juvenile hormone on pheromone responses of antenna lobe (AL) neurons in mature males. Percentage of tested AL neurons responding to 1 female gland extract equivalent in mature A.

ipsilon males deprived of juvenile hormone (2CA+oil, day-5, n=74 neurons in 15 males), or allatectomized and injected with juvenile hor-mone at day 3 and tested at day 4 (2CA+JH, day-4, n=38 neurons in 9 males) or day 5 (2CA+JH, day-5, n=70 neurons in 20 males). Bars with same letter are not statistically different (G-Test, P,0.05).

df=3, NS). The proportion of low threshold neurons in response to the other two single components was not affected by JH: there was no significant difference between 1-day-old control males and JH injected males for Z9-14:OAc (G=3.10, df=3, NS) and Z11-16:OAc (G=0.84, df=3, NS). The increase of the proportion of low threshold AL neurons in response to the blends and to the single components was already visible one day following treatment and was more pronounced after two days (Fig. 9).

The responses of the AL neurons were similar for the two blends one and two days following JH injection, but were different from responses to Z7-12:OAc. Responses to both Z9-14:OAc and Z11-16:OAc, in turn, were dif-ferent from responses to all other stimuli (Table 2).

3.5. Tuning of neurons changing with treatment

The proportion of AL neurons responding more sensi-tively to the conspecific pheromone blend was generally higher than the proportion of neurons responding more sensitively to the blend of the related species, A. segetum (Fig. 10), with the exception of neurons in 3-day-old males, which responded more sensitively to the A.

sege-tum blend than to the A. ipsilon blend. One-day-old JH

injected males, 1-day-old control males, and to a lesser extent allatectomized males, possessed almost equal pro-portions of AL neurons responding more sensitively to either pheromone blend (Fig. 10). Five-day-old control males and, more pronounced, males which had

under-gone a JH injection other than the above mentioned, had a high proportion of AL neurons which responded more sensitively to the A. ipsilon blend and a very low pro-portion of neurons responding more sensitively to the A.

segetum blend (Fig. 10).

4. Discussion

4.1. Neural processing of sex pheromone information in the AL of Agrotis ipsilon

The most common neuron types found in A. ipsilon were component-specific neurons responding to Z7-12:OAc and neurons with medium specificity. The few blend specific neurons had high response thresholds, while all other neuron types were found at different sen-sitivity levels. We did not, however, test systematically, whether the specificity of highly sensitive neurons changed with increasing stimulus amounts.

Relatively low sensitivity of blend specific neurons has been found before in other moth species, while gen-eralist neurons often exhibit high sensitivity (Ostrinia

nubilalis, Anton et al., 1997; A. segetum, Hartlieb et al.,

1997). In both moth species mentioned, specificity of neurons often changed with increasing stimulus concen-trations. The relatively low number of component-spe-cific neurons found in A. segetum (Hansson et al., 1994; Hartlieb et al., 1997) and Spodoptera littoralis (Anton and Hansson 1994, 1995), compared to the present study in A. ipsilon, might be explained by the fact that we used lower stimulus concentrations. As specificity is often lost with increasing stimulus concentrations, neurons with specific responses were possibly missed in the former studies. As in our study, a large proportion of compo-nent-specific neurons have also been demonstrated in the sphinx moth, Manduca sexta (Christensen and Hildeb-rand, 1987; Hansson et al., 1991).

The large number of Z7-12:OAc-specific neurons compared to neurons specifically responding to the other pheromone components might be due to the high pro-portion of olfactory receptor neurons (ORNs) responding to this component. More than 60% of the sensilla investi-gated on the antenna of A. ipsilon house a receptor neu-ron for this component, while only few ORNs were found to respond to Z9-14:OAc and no ORNs have so far been found to respond to Z11-16:OAc (Renou et al., 1996). The present study demonstrates that ORNs responding to Z11-16:OAc must exist on the antenna, as we found AL interneurons responding to this component. The response thresholds for the different stimuli varied within each age/treatment class. Thresholds were in general lowest for the blends, slightly higher for Z7-12:OAc, and even higher for the two other single compo-nents. Different convergence of ORNs, as mentioned

