The dopaminergic system plays a crucial role in a broad spectrum of neuropsychiatric disorders (e.g., Parkinson’s disease (PD), schizophrenia and addiction) (Poulin et al., 2020), and dopaminergic dysfunction has been hypothesized to underpin drug addiction-related behavior (Lin et al., 2016). Drug addiction is a non-degenerative disease, which results in elevated levels of extracellular monoamine neurotransmitters in all brain regions with a magnitude of increase that varies between brain regions (Zhang et al., 2001; Bubenikova-Valesova et al., 2009). Methamphetamine (MA) is one of the most widely abused psychostimulants that may cause addiction (Shi et al., 2000; Saha et al., 2014). Brain dopaminergic pathways are activated by MA exposure, which promotes dopamine (DA) release from presynaptic nerve endings, inhibits DA transporter (DAT) reuptake, results in increased extracellular DA levels (Cruickshank and Dyer, 2009; Kim et al., 2020), and damages the dopaminergic system in all animal species including humans (Moratalla et al., 2017; Granado et al., 2018). Compulsive drug-seeking and drug-taking are important drug abuse-related behaviors that have been linked to alterations in dopaminergic neurotransmission and to impaired inhibitory control (Groman et al., 2012). Regarding DA, MA activates the mesolimbic and nigrostriatal dopaminergic pathways, and the clinical response to the drug in humans includes euphoria, arousal, reduced fatigue, positive mood, and anxiety. MA administration in rats induces behavioral sensitization which is characterized by augmentation of locomotor activity (Ares-Santos et al., 2014). Behavioral sensitization has been postulated to result from enhanced DA release in the mesolimbic and nigrostriatal dopaminergic terminals. DA cell bodies in the midbrain are the primary source of dopaminergic input to the caudate putamen (CP) and nucleus accumbens (NAc), and DA neurons in the ventral tegmental area (VTA) and the substantia nigra (SN) send their axons to the NAc and CP, respectively (Lammel et al., 2014; Thanos et al., 2016). DA is the major neurotransmitter in the neostriatum (Björklund and Dunnett, 2007), and neural activity of DA neurons in the somatodendritic field may influence the synthesis and release of DA in the terminal regions (Poulin et al., 2020; Lu et al., 2021). In vitro studies have shown enhanced DA release from NAc and striatum tissue slices from rats sensitized to psychostimulants such as MA (Desai et al., 2010; Granado et al., 2010). In vivo measurement of DA function is critical for understanding how this key neuromodulator contributes to addiction and may provide objective markers of DA function in MA addiction. Currently, positron emission tomography (PET) is the main imaging method to measure DA function (Huang et al., 2019). However, PET is limited by the risk of radioactivity exposure and costly specialized infrastructure. The non-invasive measurement of DA function abnormalities by optimal imaging method remains challenging.

Dopamine neurons contain a melanin pigment known as neuromelanin (NM) that is mainly distributed within the SN (Zucca et al., 2014). NM normally accumulates with age in human SN neurons (Zecca et al., 2002). A neuronal pigment has also been observed in the SN of adult rats (Rabey and Hefti, 1990), and its concentration seems to depend upon age (DeMattei et al., 1986). Indeed, DA is one of the precursors in NM synthesis. NM is synthesized via iron-dependent oxidation of cytosolic catecholamines and subsequent reaction with proteins and lipids in midbrain DA neurons (Karlsson and Lindquist, 2013). NM biosynthesis is driven by an excess of cytosolic DA that not collected in synaptic vesicles (Sulzer et al., 2000). Excess cytosolic DA can be removed by converting it into a stable compound such as NM. The main iron compound in DA and norepinephrine neurons is the NM-iron complex, since NM is an effective metal chelator. NM serves to trap iron and provides neuronal protection from oxidative stress (Zucca et al., 2017). The biological function of NM is controversial. On one hand, NM plays a neuroprotective role in physiological states by combining with metal ions or drugs (Zecca et al., 2004). On the other hand, NM interacts with excessive Fe3 + and α-synuclein in the cytoplasm, damages Ca2 + homeostasis, and overactivates microglia, resulting in toxic effects and a pathological state (Zecca et al., 2008; Kim et al., 2020). When NM is combined with iron or other metals, other paramagnetic complexes result in T1-shortening effects in neuromelanin-sensitive magnetic resonance imaging (NM-MRI) (Trujillo et al., 2017; Salzman et al., 2021). Brain tissue signal intensity is reduced, and NM-containing nuclei show high signal intensity due to shortened T1 relaxation times (van der Pluijm et al., 2021). Thus, NM-MRI can enhance the T1-related contrast between NM and the brain tissue (Wengler et al., 2020), which allows NM-MRI to be used to detect NM accumulation. As NM is located in the cell bodies of dopaminergic neurons of the nigrostriatal pathway (Reneman et al., 2021), NM-MRI can also be used as a proxy measure for DA function. It provides the possibility for indirect measurement of DA function in non-neurodegenerative conditions.

