神经酰胺：把“酒精杀细胞”翻译成“细胞被迫增殖”的那个脂质开关
下划线 = 中文容易“变味”或丢信息的词句：红线带编号 = 13 条注释卡（悬停看、点击跳转）。这一篇问的是一个反直觉的问题：酒精除了把神经细胞杀死，会不会还逼着活下来的前体细胞多分裂几次？答案被作者压在一个脂质分子上——神经酰胺。第一作者是不是同一人（2012、2005 那两篇中文论文的作者）见“这篇在轨迹里的位置”一节，那里给了依据。英文原文附于文末
摘要（全译）
孕期酒精暴露，尤其在妊娠早期，可以导致胎儿酒精综合征。乙醇的药理学与毒理学机制，与神经酰胺（ceramide）的效应有关。本研究在野生型小鼠与神经酰胺代谢关键酶——鞘磷脂合酶 2（sphingomyelin synthase 2，SMS2）——的基因敲除小鼠中建立了酒精暴露模型。该模型每天接受 25% 乙醇灌胃，仔鼠在生后第 0、7、14、30 天（P0、P7、P14、P30）用于实验。血清学与免疫荧光染色发现：乙醇暴露以剂量依赖的方式降低两种基因型仔鼠血中的鞘磷脂水平，并增加海马齿状回的神经细胞增殖与新生神经元数量。Western blot 分析显示，乙醇暴露后两种基因型仔鼠中蛋白激酶 C α 的相对表达水平升高。与野生型仔鼠相比，鞘磷脂合酶 2 敲除仔鼠海马中，神经酰胺／神经酰胺-1-磷酸通路的重要激活蛋白——蛋白激酶 C α——的表达水平降低。我们的结果说明：孕期乙醇暴露后，神经酰胺参与了仔鼠海马齿状回中酒精诱导的神经增殖，其机制可能与蛋白激酶 C α 表达升高、进而激活神经酰胺／神经酰胺-1-磷酸通路有关。
摘要正文取自同一篇论文的 Europe PMC 记录（XML 中的 abstract 节）——抽取出来的 markdown 里这一节只剩关键词，我在方法一节末尾把这件事说清楚了。
引言（全译）
急性酒精暴露可以引起大量神经细胞凋亡，但它能不能引起神经细胞增殖？这一现象又受哪些机制调控？本文把焦点放在海马齿状回的神经细胞增殖上，因为这个区域属于古皮质，是记忆与学习的储存区。海马同时也是神经细胞增殖相对活跃的区域。
长期过量饮酒可损伤机体多个器官与系统，包括消化、心血管、免疫与神经系统。酒精能穿过血脑屏障，直接对中枢神经系统产生神经毒性，导致神经细胞凋亡与突触减少。脑功能甚至会被改变，出现认知损害与神经退行性疾病，如阿尔茨海默病与帕金森病。临床文献与动物实验显示，在子宫内的孕期酒精暴露——尤其是妊娠早期——可导致死胎、流产、早产，甚至胎儿酒精综合征。患胎儿酒精综合征的儿童可能有躯体畸形、器官发育异常、智力低下，以及中枢神经系统功能损害。
酒精神经毒理学的机制尚不清楚。一些学者认为，酒精通过抑制 NMDA 受体、激活 GABA 受体而产生效应。酒精还可通过影响 Na⁺-K⁺-ATP 酶与腺苷酸环化酶的活性、以及调节 Ca²⁺ 通道的开放，诱导细胞凋亡。另一些学者认为，其药理与毒理机制涉及神经酰胺。作为一个重要的信号转导分子，神经酰胺参与多种细胞转导通路，可调节细胞周期、细胞分化、增殖与凋亡（包括自噬）。神经酰胺还参与应激、免疫反应与炎症等多种细胞生理功能。神经酰胺是一种中间型鞘脂，其磷酸化、糖基化与酰化产物参与不同的细胞生理功能。神经酰胺的主要代谢出路，是通过鞘磷脂合酶（sphingomyelin synthase，SMS）转化为鞘磷脂；该酶在鞘脂与磷脂的代谢过程中居于中心地位，也参与调节细胞内神经酰胺与二酰甘油的水平。在哺乳动物基因组中，鞘磷脂合酶有两个同工酶：SMS1 与 SMS2。用 RNA 干扰使 SMS1 或 SMS2 基因缺失，会减少鞘磷脂的合成，导致神经酰胺堆积，进而使其细胞效应增强。本项目用 SMS2 基因敲除（SMS2-/-）小鼠构建孕期酒精暴露模型，以研究神经酰胺在酒精诱导的神经细胞增殖过程中的作用，而基因敲除这件事本身在实验设计上意味着什么，后面方法一节会再拆。
