电镜：看到的是结构，不是结论
方法学原文精译 + 判读与伪影指南｜STAR Protocols 2022｜原文见文末附录。带红色下划线的词，悬停或点击可看注释（共 12 条）；中文正文按原文步骤重排，参数均出自原文，经验值单独标注。
电镜把分辨率推到纳米级，让你第一次"看见"髓鞘的板层、突触的囊泡、线粒体的嵴。但它给的始终是一张结构快照：它不告诉你这条膜上是什么受体，也不告诉你这个突触是兴奋还是抑制，更不告诉你结构变化与行为之间谁先谁后。这一页用一篇公开的 STAR Protocols 方案（小鼠脑组织的多尺度光镜/电镜联用）做骨架，讲清三件事：样品是怎么变成一张电镜图的、每一步失败会在图上留下什么、以及一个本科生现在能亲手把哪几步做好。
本篇精译自 STAR Protocols 2022 年的方法学文章《Protocol for multi-scale light microscopy/electron microscopy neuronal imaging in mouse brain tissue》，原文见文末附录。译文按原文的 Before you begin、Step-by-step method details、Expected outcomes、Limitations 与 Troubleshooting 重排；原文中的图、参考文献编号与部分试剂表在文本抽取时未保留。
一、电镜能回答什么，不能回答什么
把一块脑组织送进透射电镜（transmission electron microscopy, TEM），你看到的不是"放大的照片"，而是电子穿过样品后的明暗分布：锇、铀、铅这些重金属沉积得多的地方，电子被散射得多，成像就暗。细胞膜、髓鞘、突触的致密带因此显出来。分辨率到纳米级——这篇 protocol 的成像像素为 1.2–16.1 nm/pixel，加速电压 80 kV——所以约 5 nm 的脂双层、20 nm 上下的突触间隙、40 nm 左右的突触囊泡，都在可分辨的范围内。这就是电镜（electron microscopy, EM）在第 3 篇里的基本定位：它把"结构"这一层单独拎出来给你看。
但"看见结构"和"知道那是什么"是两件事，中间隔着标记。电镜图里的一条膜，你可以说它是一条膜，不能说它是 γ-氨基丁酸（GABA）受体；一颗囊泡，你可以说它是一颗囊泡，不能说它装的是谷氨酸。要给结构加上分子身份，只有两条路：免疫金（immunogold）标记——抗体上挂金颗粒，金颗粒在电镜下是一个高电子密度的黑点；或者基因编码的过氧化物酶标记（本篇用的 APEX2），让特定细胞在反应后沉积电子致密物。没有标记，电镜给的是"超微结构（ultrastructure）"，不是"分子身份"。
比分子身份更容易被越过的，是功能与因果。一张图显示"突触变少了"，只能说明这个时间点、这个采样区域里，形态学上可识别的突触结构少了；它不能说明突触是"丢失"还是"变小到认不出"，更不能说明这就是记忆下降的原因。要谈功能，得用电生理、钙成像、行为学；要谈因果，得有时序、干预和对照。结构、功能、因果是三张不同的网，电镜只负责其中一张。
这一页反复要回到的一句话是：漂亮的电镜图不是证据，可重复、有对照、有抽样、有参数的图才是证据。连这篇 protocol 自己都在 Limitations 里承认：ScaleSF 透明化会"轻度但显著"地损伤超微结构，APEX2/BT-GO 反应和通透本身也可能影响结构，胞质里嗜锇的聚合物还会干扰对活性带、突触后致密物和囊泡形态的判读。作者把丑话说在前面，这正是方法学文章该有的样子。读方法学原文，先读它的 Limitations 和 Troubleshooting，比先读漂亮的 Results 更有用。
还有一层常被忽略的限制：电镜看的是"此刻"。灌流固定的那一刻，结构被定格；此后你数的每一个突触、每一层髓鞘，都属于那个时间点。所以电镜实验的对照，不只是"用药与不用药"，还包括"同一时间点、同一固定批次、同一成像参数"。跨批次比较形态学数字，是天真的：固定液的新旧、灌注的快慢、锇酸的批次，都会改变你数出来的结果。这也是为什么方法学文章要把每一步的参数写得那么细——不是为了让你照抄，而是为了让你知道，哪些变量一旦改变，结论就不再可比。
二、为什么先光镜后电镜：相关的分量
神经元有一个尺度上的矛盾：它的一端在皮层，另一端可能伸到脊髓，轴突总长可以到米级；而它与其他细胞交换信息的地方——突触——只有纳米级。用同一套成像手段同时覆盖这两个尺度是做不到的，所以多尺度成像不是"炫技"，是不得不如此。
这篇 protocol 的思路，是先用光镜找到目标，再用同一块组织做电镜。具体做法：把 AAV2/1-SynTetOff-EGFP-APEX2 立体定位注射进小鼠纹状体（caudate-putamen, CPu），表达 4 周；EGFP 让你在光镜下看见这条投射轴突，APEX2 让同一根轴突在做完 DAB 反应后，在电镜下成为一条电子致密的线。1 mm 厚的脑片先透明化、上共聚焦，确定目标位置；再把同一片组织冷冻切成 40 µm 薄片，做 DAB-Ni 反应和电镜制样。光镜和电镜看的不是"两块相似的组织"，而是同一块组织、同一段轴突。
这就是相关光镜-电镜联用（correlative light and electron microscopy, CLEM）中"相关"两个字的分量。它至少要求三件事同时成立：同一块组织、同一位置、两套图像能配准（registration）。缺一条，都只是把两张图并排放在一起，不叫 CLEM。配准靠的是地标：血管的走行、胞体与树突的形状、DAB-Ni 沉淀的分布。原文的 Expected outcomes 也把话说得很直：要用 GFP 信号、DAB-Ni 沉淀与内源标志（血管、结构形状）三者一起，才能把光镜图和电镜图对上。
为什么值得先花时间做光镜？因为电镜的视野太贵。一张超薄切片只有几十微米见方，想在其中"盲找"一根特定的轴突，几乎等于大海捞针；而在 1 mm 的透明脑片里，共聚焦可以先用低倍物镜把整条投射路径走一遍，把目标坐标记下来，再定向切片。省下来的不是时间，是"切到了但没切到目标"的整批浪费。
三、样品制备链：每一步为什么，失败长什么样
电镜制样的核心矛盾是：你既要把"软"的脑组织变成能切 60–80 nm 薄片的"硬"块，又要尽量不改动它的结构。固定、脱水、包埋，每一步都是妥协；参数不是随便定的，每一步都有它要防的那个失败。下面按原文的步骤顺序讲，关键参数加粗，并给出"这一步错了，图上会长什么样"。
| 步骤 | 要防的失败 | 关键参数（原文） | 失败在图上长什么样 |
|---|---|---|---|
| 灌注固定 | 固定不足、血管残留 | 20 mL 冰冷 PBS → 20 mL 冰冷固定液（4% PFA + 0.2% GA）；后固定 4°C 16–20 h | 膜断裂、细胞器肿胀、胞质"洗空"；血管内红细胞残留 |
| 振动切片 | 厚薄不匀、刀痕 | 4% 琼脂包埋；1 mm；速度 0.14 mm/s；振幅 1.4 mm；频率 75–77 Hz | 切片表面平行刀痕、局部破碎 |
| 通透 | 试剂进不去 | 2% BSA + 0.2% Triton X-100，4 h | 只有组织边缘有 DAB 信号 |
| APEX2/BT-GO | 无沉淀、弱、边缘化 | BT 1:5,000（常用 1:5,000–1:50,000）+ GO 3 µg/mL；β-D-葡萄糖 2 h | 全无深色沉淀；或仅边缘深、中心浅 |
| 透明化 | 不透、荧光衰减 | ScaleS0 2 h、ScaleS4 8–12 h，37°C、90 rpm；总 10.5–14.5 h | 组织发白不透光；共聚焦信号弱 |
| 锇酸与块染 | 膜对比差、脂质流失 | 1% OsO4 2 h 避光；2% UA/50% 乙醇 30 min | 膜轮廓模糊、髓鞘板层不清 |
| 脱水与包埋 | 收缩、空洞、聚合不足 | 乙醇 50/70/90/99.5/99.5/100/100%，各 20 min；环氧丙烷 2×20 min；树脂 60°C 2 天 | 组织内空洞、切片碎裂、无法连片 |
| 超薄切片 | 颤痕、厚度不均 | 60–80 nm；1% UA 20 min；铅柠檬酸 2 min | 周期性条纹（颤痕）、干涉色不匀、染色沉淀 |
第 1 步：灌注固定。先灌 20 mL 冰冷 PBS，再灌 20 mL 冰冷固定液——4% 多聚甲醛（paraformaldehyde, PFA）加 0.2% 戊二醛（glutaraldehyde, GA），溶于 0.1 M 磷酸缓冲液（phosphate buffer, PB）。PBS 的作用是冲掉血液：红细胞里有内源过氧化物酶，如果冲洗不净，后面 DAB 反应会在血管壁产生假阳性深染。PFA 靠交联蛋白把可溶成分固定成不溶网络；戊二醛（glutaraldehyde, GA）交联更强、对膜结构保护更好，但浓度一高，抗原性下降、组织变硬，所以这篇方案只用 0.2%。灌完取脑，在固定液里 4°C 后固定 16–20 h，50–100 rpm 轻摇。这一步失败，后面无法补救：固定不良的组织，膜会断、细胞器会肿、胞质会被"洗空"，你在电镜下看到的是"烂"而不是"结构"。
第 2 步：振动切片。4% 琼脂（40–45°C）包埋，冰上 30 min 凝固；振动切片机切 1 mm 厚，速度 0.14 mm/s、振幅 1.4 mm、频率 75–77 Hz。为什么是 1 mm：太薄，后面透明化和再切片都难操作；太厚，固定液和试剂透不进去。切好的片在 PBS 里 4°C 存 16–20 h。关键提醒：在 APEX2 反应之前，所有溶液都不能含叠氮钠（NaN3），它会灭活 APEX2 的酶活；反应完成之后才可以用 0.02% NaN3 保存。
第 3 步：通透（permeabilization）。2% BSA + 0.2% Triton X-100 的 PBS 溶液，4 h。目的是让后面的生物素-酪胺（biotin-tyramide, BT）和葡萄糖氧化酶（glucose oxidase, GO）进到组织内部。这一步是双刃剑：通透不足，只有切片边缘有信号；通透过度，膜结构本身会受损。原文 Troubleshooting 里"只有边缘有 DAB 信号"的第一条建议，就是确认通透和 BT 孵育时组织是否被充分摇动。