Recent studies have suggested that NM-MRI may provide a complementary non-invasive proxy measure of DA function (Sulzer et al., 2018; Cassidy et al., 2019; Martín-Bastida et al., 2019). Interindividual variability in DA function may result in varying levels of NM accumulation in the SN, which can be detected by NM-MRI (Cassidy et al., 2019). Although the potential of NM-MRI as a candidate biomarker for dopaminergic pathology in patients with neurodegenerative disorders like PD has been demonstrated, studies in patients with non-neurodegenerative diseases are sparse. Critically, NM-MRI can reliably capture not only NM depletion in the SN in those with neurodegeneration disorders (Reimão et al., 2015; Fabbri et al., 2017) but also alterations in DA function in those without neurodegeneration (Shibata et al., 2008; Jauhar et al., 2017; Cassidy et al., 2019), based on in vitro evidence that stimulating DA synthesis boosts NM synthesis (Rice et al., 2016; Ashok et al., 2017). In addition, research on addiction and psychosis-related disease has shown that NM-MRI can be used to capture increases in NM accumulation in the SN (Weinstein et al., 2017; Cassidy et al., 2020).

In this study, we aimed to investigate the feasibility of using NM-MRI to measure NM concentrations in rats following acute MA administration and to explore whether NM-MRI signal intensity can be used as potential biomarker of DA function in non-neurodegenerative diseases.

The present study assessed the effect of acute MA exposure on functional changes in dopaminergic pathways by NM-MRI. The results demonstrated that the NM-MRI signal intensity in the SN was significantly higher than that in other brain regions at all the time points. This is consistent with the fact that NM is mainly deposited in the SN as presented in a previous study (Zecca et al., 2001). This study further demonstrated that NM-MRI is sensitive enough to detect regional variability in concentrations of NM in brain tissue, which presumably depends on interindividual and interregional differences in DA function (Poulin et al., 2020). NM-MRI has the ability to directly demonstrate different signal intensities based on the NM content of their constituent neurons. This is a prerequisite for its use as a biomarker of interindividual variability in DA function following MA exposure. Since the results indicated that NM-MRI signal intensity corresponded to regional tissue concentration of NM, particularly in the midbrain region of the SN, NM accumulation probably correlated with higher DA release from nigrostriatal SN neurons. Together, these findings open the possibility that NM-MRI could be used as a proxy measure assessing DA function.

Interestingly, this study found that the lowest to highest NM-MRI signal intensity across brain regions were observed in the CC, CP, NAc, HIP and SN before and after MA administration. The SN had the highest signal intensity compared to other brain regions at each time point. NM has been shown to be richest in the SN but it is practically ubiquitous throughout the brain as it is diffusely distributed in smaller amounts in neurons of other brain areas (Zecca et al., 2008; Ward et al., 2009). Our results suggested that the deposition of NM in the CC, CP, NAc, HIP and SN, which showed a gradual increase from low to high, with highest levels in the SN and lowest levels in the CC. These results probably related to the cellular heterogeneity of midbrain dopaminergic neurons (La Manno et al., 2016), and NM-MRI has the ability to detect regional heterogeneity based on NM content. Interindividual variability in DA function may result in varying levels of NM accumulation in the SN (Cassidy et al., 2019). Therefore, the differences in NM content across brain regions may be related to the diverse array of DA functions. The heterogeneity in the cell populations in the midbrain suggests that DA function may differ substantially between neuronal tiers projecting to other regions (Joel and Weiner, 2000; Rice et al., 2016; Weinstein et al., 2017). Moreover, according to prior research, the SN contains the vast majority of DA neurons among the traditionally defined midbrain DA groups (Anderegg et al., 2015),which is consistent with higher signal intensity in the SN in our study.

