Systematic expression profiling of Dpr and DIP genes reveals cell surface codes in Drosophila larval motor and sensory neurons

Abstract

In complex nervous systems, neurons must identify their correct partners to form synaptic connections. The prevailing model to ensure correct recognition posits that cell-surface proteins (CSPs) in individual neurons act as identification tags. Thus, knowing what cells express which CSPs would provide insights into neural development, synaptic connectivity, and nervous system evolution. Here, we investigated expression of Dpr and DIP genes, two CSP subfamilies belonging to the immunoglobulin superfamily, in Drosophila larval motor neurons (MNs), muscles, glia and sensory neurons (SNs) using a collection of GAL4 driver lines. We found that Dpr genes are more broadly expressed than DIP genes in MNs and SNs, and each examined neuron expresses a unique combination of Dpr and DIP genes. Interestingly, many Dpr and DIP genes are not robustly expressed, but are found instead in gradient and temporal expression patterns. In addition, the unique expression patterns of Dpr and DIP genes revealed three uncharacterized MNs. This study sets the stage for exploring the functions of Dpr and DIP genes in Drosophila MNs and SNs and provides genetic access to subsets of neurons.

Keywords: DIP; Dpr; Motor neuron; Sensory neuron; Synaptic recognition.

Fig. 1.

Schematic of GAL4 insertion and larval body plan. (A) MiMIC or CRIMIC cassettes were inserted into a common intron or 5′UTR to capture the expression of all isoforms for each Dpr and DIP gene. MiMIC insertions were flanked by two attP sites which are later swapped by a GAL4 exon or T2A-GAL4 trojan exon. (B) Drosophila larvae are divided into three thoracic segments and nine abdominal segments, with repeated muscles, MNs and SNs. Muscles are divided into three main groups, the ventral, lateral and dorsal muscles. MNs innervating these muscles are not shown in this diagram. SNs are divided into six main classes: the es neurons, ch neurons, bd neurons, td neurons, md neuron and da neurons (Orgogozo and Grueber, 2005). In addition, da neurons are further divided into da-I, da-II, da-III and da-IV subclasses.

Fig. 2.

Dpr and DIP genes are expressed in various patterns in MNs. (A) Schematic showing the experimental procedure. Each dpr/DIP-GAL4 line was crossed to a real-time reporter (UAS-GFP or UAS-mCherry) and a permanent reporter [UAS-GFP, UAS-FLP, actin-(FRT.STOP)-GAL4] to reveal the dynamic expression of Dpr and DIP genes. (B) Example of a decrease in expression of dpr2-GAL4 in MN1-Ib (arrows) from anterior hemisegment A2 to posterior hemisegment A7. Note that the expression in nearby MN9-Ib (arrowheads) is also not robust as it was not expressed in A2 and A3 but was expressed in A4 to A7. (C) Example of an increase in expression of DIP-ζ-GAL4 in MN16/17-Ib (arrows) from anterior hemisegment A2 to posterior hemisegment A7. Note that the expression in nearby MN15/16-Ib (arrowheads) was always absent. (D) Example of temporal expression of dpr9-GAL4 in MN21-Ib. MN21-Ib was not labeled by dpr9-GAL4>GFP animals, but 50% of MN21-Ib were labeled in the cross to the permanent reporter. Dashed lines indicate muscle boundaries.

Fig. 3.

Expression map of Dpr and DIP genes in all larval MNs. Each column represents an MN including type-Ib, type-Is, type-II, type-III and the alary MN. Expression of each gene in each MN is characterized into a specific category as indicated in the key. Note that MN16/17-Ib was named as MN15/16/17-Ib by Hoang and Chiba (2001); MN6-Ib is represented only in A2 hemisegments (see further characterization below); MN7-Ib is represented only in A2 hemisegments (see further characterization below); and MN23-Ib is a newly identified neuron (see further characterization below).

Fig. 4.