(9)

Fig. 7. Effect of juvenile hormone on response thresholds of antenna lobe (AL) neurons in mature males. Percentage of tested AL neurons responding to the pheromone blends and single components at different thresholds in mature A. ipsilon males deprived of juvenile hormone (2CA+oil, day-5, 15 males), or allatectomized and injected with juvenile hormone at day 3 and tested at day 4 (2CA+JH, day-4, 9 males) or day 5 (2CA+JH, day-5, 20 males). Part of the original data for A. ipsilon blend stimulation from Anton and Gadenne (1999). For each column of histograms, those with the same letter are not statistically different (G-Test, P,0.05).

Fig. 8. Effect of juvenile hormone on pheromone responses of antennal lobe (AL) neurons in immature males. Percentage of tested AL neurons responding to one female gland extract equivalent in immature A. ipsilon males injected with juvenile hormone at day 0 and tested at day 1 (+JH, day-1, n=53 neurons in 11 males) or day 2 (+JH, day-2, n=69 neurons in 14 males). Bars with same letter are not statisti-cally different (G-Test, P=0.05).

above, can be responsible for such differences in AL neuron sensitivity.

The responses of AL neurons to the female gland extract-equivalent were very similar to the responses of the same neurons to the artificial pheromone blend at a dose of 1 ng, resembling the female gland production (Picimbon et al., 1997). This fact and its efficiency in behavioural experiments (Hill et al., 1979; Wakamura et al., 1986; Picimbon et al., 1997; Gemeno and Haynes, 1998) show that the artificial pheromone blend can replace the natural blend.

4.2. Effect of age on central nervous processing

The present results show that the proportions of low threshold AL neurons does not only increase with age in reponse to the species specific blend (Anton and Gad-enne, 1999), but also to the blend of a related species and to single pheromone components. This effect of age on the proportion of low threshold AL neurons is much more pronounced following the blend stimulations than with single component stimulations. Clearly, the increase of the proportion of low threshold neurons with age was more pronounced for stimulations of Z7-12:OAc than with that of Z9-14:OAc and of

(10)

Z11-Fig. 9. Effect of juvenile hormone on response thresholds of antenna lobe (AL) neurons in immature males. Percentage of tested AL neurons responding to the pheromone blends and single components at different thresholds in immature A. ipsilon males injected with juvenile hormone at day 0 and tested at day 1 (+JH, day-1, 11 males) or day 2 (+JH, day-2, 14 males). Part of the original data for A. ipsilon blend stimulation from Anton and Gadenne (1999). For each column of histograms, those with the same letter are not statistically different (G-Test, P,0.05).

Fig. 10. Influence of age and juvenile hormone on the specificity of antennal lobe neurons. Proportion of tested neurons which are more sensitive to either the A. ipsilon blend or the A. segetum blend. Neurons with the same sensitivity to both blends (35% to 52%) are not shown (see Section 2). G-Test, P,0.05.

16:OAc. The number of non-responding neurons to stimulations with Z9-14:OAc and with Z11-16:OAc was very high whatever the age of the male tested. We pre-viously found an increase in the proportion of low thres-hold AL neurons with age when the antenna was stimu-lated with the A. ipsilon pheromone blend (Anton and Gadenne, 1999).

The major changes in the proportion of low threshold neurons for single components and the blends occur between l-day old and 3-day old moths. Plasticity, how-ever, seems to be more pronounced for the species-spe-cific blend (significant increase of the proportion of low threshold neurons even between 3-day old and 5-day old males) than for single components and the blend of another species. For the minor component Z11-16:OAc,

no significant changes of the proportion of low threshold neurons were found during development, which might either be due to missing plasticity or might simply be difficult to detect, as there are so few neurons responding to that component in general.

Various behaviours controlled by physiological pro-cesses and correlated with certain anatomical features in the brain have been shown to be age-dependent in insects. In bees, the volume of the AL glomeruli and of the mushroom bodies changes with age and this growth is linked with changes in tasks performed by the bees (Fahrbach and Robinson, 1996; Sigg et al., 1997). In the ant, Camponotus floridanus, changes in the size of the AL and other parts of the brain were shown to be corre-lated with age and task (Gronenberg et al., 1996). In the

(11)

cricket, Acheta domesticus, the phonotactic response and linked with it, the response threshold of auditory neu-rons, were found to be age-dependent (Stout et al. 1991, 1998; Walikonis et al., 1991).