第 4 步：APEX2/BT-GO 反应。BT 按 1:5,000 稀释（原文说常用范围 1:5,000–1:50,000），GO 3 µg/mL，溶于含 2% BSA 的 0.1 M PB，4 h；然后加 12 µL β-D-葡萄糖，反应 2 h；PBS 洗 2×15 min；4% PFA 后固定 16–20 h。原理在第四节展开，这里只记一条操作纪律：反应必须在透明化之前做，因为透明化之后 APEX2 的酶活会明显下降。
第 5 步：透明化。ScaleS0 2 h、37°C、90 rpm；PBS(–) 洗 2×15 min；ScaleS4 8–12 h、37°C、90 rpm。全程 10.5–14.5 h。1 mm 厚的脑片在 14.5 h 内透明（原文数据）。透明化不是为了"好看"，是为了让共聚焦的激光进得去、荧光出得来；具体机制与边界在第五节讲。
第 6 步：装片与共聚焦。透明切片贴在腔底盖玻片上，滴 37°C 的 ScaleS4 胶，盖上盖玻片，用 Kimwipe 吸掉多余胶，再压载玻片和金属重物，4°C 30 min 固化；在 ScaleS4 里平衡 60 min 后上机。物镜用 16× 多浸物镜，数值孔径（numerical aperture, NA）0.60，工作距离 2.5 mm，校正环调到 1.47（ScaleS4 的折射率约 1.47）。校正环不调，点扩散函数失配，三维重建会糊——这是原文 Troubleshooting 里专门列的一条。
第 7 步：再切片。共聚焦看完，去掉盖玻片，4% PFA 再固定 16–20 h；PBS 洗；30% 蔗糖 4°C 泡到沉底；OCT 包埋 15 min；用液氮冷却的异戊烷速冻；冷冻切片机切 40 µm。为什么要再切：共聚焦的 1 mm 厚片没法直接做透射电镜，必须薄到电子能穿过；40 µm 是"能切、能反应、还能保住位置"的折中。这一步最常见的失败是冰晶：异戊烷要接近它的凝固点，蔗糖冷冻保护要充分，否则胞质里会出现成群的空洞。
第 8 步：锇酸、块染、脱水、包埋。1% 锇酸——也就是四氧化锇（osmium tetroxide, OsO4）——在 0.1 M PB 中 2 h、避光。锇酸有两个作用：固定脂质（尤其是细胞膜），并给膜提供电子密度；它剧毒、易挥发，必须在通风橱里操作。接着 2% 醋酸铀（uranyl acetate, UA）的 50% 乙醇溶液块染 30 min、避光；乙醇梯度 50/70/90/99.5/99.5/100/100% 各 20 min 脱水；环氧丙烷 2×20 min 过渡；树脂与环氧丙烷 1:1 渗透 1 h；纯树脂 16–20 h；夹在硅化载玻片与盖玻片之间压平，60°C 聚合 2 天。脱水不彻底，组织会收缩、出现空洞；聚合不足，切片会碎成粉末。
第 9 步：超薄切片与染色。去掉盖玻片，明场显微镜下找到感兴趣区（region of interest, ROI），修块，把树脂块粘到"rocket"上，超薄切片机配钻石刀切 60–80 nm，捞到铜网；1% UA 染 20 min，洗；铅柠檬酸染 2 min，洗、干燥。60–80 nm 不是越薄越好：太薄，切片容易碎、对比度也低；太厚，电子穿不透、结构重叠。这一步的典型失败是颤痕（chatter）——图上出现周期性平行条纹，来自刀钝、块太硬或切片速度不当；还有染色沉淀——不规则的黑色团块，和任何结构都对不上。
第 10 步：透射电镜观察。80 kV 加速电压，灯丝电流约 20 V；像素尺寸 1.2–16.1 nm/pixel；图像用 ImageJ 处理。到这里，一块脑组织才变成你在论文里看到的那张灰度图。回看整条链，你会发现：真正决定图好不好看的，往往不是电镜本身，而是前面那些"看不见的"固定、脱水、包埋和切片。
四、把分子身份写进电镜图：APEX2 与 DAB-Ni
电镜本身不认分子。要让某类细胞或某个蛋白在电镜下"自带标记"，常见两条路。第一条是免疫金（immunogold）：把 5–15 nm 的金颗粒挂在抗体上，金颗粒电子密度高，在图上就是一个清清楚楚的黑点，数黑点就能定量。它的优点是分子特异性来自抗体、空间定位直接；缺点是抗体和大颗粒都难穿透组织，尤其在致密脑组织里，标记往往只停留在切片表面。第二条是基因编码的过氧化物酶：把一种辣根过氧化物酶（horseradish peroxidase, HRP）的工程化版本——APEX2——与目标蛋白融合，或者放在特定启动子下表达，再用底物反应在酶周围沉积电子致密物。
这篇 protocol 用的是第二条：AAV2/1-SynTetOff-EGFP-APEX2，滴度 1.0×10¹¹ 感染单位/mL，立体定位注射进 8–16 周 C57BL/6J 小鼠的纹状体，表达 4 周。EGFP 负责光镜下的定位，APEX2 负责电镜下的显影——同一个融合蛋白，两种成像模式，这也是它能做 CLEM 的原因。
反应原理值得说清楚，因为它决定了你后面怎么排查失败。APEX2 在过氧化氢（H2O2）存在时，会氧化生物素-酪胺（biotin-tyramide, BT）；氧化后的酪胺产物非常活泼，会共价结合到酶附近蛋白质的酪氨酸残基上。这就是酪胺信号放大（tyramide signal amplification, TSA）：一个酶分子可以催化沉积很多个生物素，信号被放大。接着用亲和素-生物素复合物（avidin-biotin complex, ABC）结合生物素，再用二氨基联苯胺（3,3'-diaminobenzidine, DAB）加镍离子显色。DAB 被氧化后聚合，产物嗜锇（osmiophilic），在锇酸处理后被染成深色，电镜下就是一团无定形的电子致密物。
三个操作上的坑，原文都写得很明白。第一，叠氮钠（NaN3）会灭活 APEX2，所以反应前所有溶液都不能含 NaN3，反应结束后才可以用 0.02% NaN3 保存。第二，反应必须在透明化之前完成，透明化之后酶活会明显下降。第三，标记密度不是越高越好：密度过高会让局部反应物耗竭，出现"感染细胞中心反而没有染色"；而沉淀太强，会把活性带、突触后致密物和囊泡糊住，这正是原文 Limitations 里承认的干扰。
判读时还要记住：DAB-Ni 沉淀标的是"酶曾经在哪里"，不是"这个蛋白此刻精确到纳米在哪里"。酪胺产物会向外扩散，所以它的空间分辨率不如免疫金；它适合回答"这类细胞投射到哪里、形态如何"，不适合回答"这个受体是不是正好在突触后膜上"。选标记之前先想清楚你要回答哪个问题，比选哪个抗体更重要。
五、透明化与共聚焦：先看全貌，再切薄片
1 mm 厚的脑片，普通共聚焦只能看到表面几十微米，深处的荧光被散射和吸收吃掉。透明化要解决的就是这件事：把脂质去掉、把折射率（refractive index, RI）调到和物镜匹配，让光和激光都能进去。这篇 protocol 用的 ScaleSF 分两步：ScaleS0 让组织先收缩，ScaleS4 让它在高折射率介质里重新膨胀；去脂之后，组织变得半透明，荧光蛋白保留下来。
参数照原文：ScaleS0 处理 2 h，37°C、90 rpm；PBS(–) 洗 2×15 min；ScaleS4 处理 8–12 h，37°C、90 rpm；全程 10.5–14.5 h。1 mm 厚的脑片在 14.5 h 内达到透明。温度低于 37°C 或摇动不足，透明会不完全；透明液反复使用或存放过久（原文建议 4°C 保存不超过 1 个月），透明效果和荧光都会下降。
装片是个容易被低估的环节。透明切片要贴在腔底盖玻片上，滴 37°C 的 ScaleS4 胶，盖上盖玻片，用 Kimwipe 从边缘吸掉多余胶，再压载玻片和金属重物，4°C 30 min 让胶固化；上机前还要在 ScaleS4 里平衡 60 min。胶层太厚，工作距离不够；胶层有气泡，成像会出现条纹。物镜用 16× 多浸物镜，NA 0.60，工作距离 2.5 mm；校正环必须调到 1.47，因为 ScaleS4 的折射率大约是 1.47。校正环不对，点扩散函数失配，三维重建会明显变糊——原文 Troubleshooting 把它单列一条，说明这是高频错误。
共聚焦看完，还要回到"薄"。透明脑片去掉盖玻片，4% PFA 再固定 16–20 h，30% 蔗糖泡到沉底，OCT 包埋，液氮冷却的异戊烷速冻，冷冻切片机切 40 µm。这 40 µm 的切片再上共聚焦（75% 甘油封片，25×/NA 0.95 水镜或 63×/NA 1.40 油镜），确认目标还在，然后才做 ABC/DAB-Ni 反应和电镜制样。顺序不能反：先做 DAB 再做共聚焦，深色沉淀会挡住荧光；先切薄再找目标，又失去了透明化的意义。
透明化的边界也要说清楚。ScaleSF 只能透明 1 mm 厚的切片，不是全脑；去脂用的去垢剂加上 37°C、10–14 h 的处理，会"轻度但显著"地损伤超微结构。所以它是"先看全貌、再定位"的工具，不是"保结构"的工具。真正决定电镜图质量的那段超微结构，还是要靠固定、锇酸和包埋守住。
六、判读：髓鞘、突触、线粒体、自噬体
拿到一张电镜图，第一件事不是找"想看的结构"，而是先问：这张图的固定、切片、染色过关吗？过关之后，再按结构逐类判读。下面四类是神经科学里最常被数的东西，每一类都有它"长什么样"和"哪里容易看错"。
（一）髓鞘
正常中枢神经系统的髓鞘，在横切面上是同心圆状的板层：电子致密的主致密线（major dense line）与较浅的内周期线（intraperiod line）交替排列，一个少突胶质细胞可以包绕多根轴突。判读时看三件事：板层是否连续、层数是多少、轴突与髓鞘的比例是否正常。层数不能凭"看起来厚"来估，要沿着一个方向数板层，并说明在哪个放大倍数下数的。
衡量髓鞘相对厚度，用g 比（g-ratio）：g 比 = 轴突直径 ÷ 纤维（轴突加髓鞘）外径。它是一个比值，没有单位。g 比越接近 1，髓鞘越薄；越小，髓鞘相对越厚。不同脑区、物种、年龄的正常范围不同，文献里常见的参考区间是 0.6–0.8（经验值，具体以本实验室同龄对照为准）。判读时要小心两件事：一是"g 比升高"既可能是脱髓鞘，也可能是轴突本身变细；二是再生髓鞘虽然薄（g 比高），但板层结构可能仍然完整，单看横切面分不清"发育未成熟"和"损伤后再生"，需要结合纵切面、节间长度或其他标记。
