Neural Network Dysfunction Of Parkinson's Disease With Depression

Parkinson's disease(PD)is a common neurodegenerative disease that frequently occurs in the middle-aged and elderly group.With the rapid development of the population aging,the number of PD patients in China will account for about 50% of the world's PD population by 2030.Depression is one of the most common neuropsychiatric symptoms in PD patients,with a prevalence of around 35-40%.The incidence of depression in PD increases with the progression of the disease.Converging evidence indicates that depression in PD may be a consequence of the neurodegenerative process of the disease rather than a simply reactive process to the chronic,disabling symptoms.Depression associated with reduced functioning and cognitive impairment is a key determinant of poor health-related quality of life in patients with PD.Understanding depression in PD is,therefore,crucial to achieve the optimal diagnosis and treatment that is needed for this disease.Neuroimaging investigations of depression in PD can advance both the diagnosis biomarkers and treatment evaluation of this debilitating illness.Perspectives on functional segregation and integration are important in providing an interpretive framework to brain function.Functional segregation and integration are two distinct operational principles,but they supplement each other.Most previous neuroimaging studies uncovered the neural bases of depression in PD from a functional segregation perspective,and highlighted that depressed PD(DPD)patients had functional and structural abnormalities in several prefrontal,basal ganglia(BG)and limbic regions.Since the neurodegenerative processes for PD are not diffuse,random,or confluent,but instead target specific large-scale neural networks,investigation of neural mechanisms underlying PD with depression from a functional integration perspective(at the neural-system level)can help us get better understanding of the pathophysiology of this disease.In addition,given that the pathogenesis and clinical characteristics of depression in PD is not exactly the same as depression,a comparative study of neural network dysfunction in these two diseases would contribute to our knowledge about the unique biomarker for DPD patients,and provide more effective and reliable imaging evidence for the diagnosis and treatment of this disease.In the current study,we used resting-state fMRI and diffusion tensor imaging(DTI)technique to construct brain functional and structural network,and explored the neural network bases of PD with depression.Resting-state fMRI studies using seed-based functional connectivity analysis have demonstrated aberrant prefrontal-limbic circuit involved in PD with depression.However,it remains largely unknown whether other specific neural networks implicated in this disease.To address the above issue,Study One used independent component analysis(ICA)method to isolate intrinsic connectivity networks(ICNs)comprising of BG network(BGN),default-mode network(DMN),salience network(SN)and left/right frontoparietal network(LFPN/RFPN),which cover large parts of the prefrontal,BG,and limbic areas relevant to DPD patients.In consideration of ICA model order selection having a significant effect on ICN's characteristics,this study acquired ICNs at 2 decomposition levels.In sub-study 1,we assessed functional connectivity changes within each individual network in DPD patients.In sub-study 2,interactions among networks were measured by Pearson's correlation between ICN's time courses.Relationship between abnormalities of intra-and internetwork connectivity and clinical severity in PD patients was assessed by Spearman correlation analysis.The results showed that DPD patients,compared to non-depressed PD(NDPD)patients and healthy controls,had reduced functional connectivity in the BGN(putamen,caudate and thalamus)and DMN(inferior parietal lobe),increased connectivity in the LFPN(dorsolateral prefrontal cortex)and SN(anterior cingulate cortex),as well as hyperconnectivity between DMN and LFPN.Moreover,connectivity abnormalities in the DMN,LFPN and SN correlated with the depression severity,but not with the motor severity in PD patients,suggesting that dysfunction of the DMN,SN,and LFPN is responsible for the presence of depressive symptoms rather than motor symptoms in PD,and reinforcing the hypothesis of functional disruption of the DMN,LFPN and SN involved in depression in PD.These findings have provided evidence for the neural network dysfunction in PD with depression.Abnormal functional connectivity in the BGN,DMN,FPN and SN may be related to degeneration of the dopaminergic,noradrenergic and serotonergic systems.Study Two aimed to investigate functional integrity of these neurotransmitter systems in DPD patients.We using seed-based functional connectivity analysis mapped the corresponding neurotransmitter pathways,and examined functional connectivity changes in these pathways between three groups.Recent resting-state fMRI studies have confirmed that functional connectivity investigation of midbrain regions,such as substantia nigra(SN),ventral tegmental area(VTA),locus coeruleus(LC)and dorsal raphe nucleus(DRN),can give rise to the mesocorticolimbic and nigrostriatal dopaminergic,noradrenergic and serotonergic pathways in healthy subjects.In sub-study 3,the VTA and SN seed regions were identified from a published probabilistic atlas of the dopaminergic midbrain.We determined whether DPD patients showed connectivity changes in the dopaminergic systems by mapping functional connectivity of the VTA and SN respectively.In sub-study 4 and sub-study 5,we mapped functional connectivity of the LC and DRN,which would give rise to the noradrenergic and serotonergic pathways.Functional connectivity changes within the two neurotransmitter pathways were then examined.The results showed that DPD patients,compared with NDPD patients and normal controls,had altered functional connectivity between VTA and anterior cingulate cortex(ACC),LC and dorsolateral prefrontal cortex(DLPFC),LC and precuneus,as well as DRN and ACC.These findings indicated that degeneration of the mesocorticolimbic dopaminergic,noradrenergic and serotonergic systems may lead to depressive symptoms in PD patients.Moreover,connectivity between VTA and ACC,LC and DLPFC,as well