本研究有以下几个特点：（1）用 SMS2-/- 小鼠建立酒精暴露模型；（2）实验设计与亚组分析是随机的；（3）分析旨在阐明酒精暴露、神经酰胺与神经增殖三者之间的关系。
越来越多的证据已经确认，酒精能诱导神经细胞凋亡；但很少有研究去观察酒精是否能促进神经细胞增殖。我们的研究是国内首次在基因敲除小鼠中建立酒精暴露模型。本研究的目的：确定酒精能否诱导神经细胞增殖及其机制；研究 SMS2 基因敲除是否影响酒精诱导的神经增殖；并阐明神经增殖及所涉通路在上述过程中的作用。
方法（精译）
原文按 NRR 的排版把 MATERIALS AND METHODS 放在讨论之后、文章末尾。我按“动物—模型—读数—统计”重排了一遍，方便你对着自己的实验本看。
动物与基因型鉴定
SMS2 杂合子小鼠由美国纽约州立大学下州医学中心的蒋先诚（Xiancheng Jiang）教授赠送（1 雄 2 雌，平均体重 21.2 ± 1.7 g，清洁级，动物许可证号 NSJXC-79J14537）；野生型 C57BL/6J 小鼠 10 雄 20 雌（平均体重 20.7 ± 2.1 g，清洁级，许可证号 HNKF36790131）由河南大学实验动物中心提供。共使用 P0–P30 仔鼠 360 只。仔鼠由杂合子互交产生，基因型用 PCR 鉴定：凝胶上可能出现 250 bp 与 145 bp 两条带，野生型样本只有 145 bp 带，杂合子样本两条带都有，敲除样本只有 250 bp 带。所用引物信息原文列于表 1。全部实验按科技部《关于善待实验动物的指导性意见》执行，以保证实验期间动物福利。
时间与地点：2009 年 3 月至 2011 年 3 月，中国河南大学生命科学学院神经生物学研究所。
酒精暴露模型（Establishment of the alcohol exposure model）
每天早上检查 SMS2 敲除雌鼠有无阴栓，查到阴栓的那一天记为 E0（E = 受孕天数，E0 = 交配雌鼠查到阴栓；P = 生后天数，P0 = 出生后第一个 24 小时）。仔鼠出生后用 PCR 鉴定基因型，野生型与 SMS2 敲除仔鼠再分入两组。
（1）酒精处理组：雌鼠每天灌胃 25%（w/v）酒精（开封市文明（集团）有限公司，河南），剂量为 2.0 g/kg 或 4.0 g/kg，从 E5 开始，一直持续到分娩。为了让胃排空、便于酒精吸收，给药前 2 小时撤去食物。每天同一时间称重并灌胃。
（2）对照组：普通饲料组自由进食饮水，只在称重前撤食 2 小时；配对喂养（pair fed）组则灌胃与酒精组等热量、等体积的麦芽糊精。由于配对喂养组与普通饲料组在神经增殖上没有显著差异，两组数据被合并，以提高统计效能。为观察乙醇毒性对中枢神经系统发育的长期影响，仔鼠在 P0、P7、P14、P30 取材。
血酒精浓度