（二）突触
化学性突触在电镜下的"三件套"是：突触前终末，里面有一群约 40 nm 的透明囊泡；活性带（active zone），即囊泡停靠、释放处的致密带；突触间隙（synaptic cleft），宽约 20 nm。突触后膜下方若有一层电子致密的突触后致密物（postsynaptic density, PSD），就是非对称突触（Gray I 型），多为兴奋性；若 PSD 不明显、间隙较窄，就是对称突触（Gray II 型），多为抑制性。这篇 protocol 的纹状体投射就是对称突触：原文 Expected outcomes 明确写到，电镜下看到的是"无 PSD、间隙窄"的对称性突触。
计数是最容易出问题的地方。只报"一个视野里数到几个"没有意义，因为视野大小、切片厚度、抽样位置都会影响结果。规范做法是说清楚抽样框：动物数、每只动物几张切片、每张切片几个视野、视野如何随机选取；如果要推断面密度或体积密度，需要体视学（stereology）的无偏方法，比如用 disector 计数。带教老师问"你这个数是怎么来的"，比问"你数到几个"更重要。
（三）线粒体
线粒体的判读要点：外膜与内膜是否连续，嵴（cristae）是否清晰、是否肿胀或断裂，基质电子密度是否均匀。要注意，线粒体肿胀、空泡化既可能是真实的损伤，也可能是固定或脱水不足造成的伪影。区分办法是比较同一张切片里"看起来正常"的相邻结构：如果连细胞核、髓鞘也一起"烂"，那更可能是制样问题，而不是生物学表型。
（四）自噬体
自噬体（autophagosome）在电镜下是双层膜包绕的囊泡，直径大约 300–900 nm，内含胞质成分或受损细胞器。看到双层膜结构，可以说"存在自噬体样结构"；但不能仅凭电镜说"自噬流增强"——自噬体增多既可能是生成增加，也可能是降解受阻。要判断通量，得配合 LC3-II 的转换、p62 的降解、双荧光标记等实验。这是"结构≠功能"在自噬领域的典型例子。
（五）伪影识别
电镜图里的伪影有它的"形状语法"：周期性、边界锐利、与已知结构对不上、只出现在某些样本里。下面这张表是判读时的速查表。
| 伪影 | 图上长什么样 | 常见原因 | 怎么避免 |
|---|---|---|---|
| 冰晶损伤 | 胞质内成群的空洞或裂隙，边缘锐利 | 冷冻保护不足、异戊烷温度不够低 | 蔗糖泡到沉底、异戊烷接近凝固点、速冻 |
| 固定不良 | 膜断裂、细胞器肿胀、胞质"洗空" | 灌注不足、固定液穿透慢、后固定时间不够 | 充分灌注、新鲜固定液、4°C 16–20 h |
| 切片颤痕 | 周期性平行条纹，厚薄交替 | 刀钝、树脂块太硬、切片速度不当 | 换刀、修块、调整切片参数 |
| 染色沉淀 | 不规则黑色团块，与结构无对应 | 铅染后清洗不足、铅与空气中的 CO2 反应 | 充分清洗、避 CO2、现配现用 |
| 皱褶与折叠 | 条带状重叠、结构重复出现 | 捞片不当、切片漂浮 | 小心捞片、让切片自然展开 |
| 血管内源过氧化物酶 | 血管壁强染，与 DAB 信号混淆 | 灌注不充分，红细胞残留 | 冰冷 PBS 充分灌注 |
| APEX2 过度标记 | 胞质弥漫深色聚合物，遮住活性带与囊泡 | 标记密度过高、反应过强 | 降低 AAV 滴度或表达时间、调整 BT 浓度 |
最后一条经验：判读时先看低倍，确认组织没有整体性的制样问题；再看高倍，找目标结构；最后回到低倍，确认你拍的这个视野在整块组织里不是孤例。三步都走完，再下结论。
七、本科生能参与的环节
电镜不是一个人从头做到尾的实验。它更像一条流水线：取材、固定、切片、染色、观察，每一环都可以被认真对待，也可以被糊弄过去。下面这几件事，是本科生现在就能做、而且做得好就会显著提高整批数据质量的。
第一，取材与固定计时。从断头到固定液灌入的时间、从取脑到后固定开始的时间、固定液的体积与温度、后固定的时长与摇速——这些都要记。灌注不充分的脑，后面所有步骤都救不回来；而"灌没灌好"往往在几小时后才从图上表现出来，到那时已经无法重来。计时不是形式主义，它是唯一能让你在失败后回溯原因的线索。
第二，包埋块编号与台账。一块组织一个编号：动物号、脑区、左右、切片序号、ROI 坐标、树脂块号、铜网号。台账要能让别人从一张电镜图追回到原动物、原切片、原 ROI。这件事看起来琐碎，但它是"数据可追溯"的底线；电镜样本一旦编号混乱，后面所有的定量都失去意义。
第三，切片观察与拍照归档。明场下找 ROI，记录坐标；上电镜后，把加速电压、放大倍数、像素尺寸随图保存；原始文件不裁、不压、不覆盖；命名规则统一，比如"日期_动物号_脑区_切片号_视野号"。很多实验室的图找不到原始文件，不是因为没拍，是因为命名和归档没规则。
第四，测量与盲法。髓鞘厚度、g 比、突触计数，都需要先定标准：什么算一个突触、量膜的中线还是外缘、视野怎么选。标准定好之后再开始量，随机抽样，双人盲法（测量者不知道组别），用 ImageJ 之类的工具记录每一步。不要边看结果边改标准——那不是测量，是找结论。
第五，记录本。每一步的实际时间、试剂批号、异常（比如"今天固定液是上周配的""切片时听到异响"）。失败要写清楚，不然下次还会以同样的方式失败。电镜实验的周期长、成本高，记录本是唯一便宜的保险。
不能独立做的部分也要说清楚：超薄切片需要几个月的练习，最终判读需要对照和训练。但"不能独立做"不等于"只能旁观"——把取材计时、编号台账、拍照归档、测量盲法这四件事做到位，你就已经在参与真正的电镜研究了。
现在能亲手做到哪一步？把上面的内容落到一张任务清单上：① 能在带教老师监督下完成灌注取材，并准确记录从断头到固定液灌入、从取脑到后固定开始的每一个时间点，以及固定液体积、温度、后固定时长与摇速；② 能独立完成琼脂包埋、振动切片机的上样与参数记录（厚度、速度、振幅、频率）；③ 能按台账给树脂块、载玻片、铜网编号，并保证从一张电镜图追回到原动物、原脑区、原 ROI；④ 能在明场下找到 ROI、拍照存档，并把 TEM 的加速电压、放大倍数、像素尺寸随图保存；⑤ 能在 ImageJ 里按事先写好的标准测量髓鞘厚度与 g 比、按抽样框数突触，并接受双人盲法复核。还不能独立做的：超薄切片（需要长期练习）、免疫金标记的判读、以及把结构与功能、因果联系起来的解释。把"能做"的做到可追溯，把"不能做"的交给带教，这就是本科生在电镜流水线里的正确位置。
八、报告边界：一张图能说的与不能说的
一张电镜图要能被别人重复，报告里必须有哪些东西？按这篇 protocol 的要素，至少包括：固定液配方（PFA 与 GA 的浓度）与灌注方式；切片厚度与振动切片参数；透明化方法（ScaleS0/ScaleS4 的时间与温度）；APEX2/BT-GO 的浓度与孵育时间；DAB-Ni 显色时间；OsO4、醋酸铀、铅柠檬酸的处理时间；超薄切片厚度；透射电镜的加速电压与像素尺寸；动物数、切片数、视野数；抽样与盲法；统计方法。缺一项，别人就无法判断你的图是"结构如此"还是"制样如此"。
措辞上也要守住边界。不要写"电镜显示突触减少，说明记忆下降"，要写"在 X 条件下，Y 脑区每单位面积对称突触数从 A 降到 B（n=…，p=…）；该结构变化与行为学指标相关，因果关系需进一步实验验证"。前者把结构、功能、因果三件事混成一句，后者把它们分开——这不是啰嗦，是科学写作的基本诚实。
最后一句提醒：电镜最容易被浪费的用法，是把它当成"高级确认"。你先有了结论，再去找一张支持结论的图；找不到就换个视野，直到找到为止。这叫确认偏误（confirmation bias），不叫证据。真正有用的电镜数据，是在抽样之前就定好规则、在拍照之前就定好标准、在判读之前就接受"可能看不到差异"的那种。
▲ 本页翻译难点（对照学习用）
附录 · 英文原文
以下为 STAR Protocols 2022 年该方案英文全文的转录。中文译文按原文的 Before you begin（关键准备）、Step-by-step method details（组织取材与固定、APEX2/BT-GO 反应、透明化与共聚焦成像、再切片、DAB-Ni 反应、电镜制样、超薄切片）、Expected outcomes 与 Troubleshooting 的顺序重排并精译；原文中的图、图注、参考文献编号、试剂表与部分项目符号在文本抽取时未保留，个别步骤编号因此不连续。英文版权归原作者与期刊所有，此处仅作学习对照。
Protocol for multi-scale light microscopy/electron microscopy neuronal imaging in mouse brain tissue
Abstract
An imaging technique across multiple spatial scales is required for extracting structural information on neurons with processes of meter scale length and specialized nanoscale structures. Here, we present a protocol combining multi-scale light microscopy (LM) with electron microscopy (EM) in mouse brain tissue. We describe tissue slice preparation and LM/EM dual labeling with EGFP-APEX2 fusion protein. We then detail ScaleSF tissue clearing and successive LM/EM imaging. Our protocol allows for deciphering structural information across multiple spatial scales on neurons.