Having shown that NM-MRI measures the disparity in regional concentrations of NM in the SN and other brain regions, we then focused on evaluating the signal intensity changes in different brain regions over time after acute MA administration, to test whether increased NM-MRI signal intensities correlated with DA function in vivo. By analyzing the signal intensity changes at different time points after acute MA administration, the results showed that acute MA administration contributed to a continuous upward trend in NM-MRI signal intensities in four brain regions, all regions except for the CC, from baseline to 90 min, although the magnitude of increase varied across brain regions. However, this upward trend in signal intensity was not seen in the control group following the normal saline injection. These results suggested that MA administration activated the mesolimbic (NAc and HIP) and nigrostriatal (SN and CP) pathways, and enhanced DA release at the dopaminergic terminals, leading to an increase in NM-MRI signal intensity in addiction-related brain regions. One interpretation is that MA exposure is associated with a redistribution of DA between cytosolic and vesicular pools (Cassidy et al., 2020). MA inhibits vesicular monoamine transporter 2 (VMAT2) and promotes DA release from synaptic vesicles into the cytoplasm. Subsequently, available cytosolic DA is reverse transported by DATs into the extracellular space, resulting ultimately in increased extracellular DA (Fleckenstein et al., 2007; Skinbjerg et al., 2010; Nickell et al., 2014). A decrease in VMAT2 could account for the increased NM-MRI signal intensity; that is, decreased VMAT2 expression would decrease vesicular DA and increase the cytosolic DA pool from which NM is synthesized (Sulzer et al., 2000; Markov et al., 2008; Zucca et al., 2014). Excess cytosolic DA is thought to copolymerize with cysteine into a polymer of ill-defined composition which is eventually packaged into larger double-membrane bound vesicles (NM granules), leading to increased accumulation of NM (Haining and Achat-Mendes, 2017). Sulzer et al. (2000) showed that NM forms in cultured dopaminergic neurons when cytoplasmic concentrations of DA are artificially increased. Another explanation is that NM synthesis itself acts as a protective mechanism, clearing out excess cytoplasmic catecholamines and increasing NM synthesis (Zecca et al., 2002). Oxidative stress is necessitated for the handling of high concentrations of catecholamines, while the protective role of NM in dopaminergic neurons lies in its prevention of neurotoxicity from quinones that are formed during DA oxidation. When DA is oxidized to dopamine o-quinone, aminochrome and 5,6-indolequinone are formed and typically undergo polymerization to form the dark pigment NM (Muñoz et al., 2012b). The product of oxidative stress (i.e., NM) could then function to alleviate the very thing that created itself (i.e., excess DA). As has been pointed out, sequestration into vesicles apparently alleviates this toxicity and stress on the cell. This process suggests that the accumulation of NM itself could serve as a storage, protection, and rerelease mechanism for DA, possibly acting as an actual molecular memory loop. This could partly explain the reinforcement of addictive memories and behavior that are associated with DA-releasing drugs. Previous studies have shown that PD is characterized by the loss of DA neurons, resulting in decreased NM concentration and NM-MRI signal intensity (Sasaki et al., 2006; Sulzer et al., 2018; Martín-Bastida et al., 2019). However, MA exposure produces the opposite effects. Acute MA exposure is characterized by increased DA concentrations in most brain regions (Czoty et al., 2004; Desai et al., 2010), and our results indicated that the increased NM-MRI signal intensity may have been due to increased NM concentrations, which may indirectly reflect the elevated DA availability in DA neurons and is consistent with the idea that NM accumulation depends on DA function in the soma (Sulzer et al., 2000; Cebrián et al., 2014; Cassidy et al., 2019). Thus, our results supported the idea that NM-MRI captures an addiction-related aspect of DA dysfunction in the nigrostriatal pathway.

In this study, across each of the brain regions, the signal intensity growth rate within 5 min of acute MA administration was the highest compared with other time periods, which indicated that DA release in the four brain regions occurred most rapidly within 5 min after acute MA administration compared with other time periods. A previous study showed that extracellular DA concentrations in the SN, CP and NAc increased with a peak response at 30 min after acute MA injections (Zhang et al., 2001). One possible explanation for the current results was the delay in scanning. In particular, it took 20 min to complete the NM-MRI scan after the MA injection. Another explanation is that MA was administered differently in this study. Intraperitoneal injections were used in a previous study (Zhang et al., 2001), while jugular vein injections were used in this study, a route by which the drug has a faster effect.

