GABA, not BOLD, reveals dissociable learning-dependent plasticity mechanisms in the human brain.
GABA, not BOLD, reveals dissociable learning-dependent plasticity mechanisms in the human brain.
复制标题
DOI：
10.7554/elife.35854
复制
发表时间：
2018-10-25
期刊：
影响因子：
7.7
通讯作者：
Kourtzi Z
中科院分区：
文献类型：
--
作者：
Frangou P;Correia M;Kourtzi Z
经验和训练已经被证明有助于我们从杂乱的环境中提取和区分有意义的模式。然而，通过抑制嘈杂和不相关的信号来调节我们学习能力的人脑机制在很大程度上仍然未知。为了测试抑制在知觉学习中的作用，我们将联合收割机与GABA的MR光谱测量相结合，因为单独的fMRI不能让我们区分抑制机制与兴奋机制。我们的研究结果表明，任务依赖性GABA能抑制与脑功能可塑性和行为改善有关。具体而言，枕颞叶皮层中的GABA能抑制与可分离的学习机制有关：噪声过滤的GABA减少，而特征模板重新调整的GABA增加。在tDC训练期间干扰皮质兴奋性以特定于任务的方式改变表现，为抑制和行为改善之间的直接联系提供了证据。我们的研究结果提出了可分离的GABA能机制，通过训练优化我们做出感知决策的能力。当你在人群中寻找朋友或分辨同卵双胞胎时，你的视觉系统必须解决一个复杂的难题。它必须忽略所有不相关的信息（例如，人群中不熟悉的面孔）并关注关键特征（例如，你朋友熟悉的面孔），这将使你做出决定。通过练习，我们能够更好地解决复杂的视觉辨别问题。但大脑究竟是如何实现这种改进的性能还不清楚。为了回答这个问题，Frangou等人对健康志愿者进行了两项视觉任务的训练。第一个（目标检测任务）涉及定位目标（例如，由背景中随机分布的点组成的圆形），该任务类似于在人群中识别朋友。第二个（特征辨别任务）涉及在两个不同的类别中分配高度相似的形状，类似于区分同卵双胞胎。为了解决这个问题，志愿者必须识别出使他们能够识别这些形状的独特特征。在这项任务的训练过程中，他们更新和完善了这些不同特征在大脑中的表现。这使他们能够进行更精细的区分，并将每个图像正确地分配到两个类别中的一个。当志愿者接受任务训练时，Frangou等人测量了大脑处理视觉信息区域中一种名为GABA的化学物质的水平。GABA是大脑的主要抑制分子，控制神经元的活动。当志愿者学习这两项任务时，他们的大脑中GABA水平出现了相反的变化。在第一个目标检测任务中，如果训练期间GABA减少，个体的表现会更好。在第二个特征辨别任务中，如果他们的GABA在训练期间增加，他们的成绩会更好。为了证实这些发现，Frangou等人使用了第二种技术来激活或抑制大脑视觉区域的处理。激活视觉区域增强了目标检测任务的性能。抑制它们增强了精细辨别任务的表现。因此，这些变化与GABA水平中观察到的变化一致。除了揭示我们如何学会根据眼睛的信息做出决定外，这些发现还表明，调整大脑活动可以帮助患者恢复因眼睛相关或神经系统疾病而失去的技能。了解GABA在大脑可塑性中的作用也与自闭症和精神病等疾病有关，这些疾病已被证明与大脑抑制的变化有关。
Experience and training have been shown to facilitate our ability to extract and discriminate meaningful patterns from cluttered environments. Yet, the human brain mechanisms that mediate our ability to learn by suppressing noisy and irrelevant signals remain largely unknown. To test the role of suppression in perceptual learning, we combine fMRI with MR Spectroscopy measurements of GABA, as fMRI alone does not allow us to discern inhibitory vs. excitatory mechanisms. Our results demonstrate that task-dependent GABAergic inhibition relates to functional brain plasticity and behavioral improvement. Specifically, GABAergic inhibition in the occipito-temporal cortex relates to dissociable learning mechanisms: decreased GABA for noise filtering, while increased GABA for feature template retuning. Perturbing cortical excitability during training with tDCs alters performance in a task-specific manner, providing evidence for a direct link between suppression and behavioral improvement. Our findings propose dissociable GABAergic mechanisms that optimize our ability to make perceptual decisions through training. When searching for a friend in the crowd or telling identical twins apart, your visual system must solve a complex puzzle. It must ignore all irrelevant information (e.g., unfamiliar faces in the crowd) and focus on key features (e.g., your friend’s familiar face) that will allow you to make a decision. We become better at solving complex visual discriminations with practice. But exactly how the brain achieves this improved performance is unclear. To answer this question, Frangou et al. trained healthy volunteers on two such visual tasks. The first (target detection task) involved locating a target (e.g. circular shape made of dots among randomly distributed dots in the background), a task similar to identifying a friend in the crowd. The second (feature discrimination task) involved assigning highly alike shapes in two different categories, similar to telling apart identical twins. To solve this problem, volunteers had to identify distinct features that allowed them to distinguishthese shapes. During training on this task, they updated and refined the representation of these distinct features in their brain. This enabled them to make finer discriminations and assign each image correctly to one of the two categories. While the volunteers trained on the tasks, Frangou et al. measured levels of a chemical called GABA in brain areas that process visual information. GABA is the brain's main inhibitory molecule and controls the activity of neurons. As the volunteers learned the two tasks, their brains showed opposite changes in GABA levels. In the first, target detection task, individuals did better if their GABA decreased during training. In the second, feature discrimination task, they achieved more if their GABA increased during training. To confirm these findings, Frangou et al. used a second technique to activate or suppress processing in visual areas of the brain. Activating visual areas enhanced performance on the target detection task. Suppressing them enhanced performance on the fine discrimination task. These changes are thus consistent with those seen in GABA levels. As well as revealing how we learn to make decisions based on the information from our eyes, these findings suggest that adjusting brain activity could help patients regain skills lost as a result of eye-related or neurological conditions. Understanding the role of GABA in brain plasticity is also relevant to conditions like autism and psychosis, which have been shown to relate to changes in brain inhibition.