For complete details on the use and execution of this protocol, please refer to [Furuta et al. (2022)].
Before you begin
The protocol below describes the specific steps for using mouse brain tissues. However, we have also used this protocol in brain tissues of common marmoset (Callithrix jacchus) ([Furuta et al., 2022]).
Institutional permissions
All animal experiments in this protocol were approved by the Institutional Animal Care and Use Committees of Juntendo University (Approval No. 2020087, 2020088) and conducted in accordance with Fundamental Guidelines for Proper Conduct of Animal Experiments by the Science Council of Japan (2006). All experiments on live vertebrates or higher invertebrates must be performed in accordance with the Guideline for the Institutional Animal Care and Use Committee.
Stereotaxic injection of an adeno-associated virus (AAV) vector
Timing: ∼1 h
For a detailed procedure, refer to [Okamoto et al. (2021)] and [Takahashi et al. (2021)].1.Prepare AAV2/1-SynTetOff-EGFP-APEX2 vector solution (1.0 × 1011 infectious unit/mL).2.Load the AAV solution into a glass micropipette.3.Attach the micropipette to the holder of Picospritzer® III.4.Anesthetize a C57BL/6J mouse (8–16 week old) by intraperitoneal (ip) injection of a mixture of Domitor® (0.3 mg/kg), Dormicum® (4 mg/kg) and Vetorphale® (5 mg/kg).5.Mount the anesthetized animal onto a stereotaxic instrument.6.Thin the skull with a dental drill above the injection site, and remove the remaining bone with fine forceps.7.Insert the tip of the micropipette into the brain and eject the vector solution.8.Close and sterilize the wound, and administrate gentamicin locally.9.Recover the animal from anesthesia by ip injection of Antisedan® (1.5 mg/kg).10.Maintain the animal for 4 weeks under regular health check.
Preparation of a regent for tissue fixation
Timing: ∼1 h 11.16% (w/v) paraformaldehyde (PFA) in distilled deionized water (ddH2O) (100 mL).a.Warm ∼80 mL ddH2O to 70°C–80°C.b.Add 16 g PFA to the warmed water.c.Dissolve PFA by adding 2 to 3 drops of 10 N sodium hydroxide (NaOH) solution.d.Allow the solution to cool to 20°C–25°C, and bring 100 mL with ddH2O.e.Filter the solution through filter paper.f.Dispense to 10 mL each.g.Store at −20°C. The solution can be stored at −20°C for 6 months. CRITICAL: PFA is toxic and teratogenic. Avoid inhalation or contact with skin, eyes and mucous membrane. Handle it inside a fume hood with appropriate protective gear. CRITICAL: NaOH is corrosive. Handle it with appropriate protective gear. Contact with skin, eyes and mucous membrane must be avoided.
Preparation of an imaging chamber
Timing: ∼3 h
A customizable three dimensional (3D)-printed imaging chamber is designed for reliable mounting of optically cleared brain slices () ([Furuta et al., 2022]). The chamber consists of the chamber frame and bottom coverslip. Microscope stage adaptors are also designed for direct mounting of the imaging chamber on a microscope stage. 3D computer-aided design (CAD) data for the chamber frame and microscope stage adaptors are provided in [Furuta et al. (2022)].12.Prepare the chamber frame. The chamber frame is 3D-printed.Alternatives: Online 3D printing services (e.g., DMM.make [https://make.dmm.com]) provide a good alternative.13.Secure the printed chamber frame to the bottom coverslip with a pressure-sensitive adhesive.Figure 1A customizable 3D-printed imaging chamber for optically cleared tissue slices(A and B) A schema drawing (A) and picture (B) of a customizable 3D-printed imaging chamber. The imaging chamber is composed of the chamber frame and bottom coverslip. Microscope stage adaptors are designed for mounting directly on a microscope stage. ScaleSF-treated tissue slices are mounted onto the bottom coverslip and embedded in ScaleS4 gel. The chamber frame and microscope stage adaptors are customizable.Reprinted and modified from [Furuta et al. (2022)] under the Creative Commons Attribution 4.0 International License (CC BY 4.0; https://creativecommons.org/licenses/by/4.0/).
Preparation of silicone-coated glass slides and coverslips
Timing: 1 day 14.Add 500 μL of Siliconized L-25 to 9.5 mL of chloroform, and mix well by stirring.15.Spread the 5% Siliconized L-25 solution on glass slides and coverslips.16.Let them dry. CRITICAL: Chloroform is toxic and carcinogenic. Handle it inside a fume hood with appropriate protective gear. All necessary precautions should be taken when disposing.
Preparation of rocket of resin
Timing: 3 days 17.Mix well 3.8 mL of Luveak-812, 2 mL of Luveak-DDSA and 2.4 mL of Luveak-MNA by slowly stirring for 5 min.18.Add 0.12 mL of Luveak-DMP-30 to the mixture, and mix well by slowly stirring for 30 min.19.Put the mixture into silicon capsules and polymerize the resin in an oven at 60°C for 2 days. The polymerized resin, rocket of resin, can be stored at 20°C–25°C for several years.
Key resources table
REAGENT or RESOURCESOURCEIDENTIFIERBacterial and virus strainsAAV2/1-SynTetOff-EGFP-APEX2[Furuta et al. (2022)]n/aChemicals, peptides, and recombinant proteins1 N sodium hydroxide (NaOH) solutionNacalai Tesque37421-052-AminoethanolNacalai Tesque23405-553,3′-Diaminobenzidine, tetrahydrochloride (DAB)Dojindo347-0090410 N NaOH solutionNacalai Tesque94611-45AgarNacalai Tesque01028-85AgaroseTaKaRa BioL03Ammonium Nickel (II) Sulfate Hexahydrate (Nickel ammonium sulfate)Nacalai Tesque24217-82Atipamezole (Antisedan®)Nippon Zenyaku Kogyon/aß-D-glucoseWako Pure Chemical Industries049-31165Biotin–NHSCalbiochem203112Bovine serum albumin (BSA)Nacalai Tesque01863-77Butorphanol tartrate (Vetorphale®)Meiji Seika Pharman/aChloroformNacalai Tesque08402-55Dimethyl sulfoxide (DMSO)Nacalai Tesque13407-45Di-sodium hydrogen phosphate 12-water (Na2HPO4·12H2O)Nacalai Tesque31722-45D-SorbitolNacalai Tesque06286-55Ethanol (EtOH) (99.5%)Nacalai Tesque14713-95γ-cyclodextrinWako Pure Chemical Industries037-10643GentamicinNichi-Iko Pharmaceuticaln/aGlucose oxidaseNacalai Tesque16831-14Glutaraldehyde (GA) (25% aqueous solution)Nacalai Tesque17003-92GlycerolSigma-AldrichG9012Hydrochloric acid (HCl)Nacalai Tesque18321-05Hydrogen peroxide (H2O2) (31% aqueous solution)Santokun/aLead nitrateNacalai Tesque20231-02Luveak-812Nacalai Tesque20829-05Luveak-DDSANacalai Tesque14423-95Luveak-DMP-30Nacalai Tesque14425-75Luveak-MNANacalai Tesque14424-85Medetomidine hydrochloride (Domitor®)Nippon Zenyaku Kogyon/aMidazolam (Dormicum®)Astellas Pharman/aMethyl-β-cyclodextrinTokyo Chemical IndustryM1356Molecular Sieves 3A, mixed indicatorNacalai Tesque23356-05Osmium (Ⅷ) Oxide (OsO4) (2% aqueous solution)Nacalai Tesque25746-06Paraformaldehyde (PFA)Merck Millipore1.04005.1000Phosphate buffered saline (10×) (pH 7.4) (10× PBS(–))Nacalai Tesque27575-31Potassium chloride (KCl)Nacalai Tesque28514-75Potassium dihydrogenphosphate (KH2PO4)Nacalai Tesque28721-55Propylene oxideNacalai Tesque29223-55Siliconized L-25Fuji Systems0411002Sodium azide (NaN3)Nacalai Tesque31233-55Sodium dihydrogen phosphate dihydrate (NaH2PO4·2H2O)Nacalai Tesque31718-15Sodium chloride (NaCl)Nacalai Tesque31320-05Sodium pentobarbital (Somnopentyl®)Kyoritsu Seiyakun/aSodium citrateNacalai Tesque31404-15SucroseNacalai Tesque30404-45Tissue-Tek® O.C.T. Compound (OCT compound)Sakura Finetek4583Tris(hydroxymethyl)aminomethane (Tris)Nacalai Tesque35434-21Triton X-100Nacalai Tesque35501-15Tyramine hydrochlorideSigma-AldrichT2879-1GUreaNacalai Tesque35940-65Uranyl acetateMerckArt. 