Having validated that the signal intensity increased in most brain regions over time, we then proceeded to analyze the differences in increased signal intensities in different brain regions at different time points. In this study, the NM-MRI signal intensity of SN was significantly higher at the 60 and 90 min both in within-groups and inter-group comparison. There was no significant result in other brain regions, which suggested that a significant effect from MA administration in SN. The rate of increase and amplitude of NM-MRI signal intensities in the SN were higher than those in other brain regions at each time point after acute MA administration. These results suggested that the conversion of DA to NM was most significant in the SN after acute MA administration, which was consistent with the SN being where most NM accumulates. Given that a previous imaging study showed that cocaine users had higher locus coeruleus (LC) NM signal intensity but the VTA/SN did not show significant group differences (Wang et al., 2021), the finding of NM-MRI signal intensity increases being most significant in the SN provides additional insight into the pathophysiology of MA exposure. The combination of increased NM-MRI signal intensity and elevated NM in the SN suggests that DA neurons are differently distributed after MA exposure. The SN is DA-rich and contains both redox-available NM and a high iron content (Zucca et al., 2014). There are several interrelated mechanisms involving iron, DA, and NM in SN neurons (Zucca et al., 2017). NM accumulates during aging and is the catecholamine-derived pigment in the dopaminergic neurons in the SN, but it is relatively depleted in the SN of patients with PD. On one hand, NM is an effective metal chelator and serves to trap iron and provides neuronal protection from oxidative stress. The high concentration of NM in SN neurons seems to be linked to the presence of a considerable amount of cytosolic DA that has not been sequestered into synaptic vesicles. Excess cytosolic DA can be removed by converting it into a stable compound like NM, and this process rescues the cell, that is, NM serves a protective role (Muñoz et al., 2012a). Our results could be explained by this neuroprotective effect in that NM synthesis was most significant in the SN. On the other hand, many cellular redox reactions rely on iron, however, an altered distribution of reactive iron is cytotoxic (Wang et al., 2021).

In this study, we first demonstrated that NM-MRI is sensitive enough to detect regional heterogeneity in NM content and found that the SN itself accumulates the most NM. Secondly, the NM-MRI signal intensity in each brain region showed an upward trend after MA administration. consistent with the evidence that DA dysfunction is closely related to drug addiction, and thus, our data support the notion that the increase in DA concentration would lead to increases in NM deposition over time, which can be captured by NM-MRI. Finally, the conversion of DA to NM was most significant in the SN, which also confirmed that NM is mainly deposited in the SN. Our work suggested that NM-MRI has the ability to capture this addiction-related phenotype consisting of nigrostriatal DA excess by demonstrating an increase in NM accumulation in the SN. These findings promote the development of this candidate biomarker for addiction-related psychiatric disorders and treatment selection.

There are several limitations in this study. First, this study lacked an absolute gold standard. It is critical that future studies should provide direct evidence for a relationship between NM-MRI signal intensity and concentration of DA and NM measured in ex vivo rat brain tissue using biological methods. Second, similar to other neuropsychiatric diseases, our results showed that the NM-MRI signal intensity is sensitive but not necessarily diagnosis-specific for MA addiction; applications in other non-degenerative diseases remain to be studied. Third, there was no monitoring of physiological parameters during the experiment and that a potential bias due to a change in the animal’s body temperature cannot be excluded. Finally, our study was limited to animals and needs to be validated in the context of a clinical application in the future.

This study was an investigation to identify evidence for abnormal NM-MRI signal intensity in NM-containing dopaminergic neurons in the SN after acute MA exposure. Given the central role of DA in drug addiction and the non-invasiveness of NM-MRI data acquisition, NM-MRI has the potential to advance our understanding of DA alterations in the context of drug addiction. Our findings further underscore the promise of NM-MRI as a method to identify clinically useful biomarkers for non-neurodegenerative conditions associated with DA dysfunction.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.

The animal study was reviewed and approved by the institutional review board in the Second Xiangya Hospital of Central South University.

JL and FT: conception and design. JL and HL: administrative support. FT, XZ, and HZ: collection and data assembly. JL, FT, WY, YMD, YYD, and LZ: data analysis and interpretation. All authors wrote the manuscript and approved the manuscript.

YMD and LZ were employed by the company Shanghai United Imaging Healthcare Co., Ltd. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.