Dataset Information

Implementation of deep neural networks to count dopamine neurons in substantia nigra.

ABSTRACT: Unbiased estimates of neuron numbers within substantia nigra are crucial for experimental Parkinson's disease models and gene-function studies. Unbiased stereological counting techniques with optical fractionation are successfully implemented, but are extremely laborious and time-consuming. The development of neural networks and deep learning has opened a new way to teach computers to count neurons. Implementation of a programming paradigm enables a computer to learn from the data and development of an automated cell counting method. The advantages of computerized counting are reproducibility, elimination of human error and fast high-capacity analysis. We implemented whole-slide digital imaging and deep convolutional neural networks (CNN) to count substantia nigra dopamine neurons. We compared the results of the developed method against independent manual counting by human observers and validated the CNN algorithm against previously published data in rats and mice, where tyrosine hydroxylase (TH)-immunoreactive neurons were counted using unbiased stereology. The developed CNN algorithm and fully cloud-embedded Aiforia™ platform provide robust and fast analysis of dopamine neurons in rat and mouse substantia nigra.

SUBMITTER: Penttinen AM

PROVIDER: S-EPMC6585833 | biostudies-literature | 2018 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Implementation of deep neural networks to count dopamine neurons in substantia nigra.

Penttinen Anna-Maija AM Parkkinen Ilmari I Blom Sami S Kopra Jaakko J Andressoo Jaan-Olle JO Pitkänen Kari K Voutilainen Merja H MH Saarma Mart M Airavaara Mikko M

The European journal of neuroscience 20180920 6

Unbiased estimates of neuron numbers within substantia nigra are crucial for experimental Parkinson's disease models and gene-function studies. Unbiased stereological counting techniques with optical fractionation are successfully implemented, but are extremely laborious and time-consuming. The development of neural networks and deep learning has opened a new way to teach computers to count neurons. Implementation of a programming paradigm enables a computer to learn from the data and developmen ...[more]

PMID: 30144349

Similar Datasets

Project description:Dopamine midbrain neurons within the substantia nigra are particularly prone to degeneration in Parkinson's disease. Their selective loss causes the major motor symptoms of Parkinson's disease, but the causes for the high vulnerability of SN DA neurons, compared to neighbouring, more resistant ventral tegmental area dopamine neurons, are still unclear. Consequently, there is still no cure available for Parkinson's disease. Current therapies compensate the progressive loss of dopamine by administering its precursor l-DOPA and/or dopamine D2-receptor agonists. D2-autoreceptors and Cav1.3-containing L-type Ca(2+) channels both contribute to Parkinson's disease pathology. L-type Ca(2+) channel blockers protect SN DA neurons from degeneration in Parkinson's disease and its mouse models, and they are in clinical trials for neuroprotective Parkinson's disease therapy. However, their physiological functions in SN DA neurons remain unclear. D2-autoreceptors tune firing rates and dopamine release of SN DA neurons in a negative feedback loop through activation of G-protein coupled potassium channels (GIRK2, or KCNJ6). Mature SN DA neurons display prominent, non-desensitizing somatodendritic D2-autoreceptor responses that show pronounced desensitization in PARK-gene Parkinson's disease mouse models. We analysed surviving human SN DA neurons from patients with Parkinson's disease and from controls, and detected elevated messenger RNA levels of D2-autoreceptors and GIRK2 in Parkinson's disease. By electrophysiological analysis of postnatal juvenile and adult mouse SN DA neurons in in vitro brain-slices, we observed that D2-autoreceptor desensitization is reduced with postnatal maturation. Furthermore, a transient high-dopamine state in vivo, caused by one injection of either l-DOPA or cocaine, induced adult-like, non-desensitizing D2-autoreceptor responses, selectively in juvenile SN DA neurons, but not ventral tegmental area dopamine neurons. With pharmacological and genetic tools, we identified that the expression of this sensitized D2-autoreceptor phenotype required Cav1.3 L-type Ca(2+) channel activity, internal Ca(2+), and the interaction of the neuronal calcium sensor NCS-1 with D2-autoreceptors. Thus, we identified a first physiological function of Cav1.3 L-type Ca(2+) channels in SN DA neurons for homeostatic modulation of their D2-autoreceptor responses. L-type Ca(2+) channel activity however, was not important for pacemaker activity of mouse SN DA neurons. Furthermore, we detected elevated substantia nigra dopamine messenger RNA levels of NCS-1 (but not Cav1.2 or Cav1.3) after cocaine in mice, as well as in remaining human SN DA neurons in Parkinson's disease. Thus, our findings provide a novel homeostatic functional link in SN DA neurons between Cav1.3- L-type-Ca(2+) channels and D2-autoreceptor activity, controlled by NCS-1, and indicate that this adaptive signalling network (Cav1.3/NCS-1/D2/GIRK2) is also active in human SN DA neurons, and contributes to Parkinson's disease pathology. As it is accessible to pharmacological modulation, it provides a novel promising target for tuning substantia nigra dopamine neuron activity, and their vulnerability to degeneration.

Dataset Information

Implementation of deep neural networks to count dopamine neurons in substantia nigra.

Publications

Implementation of deep neural networks to count dopamine neurons in substantia nigra.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets