DLB patients are at high risk of developing major depression, with depression having a significant impact on quality of life 3. Depression in DLB correlates with α-synuclein pathology in late onset MDD, indicating that α-synuclein pathology may be a factor in MDD development in older individuals 35. The sgACC is an area intimately involved in the aetiology of depression, with structural and functional changes in early onset MDD 10, 26 and a high α-synuclein pathological burden in DLB 27. Despite high levels of α-synuclein in the sgACC, we found no significant difference between depressed and non-depressed DLB donors for α-synuclein pathology burden suggesting gross pathology within the sgACC is not a significant driver of depressive symptoms. Significant neurone loss within the sgACC in DLB is not apparent, and there is no association with depression. Although the basis for depression in DLB is complex, our results indicate reduced dopaminergic innervation in sgACC with changes in dopaminergic synaptic function and decreased levels of dopaminergic markers contributing. This may warrant clinical trials of D2 dopamine receptor agonists since pramipexole and pergolide show efficacy in reducing depression in patients with PD 36, 37, as well as symptoms of major depression in patients without PD 38, 39.
The mesolimbic dopamine pathway from the VTA to cortical and subcortical regions includes the cingulate with additional projections from the dorsal tier of the SN 33. VTA dopaminergic neurones play a role in reward and stress 20 with dopaminergic dysfunction in the mesolimbic system contributing to anhedonia and loss of motivation in MDD 21. We identified a decrease in VTA cell number in DLB cases overall, but no significant loss in relation to depression in DLB. Compared to the major loss of neurones in the SN 27, the VTA shows relative preservation of neurones in PD 27, 40. Similar to the current findings, severe cell loss and gliosis is associated with the presence of depression in PD 41, with VTA neuronal density having a significant effect 42. In PD, PDD, and DLB, higher midbrain α-synuclein burden associates with depressive symptoms 27 suggesting that mesolimbic and mesocortical projections rather than nigrostriatal changes contribute to depression in DLB, similar to PD 43.
In this study, we observed reduced sgACC dopaminergic fibre density in DLB cases with depression as a potential indicator of the reduced VTA innervation. In PD, motor symptoms appear when around 50% of SN dopaminergic neurones are lost 44, although striatal innervation at onset of motor symptoms is more severe, showing around 80% terminal loss 45. Dopaminergic axons and terminals are the main site of pathology, with the neurodegenerative process in PD thought to follow a retrograde pathway originating in the striatal terminals 46. The enhanced dopaminergic fibre loss seen in this study in DLB cases with depression in the absence of severe (~ 20%) VTA neuronal loss may represent a similar effect with axon and terminal loss prior to major cell death.
Our STED analysis showed phosphorylated α-synuclein within sgACC dopaminergic synapses and increased phosphorylated α-synuclein in sgACC tissue homogenates. In depressed DLB cases, elevated synaptic α-synuclein corresponded with elevated SNAP25, a finding also seen with serotonergic terminals. One possibility is that pathological α-synuclein sequesters SNAP25 within synapses leading to synaptic dysfunction. Alpha-synuclein modulates synaptic activity through effects on vesicle recycling and release by assisting with SNARE complex formation 34, 47. The positive correlation identified in the current study may represent a synaptic response to increase activity and maintain normal neurotransmission following reduction in cell numbers 48. Pathological fibrillar α-synuclein can however, rapidly promote aberrant synaptic activity following neuronal application 49. In forming aggregates within synapses, fibrillar α-synuclein may cause an effective depletion of functional SNARE proteins including SNAP25, reducing the effective synaptic vesicle pool. Fibrillar and oligomeric α-synuclein also has direct effects on synaptic machinery by depleting SNARE complexes 50. This may underscore the increase in SNAP25 within α-synuclein containing dopaminergic synapses, with SNAP25 increased, but non-functional due to sequestration in α-synuclein aggregates 51. The increased volumes of α-synuclein negative synapses in depressed DLB donors may be a compensatory response to reduced function of α-synuclein containing synapses in depressed cases. The combined effects of dopaminergic synapse reduction due to cell loss and reduced synaptic efficiency in remaining synapses may cause an effective depletion of dopamine to the sgACC cortex and contribute to depressive symptoms.
In this study, we found a loss of dopaminergic proteins in DLB cases with depression compared to controls, including DAT, TH and DDC. Using hierarchical clustering, DAT and DRD4 showed significant associations with depression in DLB. These reductions align with DAT fibre loss in depressed DLB cases and reinforce the general loss of dopaminergic innervation in DLB, but particularly in DLB experiencing depression 25, 27. Dopaminergic deficits including reduced striatal DAT binding has been observed in depressed patients with anhedonia 22 and additionally in PD and DLB, with DLB cases showing greater DAT loss in caudate compared to PD 52, with reduced ACC DAT binding in DLB 53. The use of dopamine agonists in MDD can have significant benefits, and may be beneficial in DLB if suitable treatment regimens are identified 54, 55.
Our results indicate minimal changes in post-synaptic D2DR in DLB cases. The highest levels of D2DR are found in the striatum, nucleus accumbens and olfactory tubercle, with significant levels in the SN, VTA, amygdala, hippocampus and cortex 56. In PD, SN neurone loss results in reduced striatal dopamine with increased striatal D2DR as a compensatory mechanism in response to low dopamine 57. Increased dopamine D2/D3 receptor binding and lower DAT activity has also been shown in MDD, potentially reflecting compensatory changes 58, 59. This does not appear to occur in DLB however, where reduced striatal and cortical D2DR occurs 60, 61.
