Summary In the Drosophila optic lobes, 800 retinotopically organized columns in the medulla act as functional units for processing visual information. The medulla contains over 80 types of neurons, which belong to two classes: uni-columnar neurons have a stoichiometry of one per column, while multi-columnar neurons contact multiple columns. We show that combinatorial inputs from temporal and spatial axes generate this neuronal diversity: All neuroblasts switch fates over time to produce different neurons. The neuroepithelium that generates neuroblasts is also sub-divided into six compartments by the expression of specific factors. Uni-columnar neurons are produced in all spatial compartments independently of spatial input; they innervate the neuropil where they are generated. Multi-columnar neurons are generated in smaller numbers in restricted compartments and later move to their final position. The integration of spatial inputs by a fixed temporal neuroblast cascade thus acts as a powerful mechanism for generating neural diversity, regulating stoichiometry and the formation of retinotopy.

Sensory systems use stochastic fate specification to increase their repertoire of neuronal types. How these stochastic decisions are coordinated with the development of their targets is unknown. In the Drosophila retina, two subtypes of ultraviolet-sensitive R7 photoreceptors are stochastically specified. In contrast, their targets in the brain are specified through a deterministic program. We identified subtypes of the main target of R7, the Dm8 neurons, each specific to the different subtypes of R7s. Dm8 subtypes are produced in excess by distinct neuronal progenitors, independently from R7. After matching with their cognate R7, supernumerary Dm8s are eliminated by apoptosis. Two interacting cell adhesion molecules, Dpr11 and DIPγ, are essential for the matching of one of the synaptic pairs. These mechanisms allow the qualitative and quantitative matching of R7 and Dm8 and thereby permit the stochastic choice made in R7 to propagate to the brain.

The ability of neurons to identify correct synaptic partners is fundamental to the proper assembly and function of neural circuits. Relative to other steps in circuit formation such as axon guidance, our knowledge of how synaptic partner selection is regulated is severely limited. Drosophila Dpr and DIP immunoglobulin superfamily (IgSF) cell-surface proteins bind heterophilically and are expressed in a complementary manner between synaptic partners in the visual system. Here, we show that in the lamina, DIP misexpression is sufficient to promote synapse formation with Dpr-expressing neurons and that disrupting DIP function results in ectopic synapse formation. These findings indicate that DIP proteins promote synapses to form between specific cell types and that in their absence, neurons synapse with alternative partners. We propose that neurons have the capacity to synapse with a broad range of cell types and that synaptic specificity is achieved by establishing a preference for specific partners.

Many sensory systems use stochastic fate specification to increase their repertoire of neuronal types. How these stochastic decisions are coordinated with the development of their target post-synaptic neurons in processing centers is not understood. In the Drosophila visual system, two subtypes of the UV-sensitive R7 color photoreceptors called yR7 and pR7 are stochastically specified in the retina. In contrast, the target neurons of photoreceptors in the optic lobes are specified through a highly deterministic program. Here, we identify subtypes of the main postsynaptic target of R7, the Dm8 neurons, that are each specific to the different subtypes of R7s. We show that during development the different Dm8 subtypes are produced in excess by distinct neuronal progenitors, independently from R7 subtype specification. Following matching with their respective R7 target, supernumerary Dm8s are eliminated by apoptosis. We show that the two interacting cell adhesion molecules Dpr11, expressed in yR7s, and its partner DIPγ, expressed in yDm8s, are essential for the matching of the synaptic pair. Loss of either molecule leads to the death of yDm8s or their mis-pairing with the wrong pR7 subtype. We also show that competitive interactions between Dm8 subtypes regulate both cell survival and targeting. These mechanisms allow the qualitative and quantitative matching of R7 subtypes with their target in the brain and thus permit the stochastic choice made in R7 to propagate to the deterministically specified downstream circuit to support color vision.

In the fly optic lobe, $800 highly stereotypical columnar microcircuits are arranged retinotopically to process visual information. Differences in cellular composition and synaptic connectivity within functionally specialized columns remain largely unknown. Here, we describe the cellular and synaptic architecture in medulla columns located downstream of photoreceptors in the dorsal rim area (DRA), where linearly polarized skylight is detected for guiding orientation responses. We show that only in DRA medulla columns both R7 and R8 photoreceptors target to the bona fide R7 target layer where they form connections with previously uncharacterized, modality-specific Dm neurons: two morphologically distinct DRA-specific cell types (termed Dm-DRA1 and Dm-DRA2) stratify in separate sublayers and exclusively contact polarization-sensitive DRA inputs, while avoiding overlaps with color-sensitive Dm8 cells. Using the activity-dependent GRASP and trans-Tango techniques, we confirm that DRA R7 cells are synaptically connected to Dm-DRA1, whereas DRA R8 form synapses with Dm-DRA2. Finally, using live imaging of ingrowing pupal photoreceptor axons, we show that DRA R7 and R8 termini reach layer M6 sequentially, thus separating the establishment of different synaptic connectivity in time. We propose that a duplication of R7/Dm circuitry in DRA ommatidia serves as an ideal adaptation for detecting linearly polarized skylight using orthogonal e-vector analyzers.