4.3. Effects of JH treatments on central nervous processing

JH acts differentially on the response of AL inter neu-rons to various pheromone stimuli. Our results from JH-manipulated male moths were clearly correlated with the age-dependant changes in proportions of AL neurons with different response thresholds observed.

Hormones play a role in many physiological processes occurring in the brains of vertebrates and invertebrates. In all above cited examples of age-dependent effects on the insect brain, JH was found to play a role in this regu-lation. However, in the case of the honeybee, age-depen-dent AL development was only indirectly JH-depenage-depen-dent by changing the bees’ activity, which in turn acted on the AL anatomy (Fahrbach and Robinson, 1996; Sigg et al., 1997), most likely through biogenic amines as mediators (Mercer et al., 1999). Another insect hormone, ecdysone, was shown to act on the antennal development of M. sexta through the control of the antennal protein pattern (Vogt et al., 1993). In mammals, steroids have been shown to play a role in the integration of phero-monal stimuli by maintaining the responsiveness of olfactory brain centres, i.e. the magnocellular division of the medial preoptic nucleus (Swann, 1997). Moreover, melatonin was recently shown to act on the volume of the song-control nucleus areas in the bird brain (Bentley et al., 1999).

The effect of JH injection in both allatectomized and young A. ipsilon males was gradual and rather slow. Some changes in the proportion of AL neurons with dif-ferent thresholds were observed after one day, other changes did not appear until two days after JH injection. Similarly, neurogenesis in allatectomized crickets injected with JH was observed only after three days (Cayre et al., 1994). Somewhat faster action of JH was found with topical application of JH in crickets. In female house crickets, in which JH was topically applied, an effect on phonotactic behaviour was observed within 12 hours post-treatment in young crick-ets, while application on old allatectomized cricket females was effective after 6 hours (Stout et al. 1991, 1999). Blocking of transcription and translation factors abolished the effect of JH, which indicates a nuclear effect (Stout et al., 1999). Another indication for nuclear effects of JH are cytosol and nuclear JH-binding proteins which were found in the brains of the cockroach

Diplop-tera punctata (King et al., 1994).

4.4. Tuning to the species specific blend

AL neurons of young males and allatectomized older males do not discriminate between the sex pheromone blend produced by conspecific females and the A.

sege-tum blend. In mature control males and in JH injected

males, more neurons show a higher sensitivity to the A.

ipsilon blend than to the A. segetum blend. This makes

sense, because at the time when A. ipsilon males are fully mature, the tuning of olfactory centres must be optimal in order to avoid interspecific matings with other close species such as A. segetum. These two species are sympatric in various parts of their geographical habitat. They do hybridize in laboratory conditions (Gadenne et al., 1997) and possess very similar pheromone blends.

Acknowledgements

We thank E. Warrant, B.S. Hansson and two anony-mous referees for their valuable comments on the manu-script. This work was supported by grants from Institut National de la Recherche Agronomique (INRA), the Swedish Council for Forestry and Agricultural Research (SJFR), and the Swedish Council for Natural Science (NFR).

References

Anton, S., Hansson, B.S., 1994. Central processing of sex pheromone, host odour, and oviposition deterrent information by interneurons in the antennal lobe of female Spodoptera littoralis. Journal of Comparative Neurology 350, 199–214.

Anton, S., Hansson, B.S., 1995. Sex pheromone and plant-associated odour processing in antennal lobe interneurons of male Spodoptera

littoralis. Journal of Comparative Physiology 176, 773–789.

Anton, S., Lo¨fstedt, C., Hansson, B.S., 1997. Central nervous pro-cessing of sex pheromones in two strains of the European corn borer Ostrinia nubilalis (Lepidoptera: Pyralidae). Journal of Experimental Biology 200, 1073–1087.