8473Critical commercial assaysVECTASTAIN Elite ABC KitVectorPK-6100Deposited dataCAD data of chamber frame for microscopic image acquisition of thick slices[Furuta et al. (2022)]Data S1CAD data of stage adaptor for microscopic image acquisition of thick slices[Furuta et al. (2022)]Data S2Experimental models: Organisms/strainsMouse: C57BL/6J (8–16 weeks old, male and female)Nihon SLCC57BL/6JJmsSlcSoftware and algorithmsImageJ[Schneider et al. (2012)]https://imagej.nih.gov/ij/ImarisBitplanehttps://imaris.oxinst.jp/imaris-viewerLeica Application Suite XLeica Microsystemshttps://www.leica-microsystems.com/products/microscope-software/p/leica-las-x-ls/Other0.22 μm filterMerck MilliporeSLGP033RB13 mL centrifuge tubeSARSTEDT60.541.54516× multi-immersion objective lensLeica MicrosystemsHC FLUOTAR 16×/0.60 IMM CORR VISIR25× water-immersion objective lensLeica MicrosystemsHC FLUOTAR L 25×/0.95 W VISIR3D printing serviceDMM.makehttps://make.dmm.com3D-printerKeyenceAGILISTA-320040× objective lensOlympusUPlanApo ×40/0.8563× oil-immersion objective lensLeica MicrosystemsHC PL APO 63×/1.40 Oil CS26-well culture plateGreiner Bio-One657160Blu-Tack®Bostikn/aBrightfield microscopeOlympusBX-51Carbon steel bladesFeatherFA-10BCoverslipMatsunami GlassC025601CryomoldSakura Finetek4566Diamond knifeDiatomeUltra 45° 3.0 mmDigital Microscope Color CameraOlympusDP74Electro freezeYamato KokiMC-802AGlass capillary (diameter 2 mm)NarishigeG-2Glass slideMatsunami GlassS1225Low temperature circulator bathEYELACCA-1112AParafilm®BemisPM-996Picospritzer IIIParker Hannifinn/aPressure-sensitive adhesiveCEMEDINENA-007Rigid acrylic resinKeyenceAR-M2Sliding microtomeLeica BiosystemsSM2000RSnow hornNippon EkitanCorporationn/aStereo microscopeNikonSMZ445Superfrost APS-coated micro slide glassMatsunami GlassAPS-01Stereotaxic apparatusNarishigeSR50Synthetic diamond knifeDiatomeHisto 45° 4.0 mmTCS SP8Leica Microsystemsn/aTEM gridNisshin EMVECO GRID 150 mesh CuTransmission electron microscopeHitachiH-7650UltramicrotomeLeica MicrosystemsUltracut UCTVibratomeDosaka EMLinear PRO7N
Materials and equipment
Phosphate buffer (PB) (pH 7.4) 0.2 MReagentFinal concentrationAmountNaH2PO4·2H2O38 mM11.8 gNa2HPO4·12H2O162 mM116 gddH2On/aup to 2 LTotaln/a2 L Phosphate buffered saline (PBS) (pH 7.4) 10×ReagentFinal concentrationAmountNaCl1.37 M80 gNa2HPO4·12H2O81 mM29 gKCl27 mM2 gKH2PO415 mM2 gddH2On/aup to 1 LTotaln/a1 L Fixative solutionReagentFinal concentrationAmountPFA 16%4% (w/v)10 mLGA 25%0.2% (v/v)320 μLPB (pH 7.4) 0.2 M0.1 M20 mLddH2On/aup to 40 mLTotaln/a40 mL CRITICAL: PFA is toxic and teratogenic. Avoid inhalation or contact with skin, eyes and mucous membrane. Handle it inside a fume hood with appropriate protective gear. CRITICAL: GA is a toxic and strong irritant, and should be handled inside a fume hood with appropriate protective gear. Contact with skin, eyes and mucous membrane must be avoided. Permeabilization bufferReagentFinal concentrationAmountBSA2% (w/v)1 gTriton X-1000.2% (v/v)100 μLPBS (pH 7.4) 10×1×5 mLddH2On/aup to 50 mLTotaln/a50 mL 2% BSA in 0.1 M PBReagentFinal concentrationAmountBSA2% (w/v)1 gPB (pH 7.4) 0.2 M0.1 M25 mLddH2On/aup to 50 mLTotaln/a50 mL Glucose oxidase (GO) solutionReagentFinal concentrationAmountGlucose oxidase1 mg/mL1 mgPB (pH 7.4) 0.2 M0.1 M500 μLddH2On/a500 μLTotaln/a1 mL Biotin-NHS solutionReagentFinal concentrationAmountBiotin-NHS95.9 mg/mL3.5 mgDMSOn/a36.5 μLTotaln/a36.5 μL Tyramine hydrochloride solutionReagentFinal concentrationAmountTyramine hydrochloride50 mg/mL15 mgDMSOn/a300 μLTotaln/a300 μL Biotinylated tyramine (BT) solutionReagentFinal concentrationAmountBiotin-NHS solutionn/a36.5 μLTyramine hydrochloride solutionn/a36.5 μL2-Aminoethanoln/a7.3 μLTotaln/a80.3 μL Biotinylated tyramine-glucose oxidase (BT-GO) reaction mixtureReagentFinal concentrationAmountBT solution1:5,0001 μLGO solution3 μg/mL15 μL2% BSA in 0.1 M PBn/a5 mLTotaln/a5 mL CRITICAL: The optimal concentration of BT solution should be determined for each experiment. We usually dilute the solution from 1:5,000 to 1:50,000 for the total. ß-D-glucose solutionReagentFinal concentrationAmountß-D-glucose200 mg/mL200 mgddH2On/a1 mLTotaln/a1 mL 4% PFA in 0.1 M PBReagentFinal concentrationAmountPFA 16%4% (w/v)10 mLPB (pH 7.4) 0.2 M0.1 M20 mLddH2On/a10 mLTotaln/a40 mL CRITICAL: PFA is toxic and teratogenic. Avoid inhalation or contact with skin, eyes and mucous membrane. Handle it inside a fume hood with appropriate protective gear. Methyl-β-cyclodextrin 100 mMReagentFinal concentrationAmountMethyl-β-cyclodextrin100 mM1.303 gddH2On/a10 mLTotaln/a10 mL γ-cyclodextrin 100 mMReagentFinal concentrationAmountγ-cyclodextrin100 mM1.297 gddH2On/a10 mLTotaln/a10 mL Triton X-100 10%ReagentFinal concentrationAmountTriton X-10010% (w/v)5 gddH2On/aup to 50 mLTotaln/a50 mL ScaleS0 solutionReagentFinal concentrationAmountD-sorbitol20% (w/v)20 gGlycerol5% (w/v)5 gMethyl-β-cyclodextrin 100 mM1 mM1 mLγ-cyclodextrin 100 mM1 mM1 mLDMSO3% (v/v)3 mL10× PBS(–)1×10 mLddH2On/aup to 100 mLTotaln/a100 mL Note: The original ScaleS0 solution contains N-acetyl-L-hydroxyproline (Skin Essential Actives, Hualian, Republic of China) ([Hama et al., 2015]). If this amino acid derivative is difficult to obtain from the manufacturer, N-acetyl-L-hydroxyproline should be omitted from the solution. N-acetyl-L-hydroxyproline provided by other manufactures can negatively affect tissue clarification ([Miyawaki et al., 2016]). ScaleS4 solutionReagentFinal concentrationAmountUrea4 M24.02 gD-sorbitol40% (w/v)40 gGlycerol10% (w/v)10 gTriton X-100 10%0.2% (w/v)2 mLDMSO25% (v/v)25 mLddH2On/aup to 100 mLTotaln/a100 mL ScaleS4 D25(0) solutionReagentFinal concentrationAmountUrea4 M24.02 gD-sorbitol40% (w/v)40 gGlycerol10% (w/v)10 gDMSO25% (v/v)25 mLddH2On/aup to 100 mLTotaln/a100 mL ScaleS4 gelReagentFinal concentrationAmountAgarose1.5% (w/v)1.5 gScaleS4D25(0)100 mL100 mLTotaln/a100 mL 30% Sucrose in 0.1 M PBReagentFinal concentrationAmountSucrose30% (w/v)30 g0.2 M PB (pH 7.4)0.1 M50 mLddH2On/aup to 100 mLTotaln/a100 mL 75% Glycerol in PBSReagentFinal concentrationAmountGlycerol75% (v/v)7.5 mLPBS25% (v/v)2.5 mLTotaln/a10 mL 2% BSA in PBSReagentFinal concentrationAmountBSA2% (w/v)1 gPBS (pH 7.4) 10×1×5 mLddH2On/aup to 50 mLTotaln/a50 mL ABC solutionReagentFinal concentrationAmountReagent A1:5020 μLReagent B1:5020 μL2% BSA in PBSn/a1 mLTotaln/a1 mL CRITICAL: ABC solution should be prepared before incubation with brain sections. Add 20 μL of Reagent A to 1 mL of 2% BSA in PBS, and then add 20 μL of Reagent B to the same solution, and mix immediately. Incubate the mixture for 30 min at 20°C–25°C with gentle agitation, and chill on ice. Tris-HCl (pH 7.6) 500 mMReagentFinal concentrationAmountTris500 mM60.57 gddH2On/aup to 1 LTotaln/a1 L Nickel ammonium sulfate 200 mMReagentFinal concentrationAmountNickel ammonium sulfate200 mM395 mgddH2On/a5 mLTotaln/a5 mL CRITICAL: Nickel ammonium sulfate is toxic and carcinogenic. Avoid inhalation or contact with skin, eyes and mucous membrane. Handle it inside a fume hood with appropriate protective gear. DAB-Ni2+ solutionReagentFinal concentrationAmountDAB0.05% (w/v)10 mgNickel ammonium sulfate 200 mM2.5 mM250 μLTris-HCl (pH 7.6) 500 mM50 mM2 mLddH2On/aup to 20 mLTotaln/a20 mL CRITICAL: DAB is carcinogenic. Waste should be oxidized before being discarded. Avoid inhalation or contact with skin, eyes and mucous membrane. Handle it inside a fume hood with appropriate protective gear. CRITICAL: Nickel ammonium sulfate is toxic and carcinogenic. Avoid inhalation or contact with skin, eyes and mucous membrane. Handle it inside a fume hood with appropriate protective gear. 1% PFA in 0.1 M PBReagentFinal concentrationAmountPFA 16%1% (w/v)2.5 mLPB (pH 7.4) 0.2 M0.1 M20 mLddH2On/aup to 40 mLTotaln/a40 mL CRITICAL: PFA is toxic and teratogenic. Avoid inhalation or contact with skin, eyes and mucous membrane. Handle it inside a fume hood with appropriate protective gear. 1% OsO4 in 0.1 M PBReagentFinal concentrationAmountOsO4 2%1% (w/v)5 mLPB (pH 7.4) 0.2 M0.1 M5 mLTotaln/a10 mL CRITICAL: OsO4 is highly volatile and reactive. Avoid inhalation or contact with skin, eyes and mucous membrane. Handle it inside a fume hood with appropriate protective gear. 2% uranyl acetate (UA) in 50% EtOHReagentFinal concentrationAmountUranyl acetate2% (w/v)20 mgEtOH (99.5%)50% (v/v)0.5 mLDDWn/a0.5 mLTotaln/a1 mL CRITICAL: Uranyl acetate is a radioactive material and all necessary precautions should be taken when disposing. Handle it inside a fume hood with appropriate protective gear. 1% UA solutionReagentFinal concentrationAmountUranyl acetate1% (w/v)10 mgDDWn/a1 mLTotaln/a1 mL CRITICAL: Uranyl acetate is a radioactive material and all necessary precautions should be taken when disposing. Handle it inside a fume hood with appropriate protective gear. Lead citrate solutionReagentFinal concentrationAmountLead nitrate2.66% (w/v)1.33 gSodium citrate3.52% (w/v)1.76 g1 N NaOH solution0.16 M8 mLDDWn/aup to 50 mLTotaln/a50 mL CRITICAL: To avoid production of lead carbonate, minimize contact of lead citrate solution with air. CRITICAL: NaOH is corrosive. Handle it with appropriate protective gear. Contact with skin, eyes and mucous membrane must be avoided. 