Labelling putative presynaptic structures in individual olfactory bulb dopaminergic neurons.

(A) Log2-normalised expression of Th and Syp mRNA in DA neurons from single cell RNA sequencing data (Brann et al., 2020). (B) Example snapshot of a 3D confocal stack of endogenous immunostaining for synaptophysin (green) and TH (magenta) in the glomerular layer of the OB. Yellow arrowheads point to small clusters where TH and synaptophysin co-localise. Scalebars: 5 μm (main) and 0.5 μm (inset). (C) Strategy to label putative presynaptic release sites in individual DA neurons. (D) Example snapshot of a 3D confocal stack of a successfully labelled DA cell. Inset 1 reveals the TH+ dopaminergic identity of the neuron (cyan), and inset 2 highlights the Syn-mRuby puncta (magenta, black). Scalebars, 5 μm. (E) Example snapshot of a 3D confocal stack of a GFP+ (green), Syn-mRuby+ (magenta, black) neuronal process co-stained with vGAT (orange, black). Yellow arrowheads indicate examples where Syn-mRuby and vGAT puncta co-localise. The last panel shows the orthogonal views of the bottom puncta. Scalebars, 1 μm.

Anaxonic DA neurons have dendritic neurotransmitter release sites.

(A) Example snapshot of a 3D confocal stack of a TRIM46-anaxonic DA neuron. Blue arrowheads point to examples of other TRIM46+ AISs (orange, black) in the same region which do not co-localise with this neuron’s GFP signal. Scalebars, 5 μm. (B) Snapshot of the same neuron in (A) showing Synaptophysin-mRuby Puncta (magenta, black) on the dendrites. Yellow inset highlights a region of the neuron with multiple mRuby+ puncta within the GFP+ (green) processes (yellow arrows). Scalebars, 5 μm and 0.5 μm. (C) Example snapshots from 3D dendritic reconstructions (green, GFP) and presynaptic puncta detection (magenta, Syn-mRuby) of anaxonic DA neurons. Note: these are not full dendritic reconstructions, but example dendrites. Dotted white circle represents the soma location. Scalebars, 5 μm.

Axon-bearing DA neurons have release sites on their intermittently myelinated axons, but not on their dendrites.

(A) Stitching of individual confocal stacks processed for maximum intensity projections of OB DA neurons co-stained with TH (green) and myelin basic protein (MBP, magenta). Yellow arrowheads point to myelinated parts of the axon, blue arrowheads show unmyelinated areas. Scalebar, 10 μm. (B) Snapshot of a 3D confocal stack of a distal DA axon stained with GFP (green), TH (cyan), MBP (orange) and synaptophysin-mRuby (magenta, black). Yellow inset highlights the location of the presynaptic bouton. Yellow arrowheads point to co-localised regions, blue arrowheads show non-co-localisation. Scalebars, 2 μm and 1 μm. (C) Snapshot of a 3D confocal stack of an axon-bearing TRIM46+ DA neuron. Yellow arrowheads show co-localised staining for GFP (green) and TRIM46 (orange, black). Scalebars, 2 μm. (D) Soma area of axon-bearing and anaxonic DA neurons. Each dot shows one cell; lines show mean ± SEM; n = 11 axon-bearing cells and n = 9 anaxonic neurons from N = 4 mice; unpaired t-test with Welch’s correction; ****, p <0.0001. (E) Snapshot of a 3D confocal stack of the same axon-bearing DA neuron showed in (C), co-stained with GFP (green) and synaptophysin-mRuby (magenta, black). Blue arrows show dendritic segments lacking mRuby label, despite the presence of clear mRuby+ puncta in neighbouring non-GFP+ processes. Scalebars, 2 μm. (F) Dendritic puncta density in axon-bearing and anaxonic DA neurons. All conventions as in D; n = 11 axon-bearing cells and n = 9 anaxonic neurons from N = 4 mice; Mann-Whitney’s t-test; **, p = 0.0001.

Anaxonic DA neurons are capable of self-inhibition.

