Development and patterning of a highly versatile visual system in spiders

Visual systems provide a key interface between organisms and their surroundings, and have evolved in many forms to perform diverse functions across the animal kingdom. Spiders exhibit a range of visual abilities and ecologies, the diversity of which is underpinned by a highly versatile, modular visual system architecture. This typically includes eight eyes of two developmentally distinct types, but the number, size, location, and function of the eyes can vary dramatically between lineages. Previous studies of visual system development in spiders have confirmed that many components of the retinal determination gene (RDG) network are conserved with other arthropods, but so far, comparative studies among spiders are lacking. We characterised visual system development in seven species of spiders representing a range of morphologies, visual ecologies, and phylogenetic positions, to determine how these diverse configurations are formed, and how they might evolve. Combining synchrotron radiation tomography, transcriptomics, in situ hybridisation, and selection analyses, we characterise the repertoires and expression of key RDGs in relation to adult morphology. We identify key molecular players, timepoints, and developmental events that may contribute to adult diversity, in particular the molecular and developmental underpinnings of eye size, number, position, and identity across spiders

Methods
Embryos were preserved in 70% ethanol and mounted in pipette tips for scanning at the TOMCAT beamline, Swiss Light Source (Paul Scherrer Institute) (Stampanoni et al. 2006). Scans used a monochromatic beam (16 keV), with propagation distances 10-100 mm and combined magnifications of 4x (effective voxel size 1.65 μm, LuAG:Ce 100 μm scintillator), 10x (0.65 μm, LuAG:Ce 20 μm), or 20x (0.325 μm, LuAG:Ce 20 μm). 2000 projections were recorded across 180° sample rotation, with exposure times of 70 ms (4x), 120 ms (10x), or 200 ms (20x). In-house software reconstructed slices from projections, with Paganin filtering (delta=1e-7, beta=3e-9) (Paganin et al. 2002) and depth reduction to 8-bit tiffs (Marone and Stampanoni 2012). Data were cropped in Fiji (Schindelin et al. 2012) and models were produced in Amira v.2021 (Thermo Fisher), using volume rendering (outer surfaces) and manual segmentation (CNS).

视觉系统是生物体与其生存环境之间的关键交互界面，在动物界中演化出多种形态以执行多样化功能。蜘蛛具备多样的视觉能力与生态特征，其视觉系统的多样性依托于高度灵活的模块化视觉系统架构。该架构通常包含8只眼，分为两种发育起源不同的类型，但不同演化支间的眼睛数量、大小、位置与功能均存在显著差异。过往针对蜘蛛视觉系统发育的研究已证实，视网膜决定基因（retinal determination gene, RDG）网络的诸多组分与其他节肢动物具有保守性，但目前仍缺乏蜘蛛类群间的比较研究。本研究对涵盖不同形态、视觉生态与系统发育位置的7种蜘蛛的视觉系统发育进行了表征，以阐明这些多样化视觉构型的形成机制与演化路径。研究结合同步辐射断层扫描、转录组学、原位杂交与选择分析，解析了关键RDG的基因家族组成及其与成体形态相关的表达模式。本研究鉴定出可能参与成体视觉系统多样性形成的关键分子因子、发育时间节点与发育事件，特别是蜘蛛类群间眼睛大小、数量、位置与身份的分子与发育基础。

方法
将胚胎保存于70%乙醇中，并固定于移液器吸头内，于瑞士光源（保罗·谢勒研究所）TOMCAT光束线进行扫描（Stampanoni等，2006）。扫描采用16 keV单色光束，传播距离为10~100 mm，组合放大倍率分别为4×（有效体素尺寸1.65 μm，搭配LuAG:Ce 100 μm闪烁体）、10×（0.65 μm，LuAG:Ce 20 μm）或20×（0.325 μm，LuAG:Ce 20 μm）。在样品180°旋转范围内记录2000幅投影图像，曝光时间分别为70 ms（4×）、120 ms（10×）或200 ms（20×）。使用自研软件从投影图像重建切片，并采用帕金滤波（delta=1e-7，beta=3e-9）（Paganin等，2002），随后将图像深度压缩为8位TIFF格式（Marone与Stampanoni，2012）。数据在Fiji软件中进行裁剪，并使用Amira v.2021（赛默飞世尔科技）构建模型，采用体绘制（外表面）与手动分割（中枢神经系统，Central Nervous System, CNS）两种方式。