在 E10 与 E15，于灌胃后 60 分钟与 120 分钟用剪刀剪开酒精处理孕鼠的尾尖，取 10 μL 血入肝素化毛细管，离心约得 5 μL 血清，用 ANALOX GL5 分析仪（德国）测血酒精浓度。每只母鼠得到四个读数（E10、E15 各两个）。这四个值中的最高值，被定义为该母鼠及其仔鼠的峰值血酒精浓度。4.0 g/kg 组的峰值落在 30–50 之间，2 g/kg 组落在 15–30 之间（原文把单位写作 mg/L），据此把两组分别称为高剂量组与中剂量组。
血清鞘磷脂测定
取各组 P0 仔鼠，心脏穿刺采血。鞘磷脂水平用酶法测定：SMase 水解鞘磷脂，生成磷酸胆碱与 N-酰基鞘氨醇；碱性磷酸酶从磷酸胆碱生成胆碱；胆碱在胆碱氧化酶催化下生成过氧化氢；过氧化氢在过氧化物酶催化下与 DAOS、4-氨基安替比林反应，生成在 595 nm 处有最大吸收的蓝色染料。用 Bio-Rad 680 自动酶标仪在 595 nm 读数，按标准曲线计算血中鞘磷脂浓度。
免疫组织化学
各年龄仔鼠用戊巴比妥钠（20 mg/kg）麻醉，4% 多聚甲醛经心脏灌注固定；取脑，4°C 继续浸固定 1–2 天；振动切片机切 50 μm 冠状脑片。0.1 mol/L PBS 洗三次后，用不同一抗及相应二抗做免疫荧光单标或双标。一抗包括：兔抗 doublecortin（DCX）多克隆抗体（1:200；#4604，Cell Signaling，Danver，MA，USA）、兔抗 Prox 1 多克隆抗体（1:400；AB5475，Chemicon，Billerica，MA，USA）、兔抗蛋白激酶 C α 多克隆抗体（1:50；P4334，Sigma，St. Louis，MO，USA）、小鼠单克隆抗 NeuN 抗体（1:500；MAB377B，Millipore，Billerica，MA，USA）。4°C 过夜后，0.1 mol/L PBS 洗三次，再加二抗：Alexa Fluor 568 驴抗小鼠 IgG（1:300；A10037，Invitrogen Life Technologies，Carlsbad，CA，USA）与 Alexa Fluor 488 驴抗兔 IgG（1:600；A21206）。室温孵育 3 小时后洗片、荧光封片剂封片。在罗丹明与 FITC 激发下，用荧光显微镜（BX61，Olympus，Tokyo，Japan）观察拍照；标记良好的切片再用激光扫描共聚焦显微镜（Olympus）观察成像。
BrdU 免疫荧光标记
BrdU（Sigma，B-5002）溶于生理盐水中的 NaOH（7.5 mg/500 μL），以 5 μg/mg 体重的剂量注射给 E18 的孕鼠，或在 P6、P13、P29 注射给仔鼠。为增强标记，3 小时后再给一次相同剂量。注射后 1 天取材（分别对应 P0、P7、P14、P30）。仔鼠经心脏灌注，脑在 4% 多聚甲醛／0.1 mol/L PBS 中 4°C 后固定过夜；振动切片机切 50 μm 冠状脑片。为显示掺入细胞核的 BrdU，切片在 2 mol/L HCl 中 37°C 孵育 30 分钟以变性 DNA；随后用硼酸（pH 8.0）充分漂洗三次（每次 10 分钟）、0.1 mol/L PBS 洗三次（每次 10 分钟），以彻底洗净切片内的 HCl。为同时显示齿状回的增殖细胞与颗粒细胞，双标使用兔抗 Prox 1 多克隆抗体（1:400；Chemicon，AB5475）与小鼠抗 BrdU 单克隆抗体（1:100；ZSGB-BIO，北京，ZM-0013）。所有一抗 4°C 过夜；洗片后二抗室温 3 小时（Alexa Fluor 568 驴抗小鼠 IgG 1:300；Alexa Fluor 488 驴抗兔 IgG 1:600）。封片后用激光共聚焦显微镜（Olympus，FV10）观察并拍摄 Prox 1 与 BrdU 阳性细胞。
蛋白激酶 C α 的 Western blot
因为神经酰胺可能通过神经酰胺-1-磷酸与蛋白激酶 C α 影响神经增殖，作者用 WB 检测这条回路上 PKCα 的水平。取各组 P7 仔鼠，颈椎脱臼处死，开颅后迅速取脑，分离海马匀浆。用蛋白提取试剂盒（上海碧云天，P0027）提取海马胞质蛋白，考马斯亮蓝 G250 法定量；做标准曲线后测各处理组的样本蛋白浓度。电泳、转膜后加兔抗 PKCα 多克隆抗体（1:20 000；Sigma，P4334），4°C 过夜；再加 HRP 标记山羊抗兔 IgG（1:1 000；北京索莱宝）室温孵育 2 小时。每次孵育前用 PBS 洗三次、每次 5 分钟。ECL 试剂（ZSGB-BIO，北京）反应 2 分钟后曝光、显影分析。以小鼠抗 β-actin 单克隆抗体（1:1 000；上海碧云天，#AA128）为内参。目标蛋白条带与内参条带的吸光度比值，即该蛋白的相对表达水平。
计数区域与统计
计数区选在齿状回的上锥体叶（suprapyramidal blade）——即靠近海马 CA1、位于 CA3 锥体细胞层上方的那部分颗粒细胞层。用 ImageJ（NIH，Bethesda，MD，USA）在 20× 下计数从颗粒下层到颗粒上层的增殖细胞（BrdU 阳性）与新生神经元（DCX 阳性），单位为 cells/mm²。数据以 mean ± SD 表示，SPSS 17.0 做单因素方差分析，继以 Student-Newman-Keuls 检验和／或最小显著差（LSD）检验，P < 0.05 视为差异有统计学意义。
读这篇时请自己核一遍的三处不自洽（原文如此，我不替它圆）：
① 结果一节说“共使用 360 只仔鼠”，方法一节却说“总计使用 230 只仔鼠”；
② 结果一节写对照组是“每天灌胃蒸馏水”，方法一节写对照组是“普通饲料组 + 等热量麦芽糊精配对喂养”，两处不是同一件事；
③ BrdU 剂量写作“5 μg/mg 体重”，按字面换算量大到不合理；血酒精浓度写作 mg/L，按实验常规这个量级更像 mg/dL。引用时照抄原单位，但心里要留个问号。
结果（一）：血里的鞘磷脂降了，齿状回的增殖升了
先说动物。30 只野生型雌鼠与 30 只 SMS2 敲除（妊娠）小鼠被随机分为三组：低剂量酒精组（每天灌胃 25% 酒精 2.0 g/kg）、高剂量酒精组（每天灌胃 25% 酒精 4.0 g/kg）与对照组（每天灌胃蒸馏水）。在 P0、P7、P14、P30 各选 15 只仔鼠，共使用 360 只仔鼠；其中每组 10 只用于免疫组化染色，5 只用于 WB。本研究是体内实验，仔鼠选出后立即处理，没有动物丢失。
先解释为什么测血里的鞘磷脂：SMS2 基因敲掉之后，神经酰胺转不成鞘磷脂，于是神经酰胺在体内（包括中枢神经系统与血）堆积。由于鞘磷脂水平与神经酰胺浓度反向相关，血清鞘磷脂水平可以粗略反映动物体内神经酰胺的堆积程度。作者测了 P0 仔鼠的血清鞘磷脂，比较 SMS2+/+ 与 SMS2-/- 仔鼠：SMS2-/- 仔鼠的血清鞘磷脂显著低于野生型（P < 0.01）。孕期酒精暴露使两种基因型的仔鼠鞘磷脂水平都下降，且呈剂量依赖（P < 0.05）。这条证据的意思是：酒精暴露可能降低了 SMS2 的活性，从而在神经酰胺通路受抑之后，造成细胞内神经酰胺的大量堆积。
再看正常发育的基线。中枢神经系统即使在成年也有干细胞池，例如齿状回的颗粒下层，这些神经干细胞有长期增殖能力。作者先观察野生型小鼠 P0、P7、P14、P30 的齿状回：总体规律是神经增殖随年龄增加而下降。P0 时，野生型仔鼠的增殖细胞广泛分布在整个海马与齿状回；P7 时主要在齿状回，包括门区、颗粒下层、颗粒层与分子层；P14 时在颗粒下层与颗粒层；到了 P30（成年），增殖细胞基本只限于颗粒下层。SMS2-/- 小鼠遵循同一规律。统计上，野生型仔鼠齿状回神经增殖随年龄逐渐下降（F = 483.84，P < 0.001），SMS2-/- 仔鼠同样如此（F = 873.92，P < 0.001）。
然后是酒精的作用。孕期酒精暴露使野生型与 SMS2-/- 仔鼠齿状回的神经增殖都显著增加，且呈剂量依赖。统计上，各酒精处理组齿状回的神经增殖显著高于对照组（P < 0.01），高剂量组又显著高于低剂量组（P < 0.05）。SMS2-/- 仔鼠呈现同样的模式（P < 0.05），说明孕期酒精暴露能以剂量依赖的方式、在较长的时间里持续诱导神经增殖。为了解神经酰胺在其中扮演的角色，作者比较了有／无酒精暴露的野生型与 SMS2-/- 仔鼠：无论对照组还是酒精组，SMS2-/- 仔鼠齿状回的神经增殖都高于野生型（P < 0.05）。这提示，神经酰胺的堆积上调了酒精诱导的神经增殖。
结果（二）：新生神经元跟着多，PKCα 跟着升
数“正在分裂的细胞”还不够，得看这些分裂有没有变成神经元。于是作者在不同处理组标记了新生神经元。Doublecortin 是一种微管相关蛋白，用于标记迁移中与分化中的新生神经元；在哺乳动物中，DCX 阳性神经元的分布模式与齿状回增殖细胞的分布相似。本研究中，齿状回的新生神经元用 DCX 特异标记，以进一步观察酒精诱导的神经酰胺堆积是否影响新生神经元的形成。结果：新生神经元的发育与分布模式跟增殖神经细胞相似，主要分布在齿状回颗粒下层，形态上呈现颗粒细胞的特征。齿状回新生神经元数量虽随年龄增加而减少，但 SMS2 敲除与野生型之间存在数量差异（P < 0.05）——同年龄时，SMS2 敲除仔鼠的新生神经元多于野生型。
而且孕期酒精暴露使野生型与 SMS2 敲除仔鼠的新生神经元都上调；高剂量酒精组的新生神经元密集聚集、突起很长。孕期酒精暴露后，两种仔鼠的新生神经元数量都显著高于对照组（P < 0.01），且这种增加呈剂量依赖（P < 0.05）。这说明孕期酒精暴露能诱导新生神经元的形成，而神经酰胺在这一过程中起了作用。
最后是分子层面的读数。蛋白激酶 C α 被作者当作“评价神经酰胺通路活性的关键蛋白与重要标志物”：抑制 PKCα 活性可以阻断神经酰胺通路以及神经酰胺-1-磷酸诱导的细胞有丝分裂。免疫细胞化学显示，PKCα 主要表达于新皮质、室管膜下区、海马、齿状回、门区与脑干细胞的胞质，提示 PKCα 可能参与神经细胞中大部分信号转导过程。PKCα（绿色）既表达于 NeuN（红色）阳性的成熟神经元，也表达于新生神经元，主要位于颗粒下层。