Anton, S., Gadenne, C., 1999. Effect of juvenile hormone on the cen-tral nervous processing of sex pheromone in an insect. Proceedings of the National Academy of Sciences USA 96, 5764–5767. Bentley, G.E., Vant Hof, T.J., Ball, G.F., 1999. Seasonal

neuroplastic-ity in the songbird telencephalon: a role for melatonin. Proceedings of the National Academy of Sciences USA 96, 4674–4679. Causse, R., Bue`s, R., Barthes, J., Toubon, J., 1988. Mise en e´vidence

expe´rimentale de nouveaux constituants des phe´romones sexuelles de Scotia ipsilon et Mamestra suasa. In: Descoins, C., Fre´rot, B. (Eds.), Me´diateurs Chimiques: Comportement et Syste´matique des Le´pidopte`res. Colloques INRA, 46. INRA publications, pp. 75–82. Cayre, M., Strambi, C., Strambi, A., 1994. Neurogenesis in an adult

insect brain and its hormonal control. Nature 368, 57–59. Christensen, T.A., Hildebrand, J.G., 1987. Male-specific, sex

phero-mone-selective projection neurons in the antennal lobes of the moth, Manduca sexta. Journal of Comparative Physiology 160, 553–569.

Duportets, L., Dufour, M.C., Be´card, J.M., Gadenne, C., Couillaud, F., 1996. Inhibition of male corpora allata activity and sexual phero-mone responsiveness in the black cutworm, Agrotis ipsilon, by the

(12)

hypocholesterolemic agent, fluvastatin. Archives of Insect Bio-chemistry and Physiology 32, 601–611.

Duportets, L., Dufour, M.C., Couillaud, F., Gadenne, C., 1998. Biosyn-thetic activity of corpora allata, growth of sex accessory glands and mating of the male moth Agrotis ipsilon. Journal of Experimental Biology 201, 2425–2432.

Fahrbach, S.E., Robinson, G.E., 1996. Juvenile hormone, behavioral maturation, and brain structure in the honey bee. Developmental Neuroscience 18, 102–114.

Gadenne, C., 1993. Effects of fenoxycarb, JH mimetic, on female sex-ual behaviour of the black cutworm, Agrotis ipsilon (Lep: Noctuidae). Journal of Insect Physiology 39, 25–29.

Gadenne, C., Renou, M., Sreng, L., 1993. Hormonal control of phero-mone responsiveness in the male black cutworm, Agrotis ipsilon. Experientia 49, 721–724.

Gadenne, C., Picimbon, J.F., Be´card, J.M., Lalanne-Cassou, B., Renou, M., 1997. Development and pheromone communication systems in hybrids of Agrotis ipsilon and Agrotis segetum (Lepidoptera: Noctuidae). Journal of Chemical Ecology 23, 191–209.

Gemeno, C., Haynes, K.F., 1998. Chemical and behavioral evidence for a third pheromone component in a north american population of the black cutworm moth, Agrotis ipsilon. Journal of Chemical Ecology 24, 999–1011.

Gemeno, C., Anton, S., Zhu, J.W., Haynes, K.F., 1998. Morphology of the reproductive system and antenna1 lobes of gynandromorphic and normal black cutworm moths, Agrotis ipsilon (Hufnagel) (Lepidoptera: Noctuidae). International Journal of Insect Mor-phology and Embryology 27, 185–191.

Gronenberg, W., Heeren, S., Ho¨lldobler, B., 1996. Age-dependent and task-related morphological changes in the brain and the mushroom bodies of the ant Camponotus floridanus. Journal of Experimental Biology 199, 2011–2019.

Hansson, B.S., Christensen, T.A., Hildebrand, J.G., 1991. Functionally distinct subdivisions of the macroglomerular complex in the antennal lobe of the male sphinx moth Manduca sexta. Journal of Comparative Physiology 312, 264–278.

Hansson, B.S., Anton, S., Christensen, T.A., 1994. Structure and func-tion of antennal lobe neurons in the male turnip moth, Agrotis

sege-tum. Journal of Comparative Physiology 175, 547–562.

Hansson, B.S., 1995. Olfaction in Lepidoptera. Experientia 51, 1003–1027.

