Introduction

Adaptive behavior requires that the nervous system be correctly wired, a process that involves complex interactions at both cellular and molecular levels. Neurons establishing the very first axonal projections, so-called pioneer neurons, have attracted particular interest because their growth cones are able to navigate without the aid of cellular cohorts. Following the original observations of Ramón y Cajal (1890) on the chick neural tube, a rich literature has developed focussing on the mechanisms regulating growth cone dynamics and guidance (see Tessier-Lavigne and Goodman 1996; Mueller 1999 for reviews). The peripheral nervous system of insects, in particular, has proven to be an advantageous model system for the study of pioneers as such cells have been shown to be individually identifiable. First reported in the appendages of the embryonic grasshopper (Bate 1976), the axons of these pioneers were described as providing a scaffold for the projections of sensory cells differentiating within the epithelium and navigating to the central nervous system. Since then, the insect appendage has continued to serve as a source of information on the cellular and molecular mechanisms of growth cone guidance (see Bentley and O’Connor 1992; Goodman 1996).

In our present study, we focus on the embryonic antennal nervous system of the grasshopper Schistocerca gregaria. Two pairs of sibling cells located at the tip of the antenna have been identified as pioneering the twin pathways—one dorsal and one ventral—which carry the axonal projections of sensory cells differentiating from the epithelium to the antennal lobe in the brain (Ho and Goodman 1982; Berlot and Goodman 1984; Seidel and Bicker 2000; Boyan and Williams 2004). These tip pioneers delaminate from the epithelium into the antennal lumen (Bate 1976; Ho and Goodman 1982), and their growth cones then navigate along the basal lamina towards the antennal base. En route, the pioneer growth cones of one pathway have been shown to contact a guidepost-like cell called the "base pioneer" stereotypically located nearer the antennal base (Ho and Goodman 1982; Berlot and Goodman 1984; Seidel and Bicker 2000). Both the tip pioneers and this base pioneer are known to express the neuron-specific antigen horseradish peroxidase (Ho and Goodman 1982; Seidel and Bicker 2000). The tip pioneers have additionally been shown to co-express the highly glycosylated cell surface lipocalin Lazarillo (Boyan and Williams 2004) which labels pioneer neurons elsewhere in the grasshopper nervous system (Ganfornina et al. 1995) and where it regulates directed axogenesis (Sánchez et al. 1995). In contrast to the tip pioneers, this base pioneer has an unknown origin and its role in axon guidance has also not yet been clearly established. Further, only a single base pioneer has been identified to date so that the cellular guidance cues regulating axogenesis in the other antennal pathway remain obscure.

We therefore re-examined aspects of pioneer cell biology in the early embryonic antenna using immunolabeling and immunoblocking techniques with the aim of resolving these various anomalies and, in particular, the role of the base pioneer in directed axogenesis.

Materials and methods

Immunolabeling for neuron-specific horseradish peroxidase (HRP, Jan and Jan 1982), for the highly glycosylated cell surface lipocalin Lazarillo (Ganfornina et al. 1995; Sánchez et al. 1995), for the epithelial cell surface marker Lachesin (Karlstrom et al. 1993), and for the mesodermal marker Mes3 (Kotrla and Goodman 1984) was undertaken essentially as previously described (Boyan and Williams 2004, 2007; Boyan et al. 2010).

For immunoblocking experiments, staged embryos (see Boyan and Williams 2004) were dissected out of the egg into sterile culture medium which consisted of Grace’s Insect TC Medium (Bio & SELL 2.12G07J), 600 mg/l l-glutamine (Roth), 20 μl/ml 20-hydroecdysone (Sigma), 25 μg/ml juvenile hormone III (Sigma-Aldrich), 120 units/ml penicillin, 120 units/ml streptomycin, and 2.5 μg/ml Fungizone (Gibco). A control batch of five embryos was allowed to grow in this medium under sterile conditions at 30 °C (whole embryo culture). Several experimental batches of five embryos of the same age (32 %) were grown in parallel in the same culture medium, but to each of which, a different dosage of Mab 10E6 (anti-Lazarillo) had been added. The data presented here refer to an antibody concentration of 2 μM. This was within the concentration spectrum previously shown to be effective in disrupting growth cone guidance in the ventral nerve cord of the grasshopper and over the same embryonic ages (Sánchez et al. 1995). Embryos in control and experimental groups grew an average of 4 % (from 32 to 36 %) over the course of the experiment (2 days). Control and experimental embryos were subsequently processed for HRP-labeling. Fluorescence and confocal microscopy was as previously described, as was subsequent image processing (Boyan and Williams 2004, 2007).