(A) Log2-normalised expression (of Th, Drd1, Drd2, Drd3, Drd4 and Drd5 mRNA in OB DA cells. (B) Snapshot of a 3D confocal stack of a virally labelled (GFP+, green) anaxonic DA (TH+, cyan) neuron stained against gephyrin (orange) and synaptophysin-mRuby (magenta). i) and ii) magnification of the boxed area; green shows reconstructed dendrite; blue arrowheads point to two neighbouring puncta, one mRuby+ and one gephyrin+. Scalebars, 3 μm and 0.5 μm. (C) Example trace of an action potential fired by a putative anaxonic DA neuron (left) and its monophasic phase-plane plot profile (right). Note the prolonged repolarisation due to Cs-based internal solution. (D) Example traces of an AEI response recorded before (magenta) and after (grey) the application of gabazine. The subtraction is shown in the orange inset trace. (E) Schematic showing the potential involvement in the AEI response of neighbouring GABAergic neurons activated by dopamine released from the patched DA cell. (F) Example traces of an AEI response before (purple) and after (green) applying D1-like and D2-like receptor blockers (hydrobromide and sulpiride, each at 10 μM). (G) AEI charge before (purple) and after (green) applying dopamine receptor antagonists; n = 3 cells from N = 3 mice; unpaired t-test with Welch’s correction: n.s., non-significant. (H) Schematic showing the potential involvement in the AEI response of neighbouring GABAergic neurons activated via gap junctions. (I) Example trace of an AEI response in the presence of the gap junction blocker carbenoxolone at 100 μM. (J) AEI charge in the presence of carbenoxolone at 100 μM. Each dot shows one cell; lines show mean ± SEM; n = 9 cells from N = 4 mice. (K) Schematic showing the potential involvement in the AEI response of neighbouring GABAergic neurons activated by depolarising GABA released from the patched cell. (L) Example trace of an AEI response in the presence of the NKCC1 channel blocker bumetanide at 20 μM. (M) AEI charge in the presence of bumetanide at 20 μM. All conventions as in F; n = 5 cells from N = 3 mice.

Axon-bearing DA neurons do not self-inhibit.

(A) Example trace of an action potential fired by a putative axon-bearing DA neuron (left) and its biphasic phase-plane plot profile (right). Arrows point to spike onset kink (left) and initial segment bump (right). Note the prolonged repolarisation due to Cs-based internal solution. (B) Example traces showing tail current responses in an axon-bearing DA neuron before (blue) and after (grey) the application of gabazine. The subtraction is shown in the orange inset trace. (C) AEI charge in axon-bearing and anaxonic DA neurons. Each dot shows one cell; lines show mean ± SEM; n = 9 axon-bearing cells, n = 31 anaxonic cells from N = 18 mice; Mann-Whitney test; ****, p < 0.0001. (D) Example traces showing spontaneous IPSCs in an axon-bearing DA neuron before (blue) and after (grey) the application of gabazine. (E), (F) Example traces showing tail current responses in axon-bearing DA neurons before (blue) and after (grey) the application of gabazine using a paired-pulse protocol (E) and prolonged depolarisation (F). Subtractions are shown in the orange inset traces. (G) Example traces showing that the tail current recorded in axon-bearing DA neurons (blue) is blocked by cadmium (ocre). The subtraction is shown in the orange inset trace.

Primary antibodies.

A single example of an axon-bearing DA neuron with potential dendritic release sites.

(A) Snapshot of a 3D confocal stack of a DA neuron with GFP (green) and TRIM46 (orange) labelling, revealing the axon-bearing identity of the neuron. Yellow arrowheads point to the GFP+/TRIM46+ co-localised zone. Scalebars, 2 μm. (B) Snapshot of a 3D confocal stack showing synaptophysin-mRuby label (magenta, black) in the dendrites of the axon-bearing neuron from (A) (green). The levels of synaptophysin-mRuby look dramatically higher and with a less defined puncta profile than in all anaxonic DA cells – note levels of somatic expression compared to Figure 2B. Yellow arrowheads point to examples of detected puncta. Scalebars, 5 μm.