Hartlieb, E., Anton, S., Hansson, B.S., 1997. Dose-dependent response characteristics of antenna1 lobe neurons in the male moth Agrotis

segetum (Lepidoptera: Noctuidae). Journal of Comparative

Physi-ology 181, 469–476.

Hill, A.S., Rings, R.W., Swier, S.R., Roelofs, W.L., 1979. Sex phero-mone of the black cutworm moth, Agrotis ipsilon. Journal of Chemical Ecology 5, 439–457.

King, L.E., Zhang, J., Tobe, S.S., 1994. Cytosolic and nuclear juvenile hormone-binding proteins from the brain of Diploptera punctata. General and Comparative Endocrinology 94, 11–22.

Lo¨fstedt, C., van der Pers, J.N.C., Lo¨fqvist, J., Lanne, B.S., Appelgren, M., Bergstro¨m, G., Thelin, B., 1982. Sex pheromone components of the turnip moth, Agrotis segetum: Chemical identification, elec-trophysiological evaluation and behavioral activity. Journal of Chemical Ecology 8, 1305–1321.

Mercer, A., Brown, S., Dunston, C., Kirchhof, B., Kokay, I., Perk, C., Sigg, D., Winnington, A., 1999. Behaviour-related plasticity in the

antennal lobes of the honey bee, Apis mellifera L. In: Elsner, N., Eysel, U. (Eds.), From Molecular Neurobiology to Clinical Neuros-cience. Thieme, Stuttgart, p. 129.

Picimbon, J.F., Gadenne, C., Be´card, J.M., Cle´ment, J.L., Sreng, L., 1997. Sex pheromone of the french black cutworm moth, Agrotis

ipsilon (Lepidoptera: Noctuidae): identification and regulation of a

multicomponent blend. Journal of Chemical Ecology 23, 211–230. Poitout, S.H., Bue`s, R., 1974. Elevage de plusieurs espe`ces de le´pi-dopte`res sur milieu artificiel simplifie´. Annales de Zoologie et d’Ecologie Animales 2, 79–91.

Renou, M., Gadenne, C., Tauban, D., 1996. Electrophysiological inves-tigations of pheromone-sensitive sensilla in the hybrids between two moth species. Journal of Insect Physiology 42, 267–277. Sigg, D., Thompson, C.M., Mercer, A.R., 1997. Activity-dependent

changes to the brain and behavior of the honey bee, Apis mellifera (L.). Journal of Neuroscience 17, 7148–7156.

Stout, J., Atkins, G., Zacharias, D., 1991. Regulation of cricket phono-taxis through hormonal control of the threshold of an identified auditory neuron. Journal of Comparative Physiology 169, 765–772. Stout, J., Hao, J., Kim, P., Mbungu, D., Bronsert, M., Slikkers, S., Maier, J., Kim, D., Bacchus, K., Atkins, G., 1998. Regulation of the phonotactic threshold of the female cricket, Acheta domesticus: juvenile hormone III, allatectomy, L1 auditory neuron thresholds and environmental factors. Journal of Comparative Physiology 182, 635–645.

Sokal, R.R. Rohlf, F.J., 1995. In: Freeman, W.H. (Eds.), Biometry, 3rd Edn. New York, p. 715.

Stout, J., Bronsert, M., Atkins, G., 1999. Neural plasticity in the post-imaginal development and regulation of phonotactic responsiveness of the cricket Acheta domesticus: behavioral plasticity. In: Elsner, N., Eysel, U. (Eds.), From Molecular Neurobiology to Clinical Neuroscience. Thieme, Stuttgart, p. 144.

Strambi, C., Cayre, M., Strambi, A., 1999. Neural plasticity in the adult insect brain and its hormonal control. International Review of Cytology 190, 137–174.

Swann, J.M., 1997. Gonadal steroids regulate behavioral responses to pheromone by actions on a subdivision of the medial preoptic nucleus. Brain Research 750, 189–194.

Vogt, R.G., Rybczynski, R., Cruz, M., Lerner, M.R., 1993. Ecdysteroid regulation of olfactory protein expression in the developing antenna of the tobacco hawk moth, Manduca sexta. Journal of Neurobiol-ogy 24, 581–597.