Results and discussion

Pioneer cells of the antennal nervous system

The antennal nervous system of the grasshopper S. gregaria features two parallel nerve pathways running from its tip to the antennal lobe in the brain and containing the projections of sensory cells differentiating from its epithelium (see Chapman 2002). These pathways are already established by mid-embryogenesis (Fig. 1a; Ho and Goodman 1982; Berlot and Goodman 1984; Seidel and Bicker 2000; Boyan and Williams 2004). Constructing this pathway begins at around 30 % of embryogenesis and involves sets of pioneer neurons located at stereotypic locations in the antenna (Fig. 1b; Bate 1976; Ho and Goodman 1982; Berlot and Goodman 1984; Seidel and Bicker 2000; Boyan and Williams 2004). These previous studies demonstrated the presence of two pairs of sibling pioneers (one pair ventral, the other pair dorsal) which differentiate at the antennal tip and express the neuron-specific epitope horseradish peroxidase (Jan and Jan 1982). Molecular expression patterns involving the antigens Lachesin (Fig. 1a), Annulin (Bastiani et al. 1992), and Lazarillo (Boyan and Williams 2004; see Fig. 3b), as well as the stereotypic locations of Repo-positive glia cells (Boyan and Williams 2004), demonstrate that the flagellum of the early embryonic antenna is segmented into three so-called meristal annuli (A1–3). The number of these annuli increases during embyrogenesis and continues postembryonically to ultimately regulate the distribution of sensory cells in the epithelium (Chapman 2002). Consistent with the epithelium of the antennal flagellum being segmented, additional pioneer cells like those at the tip (the A1 annulus) have been identified in other annuli (e.g., A2 and A3) and contribute their axons to the construction of the initial nerve pathways to the brain (Boyan and Williams 2004). Our present study focusses on navigational decisions being made by the growth cones of the tip pioneers from the A1 annulus of the antenna.

Fig. 1
figure 1

Pioneer cells involved in axogenesis of the antennal nervous system. a Confocal image at 45 % of embryogenesis shows the flagellum of the antenna following double immunolabeling against neuron-specific horseradish peroxidase (α-HRP) and epithelial cell-specific Lachesin (α-Lach). The antennal nervous system involves two parallel nerve pathways: one ventral (vN) and the other dorsal (dN), running from the tip of the antenna to its base and then into the brain (not shown). The approximate border between the outer ectodermal epithelium (Ep) and inner mesodermal lumen (Lu) is indicated (dashed white line). Each nerve comprises the axonal projections of sensory cells (white arrowheads) developing in the epithelium. There is no motor innervation because the flagellum does not possess a musculature. Early segmentation of the flagellum in the form of meristal annuli (A1, A2, and A3) is evident via the Lachesin expression and represents a progressive subdivision which continues postembryonically (see Chapman 2002). b Confocal image of the early embryonic antenna (32 %) following double immunolabeling (α-HRP, α-Lach) shows ventral and dorsal pioneer neurons (vA1P and dA1P) from the A1 annulus at the tip establishing the initial axon scaffold of the antennal nervous system. Each tip pioneer targets its own so-called base pioneer (vBP, dBP) before projecting to the brain (not shown). Inset (white rectangle) shows the dBP (white star, soma outlined dashed white) at higher magnification and extending HRP-positive filopodial processes (white arrowheads). A1 tip pioneers occur as sibling cells but may, on occasion, lie directly superimposed preventing individual imaging. The approximate border between the outer ectodermal epithelium (Ep) and inner mesodermal lumen (Lu) is indicated (dashed white line). c Confocal image following α-HRP immunolabeling at 34 % of embryogenesis reveals putative filopodial contacts between an HRP-positive ventral tip pioneer (vA1P, white arrowheads) and its target HRP-positive base pioneer (vBP, red arrowheads). d Confocal image following immunolabeling (α-HRP) at 36 % of embryogenesis shows the growth cone and filopodia of an HRP-positive dorsal tip pioneer (dA1P, white arrowheads) putatively contacting the filopodia (red arrowheads) and soma of its target base pioneer (dBP). For clarity images c, d have been rotated to the same orientation. Scale bar in a represents 50 μm in a, b; 20 μm in b, inset; 20 μm in c; and 15 μm in d

