Network recruitment of glia-derived fast-spiking interneurons into sensory circuits using 2-photon calcium imaging

直接谱系重编程已经成为将脑驻留的非神经元细胞原位转化为神经元的有力策略。除了修复急性CNS损伤（例如中风或创伤后）外，开发此类技术还可以为神经系统或神经精神疾病提供治疗选择，其中特定神经元亚型的丧失或功能减退可能参与发病机制。例如，GABA能皮质中间神经元功能障碍在各种神经精神疾病（包括ASD和精神分裂症）中的作用越来越受到关注1 -5。因此，产生特定的中间神经元亚型并了解它们如何在功能上整合到发育和成人皮质中是细胞和再生医学领域的关键目标。开创性工作表明，出生后的小鼠皮质星形胶质细胞可以在体外转化为谷氨酸能或GABA能神经元6-9。从那时起，随着发现各种重编程因子能够在体内将一系列起始细胞群转化为特定的神经元亚型，在不同的疾病模型和大脑区域中，对该领域的兴趣激增（最近在10中进行了综述）。ASCL 1是第一个前神经转录因子成功地重新编程神经胶质细胞的神经元6，8。根据起始细胞群、脑区域以及与其他重编程因子的组合，强制Ascl 1表达可导致体内各种神经元表型11 -16。例如，Parmer实验室发现Ascl 1，Lmx 1a和Nurr 1的3因子组合能够将纹状体中的NG 2-胶质细胞转化为表达小清蛋白（PV）的神经元13。Berninger实验室未发表的工作现在已经表明，六个Ascl 1磷酸化位点17 -19的丝氨酸到丙氨酸突变不仅提高了皮质星形胶质细胞体内重编程的效率，而且产生了一个快速尖峰PV中间神经元的子集，而不需要Lmx 1a或Nurr 1。此外，Ascl 1 SA 6衍生的诱导神经元（iN）表现出突触电流，表明它们在预先存在的电路中的整合。该策略利用逆转录病毒构建体和抗铁凋亡因子Bcl 2，其显示在重编程期间促进细胞存活20。然而，逆转录病毒载体仅能够抑制分裂细胞，从而减少了用于重编程的起始细胞的潜在数量。最近在直接重编程领域的一些显著的成功案例已经采用腺相关病毒（AAV）作为基因转移系统11，21，22。这导致了一个问题，即是否有可能的交付系统本身可能有助于更大的重新编程效率10，23，这是缺乏直接证据。无论哪种方式，AAV作为基因递送系统具有一些优势，因为它们能够感染非分裂细胞，免疫原性较低，并且比逆转录病毒载体具有更低的插入诱变风险。然而，这是以有限的遗传承载能力为代价的。因此，这将是重要的是要检查一些这些假定的好处，使用已经建立良好的重编程constructs.While皮质胶质细胞可以在体内重编程成iNs，我们缺乏关键的信息是否可以招募这样的iNs功能神经元电路，如引起的感觉刺激。迄今为止，功能表征和表型已进行解剖皮质或纹状体脑切片。目前尚不清楚的是，这些诱导的神经元是否以及如何影响体内区域皮质网络活动。这是更好地理解并最终将这些技术转化为临床应用的自然和必要的下一步。

Direct lineage reprogramming has emerged as a powerful strategy for in situ conversion of brain-resident non-neuronal cells into neurons. In addition to repairing acute CNS damage (e.g. following stroke or trauma), developing such technology may unlock treatment options for neurological or neuropsychiatric disorders where loss or hypofunction of specific neuronal subtypes may be involved in pathogenesis. For example, attention has been increasingly paid toward the role of GABAergic cortical interneuron dysfunction in various neuropsychiatric disorders - including ASD and schizophrenia1-5. Thus, generating specific interneuron subtypes and understanding how they functionally integrate into the developing and adult cortex is a crucial goal for the field of cellular and regenerative medicine.Pioneering work demonstrated that postnatal mouse cortical astrocytes could be converted into glutamatergic or GABAergic neuronal identities in vitro 6-9. Since then, interest in the field has exploded with the discovery of a variety of reprogramming factors able to convert a range of starting cell populations into specific neuronal subtypes in vivo, across different disease models and brain regions (recently reviewed in 10). Ascl1 was among the first proneural transcription factors shown to successfully reprogramme glia to neurons6,8. Depending on the starting-cell population, brain region and combination with other reprogramming factors, forced Ascl1 expression can result in various neuronal phenotypes in vivo11-16. For example, the Parmer laboratory found that a 3-factor combination of Ascl1, Lmx1a, and Nurr1 was able to convert NG2-glia in the striatum into parvalbumin (PV) expressing neurons13. Unpublished work by the Berninger lab, has now shown that serine-to-alanine mutations at six Ascl1 phosphorylation sites17-19 not only improves efficiency of in vivo reprogramming of cortical astrocytes but produces a subset of fast-spiking, PV interneurons, without the need for Lmx1a or Nurr1. Moreover, Ascl1SA6 derived induced neurons (iNs) exhibited synaptic currents, suggesting their integration in pre-existing circuits. This strategy utilises a retroviral construct and the anti-ferroptotic factor Bcl2, shown to promote cell-survival during reprogramming20. However, retroviral vectors are only able to transduce dividing cells, reducing the potential number of starting cells for reprogramming. Some remarkable recent success stories in the field of direct reprogramming have employed adeno-associated viruses (AAV) as a gene transfer system11,21,22. This has led to the question of whether it is possible that the delivery system itself could be contributing to greater reprogramming efficiency10,23, for which direct evidence is lacking. Either way, AAVs have some advantages as gene delivery systems, in that they are able to infect non-dividing cells, are less immunogenic and have lower risk of insertional mutagenesis than do retroviral vectors. This is at the expense of a restricted genetic carrying capacity, however. Therefore, it will be important to examine some of these putative benefits using already well-established reprogramming constructs.While cortical glia can be reprogrammed into iNs in vivo, we lack crucial information whether such iNs can be recruited into functional neuronal circuits such as elicited by sensory stimulation. To date, functional characterisation and phenotyping has been conducted in dissected cortical or striatal brain slices. What is currently unknown, is whether and how these induced neurons impact regional cortical network activity in vivo. This is a natural and necessary next step toward better understanding and, ultimately, clinical translation of these technologies.