Wakamura, S., Struble, D.L., Matsuura, H., Sato, M., Kegasawa, K., 1986. Sex pheromone of the black cutworm, Agrotis ipsilon: attractant synergist and improved formulation. Applied Ento-mology and Zoology 21, 299–304.

Walikonis, R., Schoun, D., Zacharias, D., Henley, J., Coburn, P., Stout, J., 1991. Attractiveness of the male Acheta domesticus calling song to females. III. The relation of age-correlated changes in syllabe period recognition and phonotactic threshold to juvenile hormone III biosynthesis. Journal of Comparative Physiology 169, 751–764. Wu, W.-Q., Hansson, B.S., Lo¨fstedt, C., 1995. Electrophysiological and behavioural evidence for a fourth sex pheromone component in the turnip moth, Agrotis segetum. Physiological Entomology 20, 81–92.

Wyatt, G.R., Davey, K.G., 1996. Cellular and molecular actions of juvenile hormone. II. Roles of juvenile hormone in adult insects. In: Evans, P.D. (Ed.), Advances of Insect Physiology, vol. 26. Aca-demic Press, pp. 1–157.

(1)

Table 2

Statistical analysis of the effects of the stimuli on the response of AL neurons as a function of age and JH treatmenta Stimuli

Males A. ipsilon blend A. segetum blend Z7-12:OAc Z9-14:OAc Z11-16:OAc

Day-1 a ab b c c

Day-3 a a b b c

Day-5 a ab b c c

2CA+oil a a b c c

2CA+JH, day-4 a b b bc c

2CA+JH, day-5 a b c d e

+JH, day-l a a b c c

+JH, day-2 a a b c c

a_{For each line, columns with same letters are not statistically different (G-test, P}

,0.05). See Figs. 5, 7 and 9.

males was already detectable one day post-treatment

(G=8.43, df=1, P

,

0.01; G=0.97, df=1, NS; G=1.54;

df=1, NS between injected and 1-, 3- and 5-day-old

males respectively). When JH was injected into young

day-0 immature males, the two blends and the single

components induced different responses of AL neurons

in injected males when tested after two days (Fig. 9). A

high proportion of low threshold neurons was observed

for both blends, different from 1-day-old control males

(G=40.37 and G=18.64; df=3; P

.

0.001 for A. ipsilon

and A. segetum blends respectively) but not different

from 5-day-old males (G=4.9 and G=0.89 df=3; NS for

A. ipsilon and A. segetum respectively). Two types of

responses were observed after stimulation with the single

components of JH injected males: the proportion of low

threshold AL neurons in response to Z7-12:OAc,

although not different from 1-day-old males (G=7.22,

df=3, NS), reached the level of 5-day-old males (G=7.64,

(2)

Fig. 6. Effect of juvenile hormone on pheromone responses of antenna lobe (AL) neurons in mature males. Percentage of tested AL neurons responding to 1 female gland extract equivalent in mature A.

df=3, NS). The proportion of low threshold neurons in

response to the other two single components was not

affected by JH: there was no significant difference

between 1-day-old control males and JH injected males

for Z9-14:OAc (G=3.10, df=3, NS) and Z11-16:OAc

(G=0.84, df=3, NS). The increase of the proportion of

low threshold AL neurons in response to the blends and

to the single components was already visible one day

following treatment and was more pronounced after two

days (Fig. 9).

The responses of the AL neurons were similar for the

two blends one and two days following JH injection, but

were different from responses to Z7-12:OAc. Responses

to both Z9-14:OAc and Z11-16:OAc, in turn, were

dif-ferent from responses to all other stimuli (Table 2).

3.5. Tuning of neurons changing with treatment

The proportion of AL neurons responding more

sensi-tively to the conspecific pheromone blend was generally

higher than the proportion of neurons responding more

sensitively to the blend of the related species, A. segetum

(Fig. 10), with the exception of neurons in 3-day-old

males, which responded more sensitively to the A.

sege-tum blend than to the A. ipsilon blend. One-day-old JH

injected males, 1-day-old control males, and to a lesser

extent allatectomized males, possessed almost equal

pro-portions of AL neurons responding more sensitively to

either pheromone blend (Fig. 10). Five-day-old control

males and, more pronounced, males which had

under-gone a JH injection other than the above mentioned, had

a high proportion of AL neurons which responded more

sensitively to the A. ipsilon blend and a very low

pro-portion of neurons responding more sensitively to the A.

segetum blend (Fig. 10).