Full size image

In addition to these pioneers, a single HRP-positive base pioneer (BP) located nearer the antennal base was shown to be contacted by the filopodia of the (A1) tip pioneers (Ho and Goodman 1982; Berlot and Goodman 1984; Seidel and Bicker 2000). Our current study confirms the neuronal phenotype for these cells (Fig. 1b). However, in contrast to previous studies, we show that the antenna possesses not just a single BP associated with the ventral pathway (vBP), but an additional HRP-positive BP associated with the dorsal pathway (dBP) (Fig. 1b inset). Further, the BPs on each side are not in equivalent locations along the antennal axis: the dBP is consistently closer to the base than is the vBP. This means that the growth cones of the tip pioneers must navigate different distances in order to contact their respective BPs. Comparisons (Fig. 1c, d) confirm putative filopodial contacts between both the ventral (vA1) and dorsal (dA1) tip pioneers and their respective BP targets. Further, filopodial intermingling occurs slightly earlier in the ventral pathway (34 %) than the dorsal pathway (36 %). Differential growth patterns consistent with the data we present here have previously been reported for the ventral and dorsal tip pioneers (Ho and Goodman 1982; Berlot and Goodman 1984; Seidel and Bicker 2000).

Summarizing, we now show that there is a BP possessing filopodia associated with each antennal tract and that these filopodia intermingle with those of the respective dA1/vA1 pioneers from the antennal tip. Such putative reciprocal filopodial contacts could provide a morphological substrate for a molecular recognition process which we investigate further below.

Origin of pioneers

In arthropod appendages, two tissues have been identified as generating proliferative cells. The first is an outer epithelium of ectodermal origin which, early in embryogenesis, is characterized by a posterior band of cells expressing the segment polarity gene engrailed (Patel et al. 1989; Boyan and Williams 2004) and later generates the cuticle (Chapman 1982). Sensory cells differentiate within this outer epithelium (Chapman 2002), and pioneer neurons delaminate from it into the lumen at stereotypic locations in both the leg (Keshishian 1980) and antenna (Bate 1976; Ho and Goodman 1982; Berlot and Goodman 1984; Boyan and Williams 2004). Following delamination, such pioneers lie closely adhered to the developing basal lamina which forms the interface between the outer epithelium and the inner lumen. The second tissue is the mesoderm which is represented in the antenna by the lumen and contains the coelom, the future median strand, hemocyte precursors (c.f. Ball et al. 1987 for the leg), dendritic-like cells possessing filopodial extensions (Berlot and Goodman 1984), nerve tract associated cells (Boyan and Williams 2007), Repo-positive glia (Boyan and Williams 2004), and in the leg and at the antennal base, muscle pioneers (Ho et al. 1982). Consistent with their mesodermal origin, lumenal cells such as nerve tract associated cells and hemocytes are labeled by the monoclonal antibody against the cell surface antigen Mes3 (methylesterase, Kotrla and Goodman 1984; Kuwada and Goodman 1985; Boyan and Williams 2007).

Whereas the tip pioneers of the antenna (Berlot and Goodman 1984) and leg (Keshishian 1980) have been shown to delaminate from the epithelium, the origin of the antennal base pioneer(s) has not been addressed. We therefore used Mes3 expression to distinguish cells of mesodermal from those of epithelial (ectodermal) origin and so establish whether these various pioneers derive from a common or different embryonic Anlagen.