为证明 PKCα 表达对酒精诱导神经增殖的影响，作者用 WB 半定量检测 P7 各组海马中 PKCα 的相对表达水平：与野生型仔鼠相比，SMS2 敲除仔鼠海马细胞中 PKCα 的表达相对较低；酒精暴露后，两种基因型组的 PKCα 表达都升高——换句话说，酒精处理组的 PKCα 表达水平显著高于对照组，且呈剂量依赖。半定量结果：野生型仔鼠海马细胞中 PKCα 的相对表达水平为 0.32 ± 0.03，SMS2 敲除仔鼠降至 0.24 ± 0.02（P < 0.05）；孕期酒精暴露后，两种基因型组的 PKCα 相对表达都升高（P < 0.01），并呈剂量依赖。
表 1 实验设计速查（据原文 Results 与 Methods 整理）
| 项目 | 内容 |
|---|---|
| 动物来源 | SMS2 杂合子（1 雄 2 雌，21.2 ± 1.7 g）赠自美国 SUNY Downstate 蒋先诚教授；野生型 C57BL/6J（10 雄 20 雌，20.7 ± 2.1 g）购自河南大学实验动物中心 |
| 基因型鉴定 | PCR；250 bp 与 145 bp 两条带：野生型只有 145 bp，杂合两条都有，敲除只有 250 bp |
| 孕期起点 | 查到阴栓当天记为 E0；E5 开始灌胃，直到分娩 |
| 酒精剂量 | 25%（w/v）乙醇，低剂量 2.0 g/kg·d，高剂量 4.0 g/kg·d（灌胃） |
| 对照设计 | 普通饲料组（自由进食，称重前撤食 2 小时）＋ 配对喂养组（等热量等体积麦芽糊精灌胃）；两组无差异后合并 |
| 取材时间点 | P0、P7、P14、P30 |
| 每组数量 | 每个时间点每个处理组至少 15 只仔鼠；10 只做免疫组化，5 只做 WB |
| 峰值血酒精 | 4.0 g/kg 组 30–50；2.0 g/kg 组 15–30（原文单位写作 mg/L） |
表 2 关键数字速查（只摘原文给过的值，未给的不补）
| 指标 | 结果 | 统计 |
|---|---|---|
| P0 仔鼠血清鞘磷脂（SMS2-/- vs 野生型） | SMS2-/- 显著更低 | P < 0.01 |
| 孕期酒精降血清鞘磷脂 | 两种基因型都降，剂量依赖 | P < 0.05 |
| 齿状回增殖随年龄下降 | 野生型 F = 483.84；SMS2-/- F = 873.92 | P < 0.001 |
| 酒精升高齿状回增殖 | 各酒精组 > 对照组 | P < 0.01 |
| 高剂量 vs 低剂量 | 高剂量组更高 | P < 0.05 |
| 增殖（SMS2-/- vs 野生型） | SMS2-/- 更高 | P < 0.05 |
| DCX 阳性新生神经元 | 酒精组 > 对照组；剂量依赖 | P < 0.01；剂量依赖 P < 0.05 |
| PKCα 相对表达（P7 海马） | 野生型 0.32 ± 0.03；SMS2-/- 0.24 ± 0.02 | P < 0.05 |
| 酒精升高 PKCα 表达 | 两种基因型都升，剂量依赖 | P < 0.01 |
讨论要点