4. Discussion

4.1. Neural processing of sex pheromone information

in the AL of Agrotis ipsilon

The most common neuron types found in A. ipsilon

were component-specific neurons responding to

Z7-12:OAc and neurons with medium specificity. The few

blend specific neurons had high response thresholds,

while all other neuron types were found at different

sen-sitivity levels. We did not, however, test systematically,

whether the specificity of highly sensitive neurons

changed with increasing stimulus amounts.

Relatively low sensitivity of blend specific neurons

has been found before in other moth species, while

gen-eralist neurons often exhibit high sensitivity (Ostrinia

nubilalis, Anton et al., 1997; A. segetum, Hartlieb et al.,

1997). In both moth species mentioned, specificity of

neurons often changed with increasing stimulus

concen-trations. The relatively low number of

component-spe-cific neurons found in A. segetum (Hansson et al., 1994;

Hartlieb et al., 1997) and Spodoptera littoralis (Anton

and Hansson 1994, 1995), compared to the present study

in A. ipsilon, might be explained by the fact that we used

lower stimulus concentrations. As specificity is often lost

with increasing stimulus concentrations, neurons with

specific responses were possibly missed in the former

studies. As in our study, a large proportion of

compo-nent-specific neurons have also been demonstrated in the

sphinx moth, Manduca sexta (Christensen and

Hildeb-rand, 1987; Hansson et al., 1991).

The large number of Z7-12:OAc-specific neurons

compared to neurons specifically responding to the other

pheromone components might be due to the high

pro-portion of olfactory receptor neurons (ORNs) responding

to this component. More than 60% of the sensilla

investi-gated on the antenna of A. ipsilon house a receptor

neu-ron for this component, while only few ORNs were

found to respond to Z9-14:OAc and no ORNs have so

far been found to respond to Z11-16:OAc (Renou et al.,

1996). The present study demonstrates that ORNs

responding to Z11-16:OAc must exist on the antenna, as

we found AL interneurons responding to this component.

The response thresholds for the different stimuli

varied within each age/treatment class. Thresholds were

in general lowest for the blends, slightly higher for

Z7-12:OAc, and even higher for the two other single

compo-nents. Different convergence of ORNs, as mentioned

(3)

above, can be responsible for such differences in AL

neuron sensitivity.

The responses of AL neurons to the female gland

extract-equivalent were very similar to the responses of

the same neurons to the artificial pheromone blend at a

dose of 1 ng, resembling the female gland production

(Picimbon et al., 1997). This fact and its efficiency in

behavioural experiments (Hill et al., 1979; Wakamura et

al., 1986; Picimbon et al., 1997; Gemeno and Haynes,

1998) show that the artificial pheromone blend can

replace the natural blend.

4.2. Effect of age on central nervous processing

The present results show that the proportions of low

threshold AL neurons does not only increase with age

in reponse to the species specific blend (Anton and

Gad-enne, 1999), but also to the blend of a related species

and to single pheromone components. This effect of age

on the proportion of low threshold AL neurons is much

more pronounced following the blend stimulations than

with

single

component

stimulations.

Clearly,

the

increase of the proportion of low threshold neurons with

age was more pronounced for stimulations of

Z7-12:OAc than with that of Z9-14:OAc and of

(4)

16:OAc. The number of non-responding neurons to

stimulations with Z9-14:OAc and with Z11-16:OAc was

very high whatever the age of the male tested. We

pre-viously found an increase in the proportion of low

thres-hold AL neurons with age when the antenna was

stimu-lated with the A. ipsilon pheromone blend (Anton and

Gadenne, 1999).

The major changes in the proportion of low threshold

neurons for single components and the blends occur

between l-day old and 3-day old moths. Plasticity,

how-ever, seems to be more pronounced for the

species-spe-cific blend (significant increase of the proportion of low

threshold neurons even between 3-day old and 5-day old

males) than for single components and the blend of

another species. For the minor component Z11-16:OAc,

no significant changes of the proportion of low threshold

neurons were found during development, which might

either be due to missing plasticity or might simply be

difficult to detect, as there are so few neurons responding

to that component in general.