as LC and precuneus was correlated with the depression severity,but not with the motor severity in PD patients,suggesting that the VTA-ACC,LC-DLPFC and LC-precuneus connectivity anomalies may contribute to depressive symptoms rather than motor symptoms in PD patients.Besides,the ACC and DLPFC are important node of the SN and FPN network,and also the core structure of neurotransmitter pathways.They receive dopaminergic,noradrenergic and serotonergic projections from midbrain regions(e.g.VTA,LC and DRN).Our findings confirm that functional anomalies in the prefrontal cortex(e.g.,ACC and DLPFC)in DPD patients may be related to impaired projections from the midbrain nuclei.Study One and Study Two using functional connectivity analysis have explored abnormal network integration in PD with depression,which can make further understanding of the neural mechanisms for this disease.It is well known that gray matter volumes,cortical thickness and white matter microstructure in the prefrontal,BG and limbic regions were also changed in DPD patients,suggesting that functional and structural alterations do not exist independently.Study Three aimed to uncover the neuroanatomical substrates of PD with depression.Specially,we investigated the white matter integrity as well as white matter connectivity alterations in DPD patients.In sub-study 6,we used DTI technique and tract-based spatial statistics(TBSS)analysis to investigate disrupted white matter integrity in DPD patients when compared to NDPD patients and normal controls.Nonetheless,investigation of white matter integrity cannot reveal structural connectivity changes in DPD patients.In sub-study 7,we applied fiber tracking technique to construct white matter connectivity network,and then examined structural connectivity alterations in DPD patients.The prefrontal,BG and limbic regions that were documented to be implicated in DPD patients were selected as regions of interest(ROIs).The results showed that DPD and NDPD patients,compared with healthy controls,had aberrant white matter integrity in the bilateral external capsula and superior corona radiata.Damage to the bilateral external capsula and superior corona radiata may lead to motor and cognitive impairment in PD patients.Although there was no significant difference in white matter integrity between DPD and NDPD patients,DPD patients relative to NDPD patients had increased structural connectivity between insula and putamen.These findings indicated that dysfunctional striatum-limbic circuit may be important for the occurrence of depression in PD.Uncovering neuroanatomical substrates of depression in PD is helpful for understanding the relationship between functional and structural alterations,and provides more useful information for diagnosis and treatment with this disease.The existing neuroimaging literature suggests that PD with depression and depression may share common neural circuits.However,recent studies indicated that DPD patients and depressive patients showed significant differences in brain activity.For example,significant cerebral blood flow differences were observed in the right prefrontal cortex,left frontoparietal region and parietal-occipital region between DPD and depressive patients.Furthermore,DPD and depressive patients showed significant difference in directional functional connectivity between emotional network and motor network.The above findings imply that the neurobiological bases may be different between depression and PD with depression.Study Four(sub-study 8)used large-scale brain network analysis to construct functional connectivity network of DPD,depressive patients,NDPD and normal controls,and then investigated whether DPD and depressive patients shared common neural network bases.The results showed that(1)DPD and NDPD patients,compared with normal controls,had reduced functional connectivity between bilateral postcentral gyrus and bilateral middle occipital cortex;(2)DPD and NDPD patients exhibted no significant difference;(3)depressive patients relative to normal controls had reduced functional connectivity between bilateral postcentral gyrus and bilateral middle occipital cortex,as well as medial prefrontal cortex and precuneus;(4)DPD patients compared with depressive patients had increased functional connectivity between medial prefrontal cortex and precuneus.Abnormal functional connectivity between sensorimotor and visual cortex in DPD and NDPD patients may be associated with motor and visual dysfunction in PD patients.Funtional connectivity changes in the visual cortex in depressive patients may lead to abnormal integration between visual information and other cognitive processes(e.g.,self-directed thinking).In addition,functional connectivity alterations in the sensorimotor cortex may be associated with psychomotor disturbance in depressive patients.Finally,DPD patients and depressive patients showed significant functional connectivity difference between medial prefrontal cortex and precuneus,indicating that the midline structure of DMN can be used to differentiate depression and PD with depression.Combined with the result of Study One,we infer that PD with depression and depression may be linked with different default mode sub-networks.In summary,this study using resting-state fMRI and DTI technique combined with brain connecitivity annlysis have uncovered the neural network bases of depression in PD.Integration of functional and structural imaging will provide more accurate neural model for this disease.The current findings have demonstrated that(1)depression in PD is associated with disrupted BGN,DMN,FPN and SN;(2)dysfunction of the midbrain-cortical and striatum-limbic circuits are crucial for the occurrence of depression in PD;(3)functional abnormalities in the prefrontal cortex(e.g.,ACC and DLPFC)may be related to impaired dopaminergic,noradrenergic and serotonergic projections from the midbrain nuclei;(4)the midline structure of DMN can be used to differentiate PD with depression and depression;(5)PD with depression and depression may be associated with different default mode sub-networks.The pathogenesis of depression in PD is not very clear yet.Development of potential diagnostic value of magnetic resonance imaging technique can enhance our understanding of the neural echanisms undelying depression in PD.