100% EtOHReagentFinal concentrationAmountEtOH (99.5%)n/a100 mLMolecular Sieves 3An/a5 gTotaln/a100 mL CRITICAL: Molecular sieves should be replaced when the beads color changes to beige demonstrating that the molecular sieves have reached equilibrium capacity. Epon 812 mixtureReagentFinal concentrationAmountLuveak-81245.7% (v/v)3.8 mLLuveak-DDSA24.0% (v/v)2 mLLuveak-MNA28.8% (v/v)2.4 mLLuveak-DMP-301.4% (v/v)0.12 mLTotaln/a8.32 mL Epoxy resin/propylene oxide mixtureReagentFinal concentrationAmountEpon 812 mixture50% (v/v)5 mLPropylene oxide50% (v/v)5 mLTotaln/a10 mL
Step-by-step method details
Tissue slice preparation
Timing: 2 days
This step describes how to prepare mouse brain slices for the successive LM/EM imaging with ScaleSF tissue clearing.1.Sacrifice mice injected with the AAV2/1-SynTetOff-EGFP-APEX2 vector by ip injection of overdose of Somnopentyl® (200 mg/kg).CRITICAL: Perform steps 2 through 3 in a fume hood to limit the exposure to PFA and GA.2.Perfusion fixation.a.Open the thoracic cavity with surgical scissors.b.Make an incision to the right atrial appendage with surgical scissors.c.Insert a 22-gauge needle into the left ventricle of heart.d.Perfuse with 20 mL of ice-cold PBS using a 20 mL syringe.e.Perfuse with 20 mL of the ice-cold fixative solution with another 20 mL syringe.3.Excise brains using surgical scissors and tweezers.CRITICAL: From steps 4 to 5, samples should be protected from light.4.Immerse the brain in the fixative solution and rock gently for 16 h–20 h at 4°C (50 rpm–100 rpm).CRITICAL: Samples should not be stored for extended periods of time. Procced to the next step immediately.5.Brain slice preparation.a.Embed the brain tissues in 4% agar in PBS in a standard 6-well culture plate.Note: The temperature of the agar solution should be between 40°C and 45°C.Note: The viscous property by agaropectin, a major component of agar, improves ease of cutting.b.Allow the agar to solidify on ice for 30 min.c.Fill the vibratome bath with crashed ice.d.Stick a piece of mending tape on the bottom of the specimen tray.e.Attach the tray to the bath.f.Trim the agar block with a razor blade to square off the sides.g.Mount the agar block onto the tape with superglue.h.Pour ice-cold 0.1 M PB to the specimen tray.i.Set the sectioning speed to 0.14 mm/s with 1.4 mm amplitude.j.Set the vibratome frequency to 75 Hz–77 Hz.k.Cut 1-mm-thick brain slices.l.Collect the slices in ice-cold PBS.Pause point: The slices can be stored in PBS for 16 h–20 h at 4°C.CRITICAL: NaN3 as a preservative should be avoided. NaN3 inactivates the enzymatic activity of APEX2.
APEX2/BT-GO reaction
Timing: 2 days
In this step, we explain the procedure for APEX2/BT-GO reaction, where biotin molecules are deposited with tyramide signal amplification (TSA) reaction using the peroxidase activity of APEX2 () ([Furuta et al., 2022]). This step is essential for LM/EM dual labeling resistance for ScaleSF tissue clearing. All incubations are performed in a standard 6-well culture plate at 20°C–25°C, except where noted.CRITICAL: APEX2/BT-GO should be performed prior to tissue clearing with ScaleSF. Peroxidase activity of APEX2 markedly declines following ScaleSF treatment ().CRITICAL: Samples should be protected from light.6.Permeabilization.a.Immerse the slices in 10 mL of permeabilization buffer in a 13 mL centrifuge tube.b.Rock gently for 4 h (50 rpm–100 rpm).CRITICAL: Brain tissues should be shaken thoroughly.7.Wash.a.Transfer the slices in 8 mL of 0.1 M PB.b.Wash the slices with 8 mL of 0.1 M PB for 5 min three times with gentle rocking (50 rpm–100 rpm).8.BT penetration.a.Transfer the slices in 1200 μL of BT-GO reaction mixture in a 2 mL safe-lock microcentrifuge tube.b.Rock gently for 4 h (50 rpm–100 rpm).CRITICAL: Brain tissues should be shaken thoroughly.9.BT deposition.a.Add 12 μL of ß-D-glucose solution to the BT-GO reaction mixture.b.Rock gently for 2 h (50 rpm–100 rpm).CRITICAL: Brain tissues should be shaken thoroughly.10.Wash.a.Transfer the slices in 8 mL of PBS.b.Wash the slices with 8 mL of PBS for 15 min twice with gentle rocking (50 rpm–100 rpm).11.Fixation.a.Immerse the slices in 8 mL of 4% PFA in 0.1 M PB.b.Rock gently for 16 h–20 h at 4°C (50 rpm–100 rpm).12.Wash.a.Transfer the slices in 8 mL of PBS.b.Wash the slices with 8 mL of PBS for 5 min twice with gentle shaking (50 rpm–100 rpm).Pause point: The slices can be stored in 0.02% NaN3 in PBS at 4°C for 1 week.Figure 2Schematic diagram of APEX2/BT-GO reactionAPEX2/BT-GO is a TSA system that utilizes peroxidase activity of APEX2 and H2O2 produced during oxidation of glucose by glucose oxidase (GO reaction) to deposit biotinylated tyramine (BT) onto tissues. APEX2 reacts with the H2O2 and oxidizes the phenolic part of BT to produce highly reactive intermediates, which in turn covalently bind to electron-rich moieties such as tyrosine residues at or near the APEX2. POD: peroxidase.
Tissue clearing
Timing: 10.5–14.5 h
This step describes the procedure for ScaleSF tissue clearing. The clearing schedule is shown in A. All reactions are performed in a standard 6-well culture plate.CRITICAL: Samples should be protected from light.13.ScaleSF tissue clearing (B).a.Warm ScaleS0 and ScaleS4 solutions to 37°C.b.Immerse the brain slices in 8 mL of ScaleS0 solution, and rock gently for 2 h at 37°C (90 rpm).c.Transfer the slices in 8 mL of 1× PBS (–), and wash twice for 15 min with gentle rocking at 20°C–25°C (50 rpm–100 rpm).d.Transfer the slices in 8 mL of ScaleS4 solution, and rock gently for 8 h–12 h at 37°C (90 rpm).Figure 3ScaleSF tissue clearing in a mouse brain slice of 1-mm thickness(A) The schedule for ScaleSF tissue clearing.(B) Transmission images of a mouse brain slice during ScaleSF tissue clearing. Images are acquired before and after ScaleS0 treatment, after washes with PBS(–), and after an incubation in ScaleS4 solution. The grid interval is 1 mm. Scale bar, 2 mm.
Brain slices mounting
Timing: ∼2 h
This step describes how to mount the cleared brain slices on the imaging chamber.14.Brain slice mounting.a.Place the cleared brain slices on the bottom coverslip of the imaging chamber.b.Remove excessive ScaleS4 solution around the cleared slice.15.ScaleS4 gel embedding.a.Drop ScaleS4 gel on the cleared slice to fill the imaging chamber.Note: The temperature of the ScaleS4 gel should be kept at 37°C.b.Mount a coverslip on the imaging chamber, and place a piece of Kimwipe paper and a slide glass on the coverslip in this order.c.Transfer the imaging chamber in a refrigerator at 4°C.d.Place metal weights on the slide glass, and allow the gel to solidify for 30 min.e.Remove the metal weights, slide glass, Kimwipe paper and coverslip from the imaging chamber, and wipe away excess ScaleS4 gel.16.Tissue equilibration.a.Attach the imaging chamber to the bottom of a 60-mm glass petri dish with Blu-Tack®. Adhesion at multiple points is required.b.Pour ScaleS4 solution to the dish, and incubate the embedded slice in the solution for 60 min at 20°C–25°C with gentle agitation (40 rpm–60 rpm).c.Substitute with fresh ScaleS4 solution.CRITICAL: Air bubbles on the sample surface should be removed.Optional: Using microscope stage adaptors, the imaging chambers can be mount on a microscope stage directly. In this case, the coverslip should remain attached on the imaging chamber.