Various behaviours controlled by physiological

pro-cesses and correlated with certain anatomical features in

the brain have been shown to be age-dependent in

insects. In bees, the volume of the AL glomeruli and of

the mushroom bodies changes with age and this growth

is linked with changes in tasks performed by the bees

(Fahrbach and Robinson, 1996; Sigg et al., 1997). In the

ant, Camponotus floridanus, changes in the size of the

AL and other parts of the brain were shown to be

corre-lated with age and task (Gronenberg et al., 1996). In the

(5)

cricket, Acheta domesticus, the phonotactic response and

linked with it, the response threshold of auditory

neu-rons, were found to be age-dependent (Stout et al. 1991,

1998; Walikonis et al., 1991).

4.3. Effects of JH treatments on central nervous

processing

JH acts differentially on the response of AL inter

neu-rons to various pheromone stimuli. Our results from

JH-manipulated male moths were clearly correlated with the

age-dependant changes in proportions of AL neurons

with different response thresholds observed.

Hormones play a role in many physiological processes

occurring in the brains of vertebrates and invertebrates.

In all above cited examples of age-dependent effects on

the insect brain, JH was found to play a role in this

regu-lation. However, in the case of the honeybee,

age-depen-dent AL development was only indirectly JH-depenage-depen-dent

by changing the bees’ activity, which in turn acted on

the AL anatomy (Fahrbach and Robinson, 1996; Sigg et

al., 1997), most likely through biogenic amines as

mediators (Mercer et al., 1999). Another insect hormone,

ecdysone, was shown to act on the antennal development

of M. sexta through the control of the antennal protein

pattern (Vogt et al., 1993). In mammals, steroids have

been shown to play a role in the integration of

phero-monal stimuli by maintaining the responsiveness of

olfactory brain centres, i.e. the magnocellular division of

the medial preoptic nucleus (Swann, 1997). Moreover,

melatonin was recently shown to act on the volume of

the song-control nucleus areas in the bird brain (Bentley

et al., 1999).

The effect of JH injection in both allatectomized and

young A. ipsilon males was gradual and rather slow.

Some changes in the proportion of AL neurons with

dif-ferent thresholds were observed after one day, other

changes did not appear until two days after JH injection.

Similarly,

neurogenesis

in

allatectomized

crickets

injected with JH was observed only after three days

(Cayre et al., 1994). Somewhat faster action of JH was

found with topical application of JH in crickets. In

female house crickets, in which JH was topically

applied, an effect on phonotactic behaviour was

observed within 12 hours post-treatment in young

crick-ets, while application on old allatectomized cricket

females was effective after 6 hours (Stout et al. 1991,

1999). Blocking of transcription and translation factors

abolished the effect of JH, which indicates a nuclear

effect (Stout et al., 1999). Another indication for nuclear

effects of JH are cytosol and nuclear JH-binding proteins

which were found in the brains of the cockroach

Diplop-tera punctata (King et al., 1994).

4.4. Tuning to the species specific blend

AL neurons of young males and allatectomized older

males do not discriminate between the sex pheromone

blend produced by conspecific females and the A.

sege-tum blend. In mature control males and in JH injected

males, more neurons show a higher sensitivity to the A.

ipsilon blend than to the A. segetum blend. This makes

sense, because at the time when A. ipsilon males are

fully mature, the tuning of olfactory centres must be

optimal in order to avoid interspecific matings with other

close species such as A. segetum. These two species are

sympatric in various parts of their geographical habitat.

They do hybridize in laboratory conditions (Gadenne et

al., 1997) and possess very similar pheromone blends.

Acknowledgements

We thank E. Warrant, B.S. Hansson and two

anony-mous referees for their valuable comments on the

manu-script. This work was supported by grants from Institut

National de la Recherche Agronomique (INRA), the

Swedish Council for Forestry and Agricultural Research

(SJFR), and the Swedish Council for Natural Science

(NFR).

References