The current study successfully adopted a NM-MRI imaging technique to investigate neuronal loss by directly measuring neuromelanin distribution in both SN and LC in PD patients and matched healthy controls. The study was able to employ an optimised approach to detect small signal intensity changes in the early stage of PD by combining quantitative susceptibility mapping (QSM) analysis used by Takahashi and colleagues[16], as well as applying colour scale contrast to render NM-MRI images recommended by Sasaki et al.[19] and adjust the appropriate window width and window level to achieve optimal results. The study found significant associations between decreasing NM signal intensity in SNpc and LC, and the occurrence and development of PD, which is in line with previous research[12, 14, 18, 19, 20]. This case-control study provides new evidence of the asymmetric distribution of neuromelanin loss in bilateral SN and LC in patients with PD. The study also found that the neuromelanin loss occurred earlier in the LC than that in the SN. This distinct NM distribution pattern may offer an effective differential diagnostic marker for PD and parkinsonian multiple system atrophy. Findings form this study have important implications for future work into early diagnostic markers, predictors for treatment response, and novel intervention targets.
The study examined neuromelanin signal at different sections of the SNpc including the anterior, central and posterior part, and found a decreasing tendency of neuromelanin signal intensity from ventral to dorsal side. At the same time, the distribution of neuromelanin in the SN and the LC is asymmetrical, and the signal on one side is often higher than that on the opposite side(the left side is dominant in this study, perhaps because most of the research population is right-handed), and the LC also maintained synchronization of asymmetries throughout disease loss. Interestingly, this variability was observed with similar results in previous studies[18, 21–24]. Therefore, we assume that this difference may exist objectively, rather than caused by signal inequalities due to high magnetic fields and multi-channel coils. In addition, when it comes to the left and right dominant sides, we quickly thought that when PD motor symptoms involve both limbs, they often develop in an "N" shape, and one side is more severe than the other, which may reveal possible associations among the "dominant side" of neuromelanin distribution with handedness and onset side in PD. However, this asymmetry was not observed in the SNpc of early-stage patients, but was seen in progressive-stage patients, indicating that the speed of neuromelanin loss in the SN nuclei on both sides is inconsistent in the early stage of PD. In other words, the apoptosis rate of dopaminergic neurons in the SN was not uniform on both sides in the early stage of PD. In addition, the results show that the CNR of the SN is not directly related to the UPDRS Part III motor symptom score, and some studies have also reached similar conclusions[25, 26]. This is obviously not consistent with the currently recognized etiological mechanism[2], that is, the motor symptoms of PD are caused by the depletion of the SN-striatum dopaminergic system.
For this contradiction, it is not difficult to understand as during the progression of PD, the brain has a variety of compensatory loop mechanisms for the apoptosis of dopaminergic neurons in the SN compacta to supplement the depletion of presynaptic dopamine to slow down this apoptosis[27], such as in the early stage of PD, as a neurotransmitter, NE can supplement DA to a certain extent[28]. This can also explain that the severity of motor symptoms is not directly determined by the loss of DA in the SN compacta but depends on the balance between the brain compensatory circuit and the loss of DA in the SN. If this compensatory effect could complement for the function of lost dopamine enough in early PD, then patients show mild prodromal symptoms[29]. Only in progressive stage when the loop mechanism is decompensated, patients start to experience severe motor symptoms. Correspondingly, the previous problem is easily solved, that is, in the early stage of the disease, the SN nuclei on the non-dominant side may lose slowly due to the existence of a compensatory loop mechanism, while the dominant side maintains a high loss rate, which in turn leads to the loss of neuromelanin on both left and right side. The apoptosis rate of dopamine neurons in the cytoplasmic nuclei is not uniform, and during the decompensation period, the asymmetry of neuromelanin distribution appear in both nuclei again. This also proves that the loss of neuromelanin in the SNpc and the LC is a continuous process, and other brain compensatory circuit mechanisms only play a supplementary role and affect the speed of loss but cannot reverse the loss process[30]. More research is needed to confirm specific underlying mechanisms.
The study also found negative correlations between CNR of LC and NMSS scores and the synchronous asymmetry loss of neuromelanin in bilateral nuclei, suggesting that the speed of apoptosis of noradrenergic neurons in LC was not affected by other compensatory mechanisms. This indicates that in the early stage of PD, when apoptosis occurs in NE neurons of the LC, patients experience a series of symptoms of norepinephrine system disorders, such as hyposmia, repaid eye movement sleep behaviour disorders (RBD), etc. However, LC degeneration is not seen in all patients with PD, but more common in patients with severe dementia[32], which is shown in this study, that is, the specificity and sensitivity of ROC of the LC in early stage PD patients are 66.46% and 67.12% whereas in progressive stage are 90.11% and 70.84%, indicating that the LC had a higher specificity in evaluating progressive PD[33]. It is known from histopathological evidence that Lewy body in the LC often appear several years earlier than that in the SN in PD[34]. For this reason, Braak et al. proposed the famous "Braak’s stages" of PD according to the chronological order of the occurrence of Lewy body in various parts of the brain[35] despite Burke et al. and Jellinger et al. questioned its authenticity and predictive value[36, 37]. However, it is known that non-motor symptoms (such as sleep disturbance, anxiety and depression, etc.) associated with neuronal loss in the LC tend to affect PD patients more than that dyskinesia does[38]. And almost all PD patients experience mild cognitive impairment, particularly in the early stage[39], which have been demonstrated in the current study. Animal studies have shown that tyrosine hydroxylase immunoreactive terminals and NE levels(rather than DA) in striatum, olfactory bulb and spinal cord of transgenic mice that expressed human α-synuclein A53T mutant were decreasing in an age-dependent manner, indicating that LC was more susceptible to the toxicity of abnormal α-synuclein than the SN[40]. Growing evidence suggests the important role of the LC in the development of non-motor symptoms in early stage of PD and have an advantage over the SN in evaluating the severity of the disease. However, Luppi PH et al. argued that RBD was due to degeneration of glutamatergic neurons in the dorsolateral pontine tegmental nucleus or in the ventral medullary reticular formation, instead of NE neurons in the LC[41]. Therefore, the role of LC-NE pathway in PD can be complex and diverse and how the LC degeneration contributes to different non-motor symptoms require further research in the future. It should be also noted that there are limitations in the current study. Firstly, this is only a cross-sectional study with a relatively small sample size conducted during a restricted COVID period. Longitudinal studies with a larger sample size are needed to confirm these findings. Secondly, the uncontrollable abnormal involuntary movements may cause artifacts and degrade image quality despite an optimised imaging protocol adopted.