Modelling human neuronal catecholaminergic pigmentation in rodents recapitulates age-related neurodegenerative deficits

Abstract

One key limitation in developing effective treatments for neurodegenerative diseases is the lack of models accurately mimicking the complex physiopathology of the human disease. Humans accumulate with age the pigment neuromelanin inside neurons that synthesize catecholamines. Neurons reaching the highest neuromelanin levels preferentially degenerate in Parkinson's, Alzheimer's and apparently healthy aging individuals. However, this brain pigment is not taken into consideration in current animal models because common laboratory species, such as rodents, do not produce neuromelanin. Here we generate a tissue-specific transgenic mouse, termed tgNM, that mimics the human age-dependent brain-wide distribution of neuromelanin within catecholaminergic regions, based on the constitutive catecholamine-specific expression of human melanin-producing enzyme tyrosinase. We show that, in parallel to progressive human-like neuromelanin pigmentation, these animals display age-related neuronal dysfunction and degeneration affecting numerous brain circuits and body tissues, linked to motor and non-motor deficits, reminiscent of early neurodegenerative stages. This model could help explore new research avenues in brain aging and neurodegeneration.

Fig. 1. Accumulation of human-like NM in PD-vulnerable catecholaminergic brain nuclei from TgNM mice.

A, C From left to right, unstained macroscopic view, NM-sensitive high-resolution T1-weighted magnetic resonance imaging, hematoxilin-eosin (H&E) and Masson-Fontana (M-F) staining of aged healthy human postmortem brains (A), wt (B), and tgNM (C) mice. One micrograph from more than three high-magnification images from two representative specimens was selected and displayed. C (right panel), macroscopic view of a clarified tgNM brain. B Unstained macroscopic view (left) and NM-sensitive high-resolution T1-weighted magnetic resonance imaging (right) from wild-type (wt) mice acquired ex vivo without fat saturation. Unstained NM, brown (see also Movie S1). Dashed lines outline the SN/VTA and LC regions in (A–C). Scale bars: 20 µm (H&E), 5 µm (M-F). Macroscopic human midbrain image adapted from.

Fig. 2. Brain-wide NM distribution and accumulation of human-like NM in PD-vulnerable catecholaminergic brain nuclei from TgNM mice.

A Qualitative mouse brain atlas showing NM accumulation in all catecholaminergic groups in tgNM mice (see also Supplementary Table 1). Left, Mouse brain atlas. Number of dots represents the quantity of NM-accumulating cells and dot color represents the levels of intracellular NM from very low levels (blue) to higher levels (red). Image credit: Allen Institute. Right chatecholaminergic areas immunostained for TH in wt mice and Nissl-stained in tgNM mice. Unstained NM, brown. Scale bars: 25 µm, 25 µm (inset). B Quantification of intracellular NM levels in unstained SN/A9, VTA/A10, LC/A6, and DVC/A2 brain sections from tgNM mice *p ≤ 0.05 compared with A9, #p ≤ 0.05 compared with A10, $p ≤ 0.05 compared with A2. Box plots: median, first and third quartile, min-max values and individual dots for each neuron. Raw data, genotypes, ages, sample sizes and statistical analyses are provided as a Source Data file.

Fig. 3. Dopaminergic dysfunction in tgNM mice.

A Time to cross the beam in seconds (s). B Olfactory discrimination index. C Percentage of animals vocalizing. D SNpc and VTA immunostained sections. TH, blue; unstained NM, brown. Scale bars: 500 µm, 20 µm (inset). E Cell counts of SNpc and VTA TH-positive neurons. F Left, quantification of TH downregulation measured as the percentage (%) of TH-immunonegative neurons within the total population of NM-containing neurons in tgNM mice. Right, TH-immunostained SNpc showing TH⁻NM⁺ (arrow) and TH⁺NM⁺ (arrowhead) neurons. G Cell counts of total SNpc and VTA dopaminergic neurons. A–C, E, F *p ≤ 0.05 compared with wt littermates. Box plots: median, first and third quartile, min-max values and individual dots for each neuron. Raw data, genotypes, ages, sample sizes and statistical analyses are provided as a Source Data file.

Fig. 4. PD-like neuropathology in tgNM mice.

A, B Left, SNpc and VTA sections exhibiting NM-laden neurons with cytoplasmic (A) and nuclear (B) inclusions. TH, purple; p62, green; alpha-synuclein (Syn), red; Hoechst (blue); NM, dark gray. Scale bar: 10 µm. Right, quantification of SNpc and VTA NM-laden neurons with p62-positive cytoplasmic (A) and nuclear (B) inclusions. C Percentage of SN-VTA cytoplasmic inclusions positive or negative for Syn in tgNM at different ages. D Left, reactive (top) and non-reactive (middle) Iba-1-positive microglia (blue) and GFAP-positive astrocytes (bottom, blue). Scale bar: 20 µm. Right, quantification of reactive Iba-1- and GFAP-positive cells in SNpc and VTA sections. B *p ≤ 0.05 compared with young tgNM. D *p ≤ 0.05 compared with wt littermates. Box plots: median, first and third quartile, min-max values and individual dots for each neuron. Raw data, genotypes, ages, sample sizes and statistical analyses are provided as a Source Data file.