D3DR are expressed pre- and postsynaptically and have the highest affinity for dopamine 62. DRD3 is a clinically relevant treatment target following identification of DRD3 downregulation in MDD 63. A reduction in DRD3 in the ventral striatum has been observed in PD cases 64, but with either downregulation 65 or no change 60 in DLB. Our results indicate that reduced DRD3 levels may play a role in depression in DLB, and DRD3 specific dopaminergic agents may be beneficial. Pramipexole and ropinirole improve depressive symptoms in PD and may provide benefit for depression in DLB, although careful monitoring may be needed to prevent unwanted side effects including impulse control disorders and hallucinations 54, 66, 67. In PD, improved working and episodic memory occurs with dopamine agonists, indicating that dopaminergic agonists may additionally improve cognition 68, 69.
Our findings show no significant changes in sgACC DRD4 levels in DLB despite an association based on hierarchical clustering. DRD4 shows the lowest levels of expression of dopamine receptors in the brain, with highest densities in anterior limbic and cortical forebrain suggesting cognitive, executive, and motivational regulation 70. Similar to previous studies indicating that DRD4 are not implicated in the modulation of depressive-like behaviours 71, our findings suggest that DRD4 may not be a treatment target.
In contrast to dopaminergic changes, serotonergic involvement in depression in DLB was limited, comparable with similar studies 42, 72. Whilst the dorsal raphe nucleus (DRN) shows LB pathology in DLB and PD, studies of neurone loss are variable with reduction 73 or no change in DRN neurones in PD or PDD 72, 74. In one report in DLB, neuronal loss is present within both the DRN and median raphe 75. In the current study, due to the tissue sampling strategy and differences in DRN projection sites (Ren et al., 2018), DRN neurone numbers were not assessed. We however found no change in sgACC serotonergic fibre density in DLB similar to findings in the amygdala and dorsal prefrontal cortex in LBD donors with and without depression 72. We did however find α-synuclein associated with serotonergic synapses and a positive correlation between SNAP25 volume and α-synuclein volume. This finding in 5HT synapses may indicate 5HT terminal dysfunction without a change in innervation in the sgACC since we saw no change in fibre density or 5HTT protein levels. Other studies have also showed no association with 5HTT levels with depression or anxiety in PD 76. Future studies should however determine the effects of α-synuclein pathology in DLB in the brainstem and midbrain serotonergic system using appropriate stereological approaches.
There has been considerable in vivo use of 5HTT ligands in PD and to a lesser extent DLB. In drug naïve PD, there were no changes in non-apathetic PD but reductions in [(11)C-N,N-dimethyl-2-(-2-amino-4-cyanophenylthio)-benzylamine (DASB)] retention in apathetic patients in the sgACC when compared to non-apathetic PD suggesting reduced serotonergic innervation 77. Similarly, reduced SERT assessed with DASB is seen in caudate and ACC in early stage PD and throughout the disease course 78–80 and at post mortem 81. In depressed PD patients however, increased DASB retention was observed in several cortical areas although more generally there was no change in DASB retention in ACC in MDD 82, 83. In vivo studies in DLB have used the combined DAT and SERT ligand 123I-N-ω-fluoropropyl-2β-carbomethoxy-3β-(4-iodophenyl) nortropane (123I-FP-CIT) where the assumption is that cortical retention only represents 5HT innervation despite DA terminals being present. Studies with 123I-FP-CIT have shown either reduced cortical SERT 84–86, or no cortical changes 87. No reduction in SERT was seen in the ACC at post mortem in PDD, although brain regions such as caudate showed reductions 88. Similarly, in DLB compared to control or non-depressed patients, SERT assessed with cyanoimipramine showed preserved cortical binding in depressed DLB donors 89, 90. Whilst these prior studies have shown variable outcomes, our findings indicate SERT reductions in DLB are mild or absent and contributions to development of depression are limited.
Decreased serotonin 5HT1A binding is inidcated in the ACC in MDD patients using PET 14 however, we observed no alteration of 5HT1A protein levels in DLB. This contrasts with findings using the 5HT1a ligand 8-Hydroxy-2-Dipropylaminotetralin showing increased binding but altered KD in DLB and PD 91. This may suggest a change in receptor occupancy due to changes in protein conformation but not protein levels. However, our finding of decreased 5HT2A protein in sgACC in DLB donors with depression corresponds with reduced 5HT2A binding in ACC and DLPFC in MDD 15 and reduced [3H]-ketanserin binding in DLB and PDD 92, 93. These changes in 5HT2A but not 5HT1A suggests region specific serotonergic receptor changes that may benefit from selective targeting. Pimavanserin, a 5HT2a selective inverse agonist and antagonist, has been used in psychosis treatment in PD 94, 95. Despite mixed benefits in MDD 96, 97, an open label study of Pimavanserin for depression in PD in the absence of psychosis showed delayed benefit in an 8-week trial with 60% of patients showing improved mood 98 although longer-term use of Pimavanserin may show reduced efficacy 99. The possibility of using Pimavanserin in the treatment of depression in may be warranted in certain DLB patients 100.
In summary, these results suggest that there is primarily a reduced dopaminergic innervation in the sgACC in DLB, along with reduced levels of dopaminergic markers and receptors, and changes in dopaminergic synaptic function. Careful treatment with selective dopaminergic agonists or positive allosteric modulators may be beneficial in alleviating depressive symptoms in DLB.