Confocal laser scanning microscopy (CLSM) imaging in cleared tissues
Timing: 2–24 h 17.Set the correction collar of a 16× multi-immersion objective lens (numerical aperture (NA) = 0.60, working distance [WD] = 2.5 mm) to 1.47. CRITICAL: Refractive index (RI) mismatch-induced aberrations can highly degrade the image formation. ScaleS4 solution has a RI of around 1.47 ([Hama et al., 2015]; [Miyawaki et al., 2016]). 18.Mount the imaging chamber on a microscope stage.19.Immerse the objective lens in ScaleS4 solution, and make it approach to the cleared slice slowly. CRITICAL: Air bubbles trapped on the tip of the objective lens should be removed. 20.Adjust imaging acquisition settings such as laser power, scan speed, pinhole diameter, detector gain, amplifier offset/gain, xy and z-axis resolution and bit intensity resolution.21.Collect images by a CLSM, and record captured images.22.Image processing is done with Leica Application Suite X and ImageJ.
Re-sectioning
Timing: 2 days
Here, we explain the procedure for re-sectioning from the slices imaged with CLSM. All reactions are performed in a standard 6-well culture plate at 20°C–25°C, except where noted.23.DeScale.a.Remove ScaleS4 gel around the imaged brain slices.b.Immerse the slice in 8 mL of 1× PBS (–), and wash twice for 15 min with gentle rocking at 20°C–25°C (40 rpm–60 rpm).24.Fixation.a.Immerse the slices in 8 mL of 4% PFA in 0.1 M PB, and rock gently for 16 h–20 h at 4°C (50 rpm–100 rpm).25.Wash.a.Transfer the slices in 8 mL of 1× PBS (–), and wash the slices with 8 mL of 1× PBS (–) for 5 min twice with gentle shaking (40 rpm–60 rpm).Pause point: The slices can be stored in 0.02% NaN3 in PBS at 4°C for 1 week.26.Cryoprotection.a.Immerse the slices in 8 mL of 30% sucrose in 0.1 M PB at 4°C until sunk (a sign of complete immersion).CRITICAL: Permeate brain slices with the cryoprotectant agent adequately. Adequate cryoprotection is of importance to reduce ice crystal formation and preserve cellular structure.Pause point: The slices can be stored in 30% sucrose in 0.1 M PB for 16 h–20 h at 4°C.27.Freezing brain slices.a.Immerse the slices in OCT compound for 15 min.b.Place the slice in a cryomold containing OCT compound.c.Cool isopentane in a metal container on liquid nitrogen.CRITICAL: Cool isopentane almost to its freezing temperature. Rapid freezing is of importance to reduce ice crystal formation and preserve cellular structure.d.Immerse the cryomold in the cooled isopentane to freeze.e.Transfer the cryomold in a freezer at −20°C.Pause point: The blocks can be stored at −20°C for a week.28.Cryosectioning.a.Mount the block onto the stage of a freezing microtome with OCT compound.b.Quickly freeze the OCT compound using dry ice powder and leave the block in place for 5 min.c.Cut the slice into 40-μm-thick sections and collect in ice-cold 0.1 M PB.Pause point: The sections can be stored in 0.02% NaN3 in PBS at 4°C for 1 week.Optional: Imaged slices can be embedded in agar or gelatin solution, and re-sectioned with a vibratome.
CLSM imaging in re-sections
Timing: 1 day
Here, we explain the procedure for CLSM imaging in re-sections with a high NA objective lens, which allows for high-resolution imaging of targeted structures.CRITICAL: Be sure to keep re-sections wet during this step.29.Section mounting.a.Place a re-section on a slide with a brush.b.Apply 75% glycerol in PBS to the re-section.c.Lower a coverslip onto 75% glycerol in PBS to avoid trapping any air bubbles.30.CLSM imaging.a.Mount the slide on a microscope stage.b.Adjust imaging acquisition settings such as laser power, scan speed, pinhole diameter, detector gain, amplifier offset/gain, xy and z-axis resolution and bit intensity resolution.d.Collect images by a CLSM equipped with high NA objective lenses, such as a 25×/NA 0.95 water-immersion and 63×/NA 1.40 oil-immersion objective lenses.e.Record captured images.f.Image processing is done with Leica Application Suite X and ImageJ.31.Wash.a.Remove the coverslip from the slide.b.Immerse the re-section in 8 mL of PBS in a standard 6-well culture plate, and wash twice for 15 min with gentle rocking at 20°C–25°C (40 rpm–60 rpm).Pause point: The sections can be stored in 0.02% NaN3 in PBS at 4°C for 1 week.
ABC/DAB-Ni2+ reaction
Timing: 2 days
This step describes the procedure for ABC/DAB-Ni2+ reaction. All incubations are performed in a standard 12-well culture plate at 4°C, except where noted.32.Wash.a.Immerse the sections in 3 mL of PBS, and wash twice for 10 min with gentle rocking (50 rpm–100 rpm).33.ABC reaction.a.Immerse the sections in 500 μL of ABC solution in a standard 24-well plate, and rock gently for 24 h (50 rpm–100 rpm).34.Wash.a.Immerse the sections in 3 mL of PBS, and wash twice for 10 min with gentle rocking (50 rpm–100 rpm).35.DAB-Ni2+ reaction.a.Immerse the sections in 1 mL of DAB-Ni2+ solution in a standard 24-well plate on ice, and incubate for 10 min on ice with gentle agitation (40 rpm–60 rpm).b.Add 3 μL of 0.1% H2O2 into DAB-Ni2+ solution, and incubate on ice with gentle agitation until color development (40 rpm–60 rpm).c.Immerse the sections in 3 mL of 2% NaN3 in PBS, and incubate for 15 min at 20°C–25°C with gentle agitation (40 rpm–60 rpm).36.Wash.a.Immerse the sections in 3 mL of PBS at 4°C and wash twice for 10 min with gentle rocking (50 rpm–100 rpm).Pause point: The sections can be stored in 1% PFA in 0.1 M PB at 4°C for 1 month.
Processing sections for EM
Timing: 3 days
This step describes processing tissue sections for ultra-thin sectioning and EM observation. This includes osmification, uranyl acetate staining and epoxy-resin embedding. All incubations are performed at 20°C–25°C, except where noted.37.Wash.a.Immerse the sections in 3 mL of 0.1 M PB, and wash twice for 10 min.38.Osmification.a.Immerse the sections in 1 mL of 1% OsO4 in 0.1 M PB, and incubate for 2 h in dark.CRITICAL: The reaction should be protected from light.39.Wash.a.Immerse the sections in 3 mL of 0.1 M PB, and wash three times for 10 min.b.Immerse the sections in 10 mL of ddH2O, and wash five times for 2 min with gentle rocking (50 rpm–100 rpm).40.Block staining.a.Immerse the sections in 1 mL of 2% UA in 50% EtOH, and incubate for 30 min in dark.CRITICAL: The reaction should be protected from light.41.Dehydration.a.Incubate the sections in 5 mL of increasing concentrations of ethanol in ddH2O for 20 min in the following order: 50%, 70%, 90%, 99.5%, 99.5%, 100%, and 100%. Incubations are performed in a glass vial.b.Incubate the sections in 10 mL of propylene oxide for 20 min twice.42.Epoxy resin embedding.a.Immerse the sections in 5 mL of Epoxy resin/propylene oxide mixture and incubate for 1 h.b.Immerse the sections in Epoxy resin and incubate for 16 h–20 h.c.Place the sections onto silicone-coated slide glass, and mount a silicone-coated coverslip on the slide.d.Place metal weights on the coverslip to make the sections flat.e.Remove the resin oozing from the slide.f.Polymerize the resin by heating in an oven at 60°C for 2 days.Pause point: The flat-embedded sections can be stored at 20°C–25°C in a desiccator for several months.
Ultra-thin sectioning
Timing: 1 day
Here, we describe how to prepare ultra-thin sections.43.Remove the coverslip from the slide with a razor blade. The embedded sections stick to either the coverslip or the slide.44.Identify the region of interest (ROI) with a brightfield microscope.45.Trim the ROI using a razor blade under a stereo microscope.46.Mount the trimmed resin block on a flat-tipped rocket of the resin with a glue.47.Trim away excess resin with a razor blade.48.Mount the rocket on the stage of an ultramicrotome, and align the block face with a synthetic diamond knife.49.Remove the empty resin from the top of the sample using the synthetic diamond knife to expose the surface of the tissue.50.Make ultrathin sections (60–80-nm thickness) with a diamond knife.51.Collect short ribbons of ultrathin sections on TEM grids of copper, and let them dry.52.Drop 1% UA solution on Parafilm®, and place the grids on the droplet for 20 min.53.Wash the grids three times by dipping them into 10 mL of ddH2O.54.Place the grids on the droplet of lead citrate solution on Parafilm® for 2 min.55.Wash the grids three times by dipping into 10 mL of ddH2O, and let them dry.Pause point: The ultrathin sections can be stored in a grid box at 20°C–25°C for several years.
Transmission electron microscopy (TEM) imaging
Timing: 1 day
This step describes a procedure for TEM imaging. We usually acquire images at a resolution of 1.2–16.1 nm/pixel.56.Set an accelerating voltage of 80 kV and a filament voltage of approximately 20 V.57.Load the grids onto the grid holder.58.Locate the ROI under a lower magnification mode.59.Increase the magnification, and acquire individual images.60.Record captured images.61.Image processing is done using ImageJ software.