Double-labeling with antibodies against HRP and Mes3 (Fig. 2a) reveals the dA1/vA1 pioneers of the antennal tip to be HRP-positive/Mes3-negative, confirming they are neurons and derive from the epithelium (ectoderm). This same double-labeling also confirms the presence of two BPs, one for each antennal pathway (Fig. 2a). Rotating the confocal Z-stack to a side view (Fig. 2b) clearly shows the ventral tip pioneer extending a process to a ventral BP and the dorsal tip pioneer likewise extending a process to a dorsal BP. Further, both the BPs appear to be HRP-positive/Mes3-positive, consistent with a derivation from the mesoderm and not the ectoderm. Confocal images at higher resolution (Fig. 2c) confirm that both BPs, but not the A1 tip pioneers, are Mes3-positive and, therefore, derive from the mesoderm which is represented by the lumen of the antenna.

Fig. 2
figure 2

Molecular markers confirm different ontogenies for tip and base pioneers. a Confocal image of the early embryonic antenna (32 %) following double immunolabeling with the neuron-specific marker HRP and the mesodermal marker Mes3. Tip pioneers (vA1P, dA1P) appear HRP-positive/Mes3-negative, whereas the base pioneers (vBP, dBP) appear HRP-positive/Mes3-positive. Mes3-positive putative hemocytes (white stars) likely secrete basement membrane components for the appendages (Ball et al. 1987). b Confocal image of the same antenna as in a but in side view confirms that there is a tip and base pioneer associated with each of the ventral (vA1P, vBP) and dorsal (dA1P, dBP) pathways. Growth cones of the vA1P (black arrowhead) and dA1P (white arrowhead) are in the process of establishing contact with their respective target base pioneers. c Confocal images following double labeling with the neuron-specific marker HRP and the mesodermal marker Mes3 reveal both the vBP (i–iii) and dBP (iv–vi) that are HRP-positive/Mes3-positive, consistent with a derivation from the mesoderm. The representative tip pioneer (dA1P) is HRP-positive/Mes3-negative (vii–ix), consistent with a derivation from the ectoderm. Scale bar in a represents 25 μm in a, b and 15 μm in c

Full size image

The base pioneers in axon guidance

Lazarillo is a highly glycosylated glycoprotein with internal disulfide bridges bound to the extracellular side of the cell membrane via a glycosylphosphatidylinositol anchor (Cross 1990). In a study of axogenesis in the ventral nerve cord of the grasshopper S. gregaria, Ganfornina et al. (1995) and Sánchez et al. (1995) used the expression pattern of the Lazarillo antigen to identify a subset of neurons which pioneer specific axon pathways in each segmental ganglion. Removal of such glycosylphosphatidylinositol-anchored cell surface proteins (Chang et al. 1992) or immunoblocking of the Laz epitope (Sánchez et al. 1995) disrupts pioneer growth cone guidance in the embryonic grasshopper nervous system.

In the early embryonic antenna, double immunolabeling reveals that the tip pioneers (Fig. 3a (i–vi)) and base pioneers (Fig. 3b (i–vi)) are HRP-positive/Lazarillo-positive, consistent with their being pioneer neurons. However, the temporal expression pattern of the Lazarillo antigen differs among the various pioneers. Whereas Lazarillo expression in the tip pioneers is continuous during embryogenesis (Fig. 3a (iii, vi) and see Boyan and Williams (2004) for later ages), the base pioneers initiate Lazarillo expression at the time their filopodia and those of the tip pioneers first recognize one another (Fig. 3b (iii)). They subsequently downregulate this antigen (Fig. 3b (ii, v)) although they still express the HRP antigen (Fig. 3b (i, iv)) allowing them to be identified. However, Seidel and Bicker (2000) report that even the HRP antigen is also subsequently downregulated in their one base pioneer, suggesting parallel molecular fates with respect to HRP and Lazarillo.