Fig. 5. Noradrenergic neurodegeneration and PD-like neuropathology in tgNM mice.

A Percentage of time spent and distance traveled in the periphery compared with the center of an open field. B Time spent on the platform of a step-down test. C Number and mean duration (s, seconds) of paradoxical sleep (PS) bouts. D Left, LC sections immunostained for TH. TH, blue; unstained NM, brown. Scale bars: 200 µm, 20 µm (inset). Right, cell counts of LC TH-positive neurons and total NA neurons at different ages. E Number of eNM granules in tgNM mice. F LC and prefrontal cortex (PFC) NA levels in young/old tgNM and wt mice. G Left, LC section exhibiting a NM-laden neuron with a cytoplasmic inclusion. TH, purple; p62, green; alpha-synuclein (Syn), red; Hoechst (blue); NM, dark gray. Scale bar: 5 µm. Right, quantification of LC NM-laden neurons with p62-positive cytoplasmic inclusions in tgNM mice. H Images and quantification of reactive Iba-1- and GFAP-positive cells in LC sections. A–D, F, H *p ≤ 0.05 compared with wt littermates. Box plots: median, first and third quartile, min-max values and individual dots for each neuron. Raw data, genotypes, ages, sample sizes and statistical analyses are provided as a Source Data file.

Fig. 6. Cholinergic and serotonergic alterations in tgNM mice.

A, B Left, NBM and PPN sections immunostained for choline acetyltransferase (ChAT). ChAT, blue; unstained NM, brown. Scale bars: 500 µm, 20 µm (inset). Right, cell counts of NBM and PPN ChAT-positive neurons in old tgNM and wt mice. C PPN and prefrontal cortex (PFC) acetylcholine (Ach) levels. D Left, dorsal raphe (DR) sections immunostained for tryptophan hydroxylase (TPH). TPH, blue; unstained NM, brown. Scale bars: 200 µm, 20 µm (inset). Right, Cell counts of DR TPH-positive neurons in old tgNM and wt mice. E DR and PFC serotonin (5HT) levels. F Immobilization time in the tail suspension test (s, seconds). A–C, F *p ≤ 0.05 compared with wt littermates. Box plots: median, first and third quartile, min-max values and individual dots for each neuron. Raw data, genotypes, ages, sample sizes and statistical analyses are provided as a Source Data file.

Fig. 7. Medullary catecholaminergic nuclei alterations in tgNM mice.

A Medullary sections immunostained for TH and ChAT. TH, red; ChAT, blue; unstained NM, brown. Scale bars: 200 µm, 10 µm (insets). Dorsal motor nucleus of the vagus nerve, DNV; dorsal vagal complex, DVC. B, C Cell counts of TH-positive (B) and total (C) pigmented catecholaminergic cells. D Left, image of eNM granules (white arrowhead). Scale bar: 25 µm. Right, number of eNM granules in A2 neuronal group in tgNM mice. E Images and quantification of Iba-1-positive reactive microglia in tgNM and wt mice. F Top, DVC section exhibiting a NM-laden neuron with a cytoplasmic inclusion. TH, purple; p62, green; alpha-synuclein (Syn), red; Hoechst (blue); NM, dark gray. Scale bar: 5 µm. Bottom, quantification of DVC NM-laden neurons with p62-positive cytoplasmic inclusions in tgNM mice. B, D *p ≤ 0.05 compared with young tgNM. E *p ≤ 0.05 compared with wt littermates. F p ≤ 0.05 compared with adult and old tgNM. Box plots: median, first and third quartile, min-max values and individual dots for each neuron. Raw data, genotypes, ages, sample sizes and statistical analyses are provided as a Source Data file.

Fig. 8. Autonomic dysfunction and decreased lifespan in tgNM mice.

A Schematic diagram of the peripheral A2 innervation sites. B Heart rate (bpm, beats per minute) in tgNM and wt mice. C Systolic and diastolic blood pressure in old tgNM and wt mice. D Respiratory rate (number of breaths/min) in old tgNM and wt mice. E Intestinal transit time (min, minutes) in old tgNM and wt mice. F Levels of DA, NA and Ach in DVC and peripheral organs from tgNM and wt mice. G Survival analysis of tgNM mice (black) compared with wt (gray). Dotted lines represent 95% confidence interval for each genotype. m, months. B–G *p ≤ 0.05 compared with wt littermates. Box plots: median, first and third quartile, min-max values and individual dots for each neuron. Raw data, genotypes, ages, sample sizes and statistical analyses are provided as a Source Data file.