为什么盯着神经酰胺。孕期酒精暴露是发育中神经系统的有效毒物，可造成广泛的神经病理损伤，甚至胎儿酒精综合征，严重影响胎儿多器官系统发育，尤其是中枢神经系统。酒精对中枢神经系统的作用主要通过某些神经递质受体（谷氨酸 GluR2/3、NMDAR，γ-氨基丁酸，神经型乙酰胆碱，多巴胺受体）。酒精还能激活下游信号通路，如 c-Jun 端激酶（JNK）通路，诱导细胞凋亡。而在这一片“怎么死”的研究之外，Saito 等人发现酒精可以通过神经酰胺的从头合成（de novo synthesis）诱导神经细胞凋亡，最终导致病理性神经退行与代偿性神经增殖。在神经酰胺通路里，神经酰胺-1-磷酸（由神经酰胺激酶磷酸化神经酰胺产生）与细胞外信号调节蛋白激酶，是诱导凋亡与代偿性增殖的关键要素。但酒精诱导的神经增殖是否专门通过神经酰胺通路实现，此前并不确定。这篇就是用 SMS2 敲除仔鼠来证明一个假设：酒精诱导的神经酰胺堆积会上调海马神经增殖。
血鞘磷脂为什么能当读数。SMS 在哺乳动物体内有两个同工酶：SMS1 主要集中在高尔基体，SMS2 主要分布在细胞膜；神经酰胺主要被转运到细胞外，由 SMS2 催化合成鞘磷脂。有证据表明，SMS2 的缺失或失活会阻断鞘磷脂合成，导致血中鞘磷脂水平下降与细胞内神经酰胺堆积，进而通过下游信号级联产生不同的细胞效应。所以“酒精让血鞘磷脂下降”这句话，被作者读成“神经酰胺参与了酒精毒性的机制”。
为什么增殖是“代偿”。发育中神经增殖的调控机制了解得并不充分，已知可能参与的通路有 MAPK、Wnt、转化生长因子-β、Shh；近期研究显示神经酰胺-1-磷酸通路参与神经增殖的上调。成年后，中枢神经系统只有少数区域保留细胞增殖能力：侧脑室下区、齿状回颗粒下层、嗅球吻侧迁移流。这篇选齿状回作为观察窗口，看到的是一条“长期、剂量依赖”的效应。作者的解释是：代偿性神经增殖很可能是对酒精诱导的神经细胞凋亡的代偿——增殖性干细胞以有丝分裂与细胞周期调节回应凋亡，分化后的多形性神经元再迁移到海马或皮质，替代部分凋亡细胞，维持神经系统功能的完整性。
分子链条怎么讲圆。神经酰胺是一种生物活性鞘脂，调节凋亡与细胞增殖等关键功能；神经增殖主要通过它的磷酸化产物神经酰胺-1-磷酸实现。这条链里 PKCα 的位置很关键：PKCα 的磷酸化需要 SMS 激活，磷酸化后的 PKCα 又能增强磷脂酰肌醇 3-激酶／蛋白激酶 B（PI3K/Akt）的激活，从而直接调节下游 C1P 的活性；最终 C1P 引起 Ca²⁺ 内流，刺激细胞凋亡与增殖。所以“导致细胞增殖被激活的一个主要因素，是 PKCα 同工酶的活化，而这类活化需要它们被招募到细胞膜上”。C1P 能诱导 PKCα 这一重要位点的磷酸化——正因如此，测 PKCα 的表达量，可以帮助理解神经酰胺如何调节酒精诱导的神经增殖。免疫细胞化学显示 PKCα 在多数神经细胞中表达，尤其在增殖细胞里，提示它既是神经元活动、也是神经增殖所需要的。半定量结果显示 SMS2 敲除仔鼠海马神经元里 PKCα 表达相对较低，作者归因于“缺 SMS 会降低 PKCα 活性，从而拖慢神经酰胺／C1P 级联”；反过来，酒精能以剂量依赖的方式诱导 PKCα 表达。
作者自己的一段总结（原文 In summary 段）。SMS2 敲除仔鼠的血清鞘磷脂显著低于野生型；酒精暴露使两种基因型的鞘磷脂水平都以剂量依赖的方式下降；有意思的是，SMS2 敲除仔鼠齿状回的神经增殖比野生型更多，而酒精还能进一步以长期、剂量依赖的方式诱导增殖。PKCα 的磷酸化可能是神经酰胺通路被激活的关键一步；酒精诱导的 PKCα 表达，可以激活 C1P 信号级联，从而增强神经增殖与新生神经元的形成。
期刊自带的四段式小结（原文末尾，译）