Fig. 3
figure 3

Dynamic expression of a cell surface antigen involved in axon guidance by pioneer neurons. a Confocal images of representative tip pioneers (vA1P) following double labeling with the neuron-specific marker HRP and the cell surface guidance molecule Lazarillo (Laz) at two different ages (33 and 35 %) during embryogenesis. At 33 % (i–iii) and 35 % (iv–vi), the vA1P co-expresses HRP and Lazarillo. b Confocal images of representative base pioneers (vBP) from the same antennal pathway as in a following double labeling with the neuron-specific marker HRP and the cell surface guidance molecule Lazarillo (Laz) at two different ages (33 and 35 %) during embryogenesis. The vBP co-expresses HRP and Lazarillo at 33 % (i–iii) but downregulates Laz expression at 35 % (iv–vi, white star in v indicates soma location). Scale bar represents 5 μm

Full size image

In addition to guidepost cells and molecular gradients (see Bentley and O’Connor 1992; Goodman 1996 for reviews), a number of other factors such as the basal lamina (Berlot and Goodman 1984; Anderson and Tucker 1988; Condic and Bentley 1989) and the neuromodulator nitric oxide (NO) (Seidel and Bicker 2000) have been shown to regulate axon guidance in the peripheral nervous system of the leg or antenna. We examined whether Lazarillo also contributes to axon guidance during early axogenesis in the antennal nervous system, consistent with its role in the ventral nerve cord (Sánchez et al. 1995). Immunoblocking of the Lazarillo epitope under whole embryo culture conditions was indeed found to disrupt axogenesis in the early embryonic antenna (Fig. 4). Blocking Lazarillo expression caused the growth cones of both dorsal (Fig. 4b) and ventral (Fig. 4c) tip pioneers to miss their correct basal pioneers despite their maintained physical presence (c.f. Fig. 4a). We interpret this as a failure to molecularly identify the respective targets. In our experiments, the growth cones of tip pioneers from both pathways stalled and extended uncharacteristic wide-ranging filopodia in an apparent search process. This response mirrors the delayed growth and altered filopodial projection pattern observed in pioneer growth cones of the ventral nerve cord as reported in a previous study whose protocol we followed (Sánchez et al. 1995). Further culturing experiments with older antennae must determine whether this disruption represents a permanent growth arrest or a slowing of growth as also occurs in NO-regulated guidance of axogenesis by these same pioneers (Seidel and Bicker 2000).

Fig. 4
figure 4

Disrupted target recognition following antibody blocking. a Normal growth. Fluorescence microscope image of the antenna (edge is outlined white) from an embryo grown in culture for 2 days (from 32 % up to 36 % of embryogenesis) and subsequently immunolabeled with neuron-specific HRP. The growth cone filopodia (white arrowheads) of the sibling tip pioneers (vA1P) have contacted a target HRP-positive vBP en route to the brain (not shown). Other putative guidepost cells (white stars) are also labeled. b, c Disrupted target recognition in the antennae of embryos of the same age and grown under the same conditions, as in a except that Lazarillo antibody was added to the initial culture medium (see Materials and methods). b Fluorescence microscope image following neuron-specific HRP immunolabeling shows a sibling pair of dA1P neurons whose growth cone appears to have stalled and whose extensive filopodia (white arrowheads) have apparently failed to recognize the target dBP. c Confocal image following neuron-specific HRP immunolabeling shows a sibling pair of vA1P neurons whose respective growth cones (open, solid white stars) have by-passed their target vBP, stalled, and now extend wide-ranging filopodia (open, solid white arrowheads) in search of the target. Antenna orientation in a applies to all panels. Scale bar in c represents 18 μm in a, 17 μm in b, and 15 μm in c

Full size image

Our study demonstrates for the first time that pioneer neurons of the peripheral nervous system in an insect appendage represent a heterogeneous population with respect to both their origin from the epithelium (A1 tip pioneers) or mesoderm (base pioneers) and temporal molecular expression patterns. A range of other questions concerning the base pioneers of the grasshopper antennal system remain open and form the subject of future studies. These include the exact mechanism of their neurogenesis and their fates—whether they die (Berlot and Goodman 1984) or simply downregulate cell surface antigens such as HRP (Seidel and Bicker 2000) and Lazarillo (Fig. 3), rendering them unidentifiable at later ages.