Expected outcomes
By following the clearing protocol of ScaleSF, one can make mouse brain slices of 1-mm thickness transparent within 14.5 h after transient shrinkage and expansion (). A multi-scale neuronal imaging in the mouse striatofugal projection system is shown in. The mouse striatofugal projection is labeled by an injection of a LM/EM dual labeling vector, AAV2/1 SynTetOff EGFP-APEX2, into the caudate-putamen (CPu). Following tissue slice preparation, APEX2/BT-GO reaction and ScaleSF tissue clearing, the cleared brain slice is subjected to neural circuit mapping using a CLSM equipped with a multi-immersion objective lens of long WD (A). EGFP-labeled axons arising from the CPu extended caudally to the brainstem, forming dense terminal fields in the external segment of the globus pallidus (GPe) and substantia nigra (SN). After re-sectioning the imaged brain slice perpendicularly, high-resolution image stacks collected with a high NA objective lens show varicose axon arborizations of the EGFP-labeled neurons in the SN (C1). Then, the imaged re-section is processed for ABC/DAB-Ni2+ reaction using biotin molecules deposited by APEX2/BT-GO reaction. One should see DAB-Ni2+ precipitates in the re-section. Following epoxy resin embedding (C2) and ultra-thin sectioning, an axon terminal filled with DAB-Ni2+ precipitates is imaged with TEM (D1 and 4D2). TEM imaging shows a symmetric synapse, which is characterized by the absence of PSD and the narrow synaptic cleft, on a dendrite of a SN neuron (D2). For accurately tracking targets across multiple spatial scales, we took advantage of GFP fluorescence, DAB-Ni2+ labeling and/or endogenous landmarks such as shape of brain structure and blood vessels. Unambiguous correlation of LM and EM datasets was achieved by LM/EM dual labeling with a correlative light and electron microscopy (CLEM) probe, EGFP-APEX2.Figure 4Multi-scale LM/EM neuronal imaging of the mouse striatofugal projection system(A) The procedure for multi-scale LM/EM neuronal imaging.(B) Neural circuit mapping in a 1-mm-thick brain slice cleared with ScaleSF. The AAV2/1-SynTetOff-EGFP-APEX2 vector is injected into the CPu. A re-section of 50-μm thickness is cut along the dotted line. Scale bar, 500 μm. MFB: medial forebrain bundle.(C1) High-resolution subcellular imaging in the re-section at the level of SN. Scale bar, 200 μm.(C2) DAB-Ni2+ labeling with APEX2/BT-GO reaction. The re-section is embedded in epoxy resin. Scale bar, 50 μm. BV: blood vessel.(D1) A TEM image of the rectangle in (C2). Scale bar, 10 μm. Arrows in (C2 and D1) indicate the identical blood vessel.(D2) A high magnification image of the rectangle in (D1). Arrowheads indicate postsynaptic membrane. Scale bar, 500 nm. AT: axon terminal.Reprinted and modified from [Furuta et al. (2022)] under the Creative Commons Attribution 4.0 International License (CC BY 4.0; https://creativecommons.org/licenses/by/4.0/).
Limitations
In the present protocol, we provide a detailed procedure for successive LM/EM for mouse neuronal cells with ScaleSF tissue clearing. ScaleSF is an isometric and rapid tissue clearing method and archives a high-level of preservation of ultrastructure and fluorescence signals as well as potent clearing capability. However, two limitations remain in ScaleSF tissue clearing. The first is the ultrastructural preservation. A lipid-extracting detergent in ScaleS4 solution and/or incubations of brain tissues at 37°C for a total of 10–14 h can potentially damage the ultrastructural integrity. Indeed, we found a slight but statistically significant degradation of the ultrastructure accompanied by ScaleSF tissue clearing ([Furuta et al., 2022]). The second is the clearing capability of ScaleSF: brain slices of 1-mm thickness, but not a whole brain, can be cleared with the clearing protocol of ScaleSF. Although 1-mm-thick brain slices can provide good knowledge of dendritic and local axonal arbors, information about long-range projections is fragmentary and incomplete in these slices ([Stepanyants et al., 2009]).
In the protocol described here, LM/EM dual labeling is implemented by coupling a genetically encoded CLEM probe, EGFP-APEX2, with APEX2/BT-GO reaction. Despite its potent LM/EM dual labeling and resistance for ScaleSF tissue clearing, APEX2/BT-GO reaction itself and/or permeabilization for intracellular staining with a lipid-extracting detergent can potentially negatively affect the cellular ultrastructure. The LM/EM dual labeling gave strong EM contrast that was introduced in the form of osmiophilic polymers throughout the cytoplasm (). Although the cytoplasmic labeling facilitates the identification of targeted structures, the labeling may interfere with interrogation of ultrastructural features of synapses such as active zones, PSDs and synaptic vesicle morphologies.
Troubleshooting
Problem 1
Brain tissues look opaque after clearing (step 13).
Potential solution
Use ScaleS solutions that are freshly prepared. ScaleS solutions can be stored up to 1 month at 4°C. Consider prolonged incubation in ScaleS4 solution.
Problem 2
Weak or very light fluorescent signal (step 21).
Potential solution
Use ScaleS solutions that are freshly prepared. Fluorescent proteins are sensitive to certain environmental conditions, such as pH, ions, and redox state.
Problem 3
3D reconstruction of CLSM images is not acceptable (step 22).
Potential solution
Set the correction collar of the multi-immersion objective lens to 1.47. ScaleS4 solution has a RI of 1.47 ([Hama et al., 2015]; [Miyawaki et al., 2016]). RI mismatch-induced aberrations can highly disturb the image formation.
Problem 4
Ice crystal damage is evident in re-sections (step 28).
Potential solution
Cool isopentane almost to its freezing temperature. Permeate brain tissues with the cryoprotectant agent adequately. This problem is caused by too slow a freezing rate and/or inadequate cryoprotection of tissues.
Problem 5
No DAB-Ni2+ precipitation (step 35).
Potential solution
NaN3 as a preservative should be avoided before APEX2/BT-GO reaction. NaN3 inactivates the enzymatic activity of APEX2. Use BT-GO reaction mixture that is freshly prepared.
Problem 6
Weak or very light DAB-Ni2+ reaction (step 35).
Potential solution
Consider increasing the concentration of BT solution. We usually dilute the solution from 1:5,000 to 1:50,000 for the total. Try to increase the incubation period of DAB-Ni2+ solution. Additional H2O2 can be supplied to the DAB-Ni2+ solution during color development. Following longer storage before APEX2/BT-GO reaction, the staining usually becomes less intense.
Problem 7
Strong DAB-Ni2+ reaction is observed only at the edge of tissues (step 35).
Potential solution
This problem is caused by poor penetration of BT into brain tissues. Verify that brain tissues are shaken thoroughly in both permeabilization buffer and BT-GO reaction mixture. For penetration of BT, it is important to shake brain tissues thoroughly in both solutions. Consider increasing the incubation period of permeabilization buffer and/or BT-GO reaction mixture. Prolonged incubation in permeabilization buffer, however, can potentially damage ultrastructure in brain tissues.
Problem 8
No DAB-Ni2+ staining in the center of the infected cells (step 35).
Potential solution
Consider decreasing the AAV labeling density. This problem can occur when the labeling density of APEX2 is extremely high. This might be due to the local reactant depletion ([Zhang et al., 2019]).
Problem 9
Very high background (step 35).
Potential solution
Try to decrease the concentration of BT solution. The optimal concentration of BT solution should be determined for each experiment. We usually dilute the solution from 1:5,000 to 1:50,000 for the total.
Problem 10
Strong DAB-Ni2+ reaction is observed in vasculature (step 35).
Potential solution
Perfuse brain tissues with ice-cold PBS adequately to remove red blood cells. Endogenous peroxidases in red blood cells retained in vasculature can react with H2O2 to deposit BT onto tissues.
Problem 11
Poor ultrastructural preservation (step 56).
Potential solution
Consider increasing the concentration of GA in the fixative solution. High concentration of GA fixation provides superior ultrastructure preservation in brain slices cleared with ScaleSF ([Furuta et al., 2022]). Try to decrease the incubation period of ScaleS4 solution and/or permeabilization buffer. These solutions contain a lipid extracting detergent that deleteriously affects membrane integrity. Sample handling at low temperature can minimize ultrastructural damage.
Resource availability
Lead contact
Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Hiroyuki Hioki (h-hioki@juntendo.ac.jp).
Materials availability
This study did not generate new unique reagents.
Acknowledgements
AcknowledgmentsWe thank Keiko Okamoto-Furuta and Haruyasu Kohda (Division of Electron Microscopic Study, Center for Anatomical Studies, Graduate School of Medicine, Kyoto University) for technical assistance in EM. We are also grateful to Yoko Ishida (Juntendo University) for AAV vector production and Kisara Hoshino (Juntendo University) for technical assistance. This study was supported by 10.13039/501100001691JSPS KAKENHI (JP20K07231 to K.Y.; JP21H03529 to T.F.; JP20K07743 to M.K.; JP21H02592 to H.H.) and Scientific Research on Innovative Area “Resonance Bio” (JP18H04743 to H.H.). This study was also supported by the 10.13039/100009619Japan Agency for Medical Research and Development (AMED) (JP21dm0207112 to T.F. and H.H.), Moonshot R&D from the Japan Science and Technology Agency (JST) (JPMJMS2024 to H.H.), 10.13039/501100020964Fusion Oriented Research for Disruptive Science and Technology (FOREST) from JST (JPMJFR204D to H.H.), Grants-in-Aid from the Research Institute for Diseases of Old Age at the Juntendo University School of Medicine (X2016 to K.Y.; X2001 to H.H.), and the Private School Branding Project.Author contributionsConceptualization, K.Y., T.F., and H.H.; Methodology, K.Y., T.F., M.T., and H.H.; Investigation, K.Y., T.F., and S.O.; Formal Analysis, K.Y., T.F., M.T., and H.H.; Writing – Original Draft, K.Y., T.F., and H.H.; Writing – Review & Editing, K.Y., T.F., M.K., and H.H.; Project Administration, H.H.; Funding Acquisition, K.Y., T.F., M.K., and H.H.Declaration of interestsThe authors declare no competing interests.