Using the G-TRACE system to probe expression of Dpr and DIP genes in muscles and glial cells. (A) Schematic showing the cross between dpr/DIP-GAL4 and the G-TRACE reporter. Red signal represents real-time GAL4 expression and green signal represents earlier GAL4 expression. (B,C) dpr10 is consistently expressed in most muscles (B) but absent in transverse muscles (C) and some deeper ventral muscles. Expression in some muscles is not consistent. For example, in some hemisegments m5 nuclei are not labeled (arrowhead), but an adjacent hemisegment shows labeling of m5 nuclei (arrows). dpr10 expression is maintained throughout development as revealed by co-labeling with GFP and RFP. (D,E) dpr19 is expressed in all muscles (D), including transverse muscles (E). Compared with dpr10, these nuclei have less RFP intensity, which may indicate that dpr19 is temporally expressed in early development and turned off later. (F) Expression map of dpr10 and dpr19 in muscles.

Fig. 5.

Expression map of Dpr and DIP genes in all larval SNs. Each column represents an SN. Expression of each gene in each SN is characterized into a specific category as indicated in the key.

Fig. 6.

Hierarchical clustering of SNs and MNs reveals shared expression patterns of Dpr and DIP genes in neurons from the same class. (A) SNs from the same class are clustered together based on their expression pattern of Dpr and DIP genes. For example, most es neurons (gray), all chordotonal neurons (purple), and da neurons fall into distinct clusters. (B) Modulatory MNs (II and III) and type-Is MNs are distinct from the main type-Ib cluster. However, individual type-Ib MNs are not easily distinguished based on their expression of Dpr and DIP genes.

Fig. 7.

Differentially expressed Dpr and DIP genes reveal an MN that solely innervates m23. (A) Schematic of transverse muscles 22, 23 and 24 (gray) with previously identified MN22/23-Ib (green), MN23/24-Ib (red) and newly identified MN23-Ib (blue). (B) Representative image showing dpr13-GAL4 expression in both MN23/24-Ib (red arrowheads) and MN23-Ib (blue arrows). Thus, all boutons on m23 and m24 are labeled by GFP. (C) Representative image showing DIP-β-GAL4 expression in MN23-Ib (blue arrow). Boutons underneath m23 and boutons from m22, m24 (red arrowheads) are not labeled by GFP, thus DIP-β-GAL4 is not expressed in MN22/23-Ib and MN23/24-Ib. (D) Representative image showing dpr5-GAL4 expression in MN22/23-Ib and MN23/24-Ib (red arrowheads), but not in MN23-Ib (blue arrow). The lack of GFP in the arbor on m23 indicated the existence of an MN that solely innervates m23.

Fig. 8.

Differentially expressed Dpr and DIP genes reveal MN6-Ib and MN7-Ib in segment A2. (A) Schematic of MN6-Ib (red) and MN7-Ib (blue) in segment A2, and MN6/7-Ib in A3-A7 (green). MN6-Ib preferentially innervates m6 but also forms a small NMJ on m7, whereas MN7-Ib prefers m7 but also forms a small NMJ on m6. (B) Representative images showing that DIP-β, DIP-ε and DIP-γ are specifically expressed in MN6-Ib (red arrows), but not in MN7-Ib (blue arrowheads). Note that MN6-Ib forms boutons with both m6 and m7, as there is a small GFP-positive type-Ib NMJ on m7 (red arrows on m7). Conversely, the lack of GFP in most m7 type-Ib NMJ and the small m6 type-Ib NMJ (blue arrowheads) also indicate dual innervation of both muscles by MN7-Ib. (C) Representative image showing that dpr15 is specifically expressed in MN7-Ib (blue arrowheads) but not in MN6-Ib (red arrows). MN6-Ib and MN7-Ib also show dual innervation patterns in this genetic background.

Fig. 9.

Further characterization of MN6-Ib and MN7-Ib. (A) Quantification of the dual innervation frequencies of MN6-Ib and MN7-Ib: 68.2% of MN6-Ib also innervate m7 and 72.7% of MN7-Ib also innervate m6 (n=21 hemisegments). (B) Quantification of MN6-Ib and MN7-Ib NMJ sizes on both muscles (n=21 hemisegments). (C,D) A pan-MN driver OK6-GAL4 driving MCFO revealed the dendritic morphology of MN6-Ib and MN7-Ib in the VNC. (E,F) Corresponding NMJ images from the same neuron shown in C (MN6-Ib) and D (MN7-Ib).