Binding of high mobility group A proteins to the mammalian genome occurs as a function of AT-content

Genomic location can inform on potential function and recruitment signals for chromatin-associated proteins. High mobility group (Hmg) proteins are of similar size as histones with Hmga1 and Hmga2 being particularly abundant in replicating normal tissues and in cancerous cells. While several roles for Hmga proteins have been proposed we lack a comprehensive description of their genomic location as a function of chromatin, DNA sequence and functional domains. Here we report such a characterization in mouse embryonic stem cells in which we introduce biotin-tagged constructs of wild-type and DNA-binding domain mutants. Comparative analysis of the genome-wide distribution of Hmga proteins reveals pervasive binding, a feature that critically depends on a functional DNA-binding domain and which is shared by both Hmga proteins. Assessment of the underlying queues instructive for this binding modality identifies AT richness, defined as high frequency of A or T bases, as the major criterion for local binding. Additionally, we show that other chromatin states such as those linked to cis-regulatory regions have little impact on Hmga binding both in stem and differentiated cells. As a consequence, Hmga proteins are preferentially found at AT-rich regions such as constitutively heterochromatic regions but are absent from enhancers and promoters arguing for a limited role in regulating individual genes. In line with this model, we show that genetic deletion of Hmga proteins in stem cells causes limited transcriptional effects and that binding is conserved in neuronal progenitors. Overall our comparative study describing the in vivo binding modality of Hmga1 and Hmga2 identifies the proteins’ preference for AT-rich DNA genome-wide and argues against a suggested function of Hmga at regulatory regions. Instead we discover pervasive binding with enrichment at regions of higher AT content irrespective of local variation in chromatin modifications.

基因组定位可为染色质结合蛋白（chromatin-associated proteins）的潜在功能及招募信号提供参考依据。高迁移率族蛋白（High Mobility Group, HMG）与组蛋白大小相近，其中Hmga1与Hmga2在增殖正常组织及癌细胞中含量尤为丰富。尽管已有多项研究提出Hmga蛋白的多种潜在功能，但目前仍缺乏对其基因组定位随染色质、DNA序列及功能结构域变化的全面阐述。本研究以小鼠胚胎干细胞为模型，构建了野生型（wild-type）及DNA结合结构域（DNA-binding domain）突变体的生物素标记（biotin-tagged）重组载体，并完成了相关表征分析。对Hmga蛋白全基因组分布的比较分析显示，二者均存在广泛的染色质结合现象，这一特征严格依赖于具备功能活性的DNA结合结构域，且为两类Hmga蛋白所共有。对指导该结合模式的潜在序列特征进行分析后发现，AT富集度——即腺嘌呤（A）与胸腺嘧啶（T）碱基的高占比——是决定局部结合的核心标准。此外，本研究证实，在干细胞及分化细胞中，与顺式调控区域（cis-regulatory regions）相关的其他染色质状态对Hmga蛋白的结合几乎无影响。因此，Hmga蛋白优先富集于AT富集区域，如组成型异染色质区域，而无法结合增强子（enhancers）与启动子（promoters）区域，这表明其在单个基因调控中的作用十分有限。与该模型相符，本研究发现干细胞中Hmga蛋白的基因敲除仅会引发有限的转录组变化，且其结合模式在神经前体细胞中保守存在。综上，本项针对Hmga1与Hmga2体内结合模式的比较研究，明确了两类蛋白在全基因组范围内对AT富集DNA的偏好性，并驳斥了此前提出的Hmga蛋白在调控区域发挥功能的假说。相反，本研究发现两类蛋白存在广泛的染色质结合，且富集于AT含量更高的区域，不受染色质修饰局部差异的影响。