研究背景 大量临床资料与动物实验证明，孕期酒精暴露（尤其妊娠早期）可导致死胎、流产、早产，甚至胎儿酒精综合征。酒精的神经毒理学机制尚不十分清楚。
研究前沿 一些学者认为，酒精可通过影响 Na⁺/K⁺-ATP 酶与腺苷酸环化酶的活性、以及调节 Ca²⁺ 通道的开放来诱导细胞凋亡；但酒精的药理与毒理机制可能取决于神经酰胺。
临床意义 本研究的结果或可为胎儿酒精综合征的发病机制提供证据，并有助于进一步理解神经酰胺在神经细胞中的生理功能与意义。
学术名词 胎儿酒精综合征：母亲在孕期大量饮酒，造成胎儿及出生后的生长发育障碍。
同行评议意见 这是国内首个用 SMS2 敲除小鼠建立酒精暴露模型的研究。该模型丰富了关于酒精毒理机制的认识，并进一步揭示了防治胎儿酒精综合征的新思路。
这篇在轨迹里的位置
先说署名这件事，不含糊。2013 这篇的作者表里有 Jinbo Deng（邓锦波）——正是 2012 年那篇中文论文（《小鼠孕期酒精暴露可诱导仔鼠齿状回苔藓细胞的丢失》，神经解剖学杂志）的合作者；两篇的第一作者姓名同为王志新；2013 年的实验在河南大学生命科学学院神经生物学研究所完成（2009 年 3 月—2011 年 3 月），而 2012 年那篇中文论文的第一作者单位写的是郑州大学基础医学院人体解剖学教研室。同一个姓名、同一个合作者、同一条酒精暴露线，判定为同一人。
把三篇按顺序摆开，能看到一个人的问题在收缩：2005 年（河南大学学报医学版，《黄芪对神经细胞缺氧缺血损伤的保护作用》）问的是“损伤能不能被人为减轻”——在培养的神经细胞上试一味中药。2012 年（神经解剖学杂志）换成整只小鼠、孕期酒精、齿状回苔藓细胞，把“酒精真的会改变某一群具体细胞的数量”钉了下来。2013 年（这一篇）往下走了一层：不再只问“少了什么细胞”，而问“酒精在细胞层面到底干了什么、由哪个分子递话”。
这一篇往前推的那一格，是方向。前两篇的逻辑都是“酒精 → 损伤”；这一篇把同一批动物身上同时发生的两件事并排放：细胞在死，前体细胞也在多分裂。作者用 compensatory（代偿）把它们连成一条因果——死逼出了生。然后给这条因果配了一个分子开关：SMS2 敲掉 → 神经酰胺堆着花不出去 → C1P／PKCα 这条支路被点亮 → 增殖上去。所以它不是一篇“酒精有害”的论文，而是一篇“损伤与修复共用一条通路”的论文。
它也确实是这条线的最后一格：目前能公开检索到的王志新论文就到这里（专区首页的时间线停在 2013）。三篇、八年、一个问题——从“能不能救”到“坏在哪”再到“坏与修之间那根线是什么”。这种节奏在今天的评价体系里几乎不可能出现，但它恰好说明一件事：一个人可以对一条线认真到不追热点。
读法上的一个提醒：这一篇的方法部分排在讨论之后（NRR 的排版习惯），你要是照着这个顺序写自己的稿子，编辑会退回来。
给刚进组的你
1. 先把这条时间轴画出来。E0（阴栓）→ E5（开始灌胃）→ 分娩 → P0／P7／P14／P30。这篇的每一个结论都挂在这条轴上：增殖随年龄下降、酒精的效应能持续到 P30、取材点决定你数到的是哪一批细胞。你以后做任何发育或暴露实验，第一件事都是画这条轴，而不是先想抗体。
2. 这篇里没有神经球。没有体外培养、没有 neurosphere、没有电生理，只有两样读数：免疫荧光数细胞 + WB 测一个蛋白。如果你在别处的清单里看到给这篇挂“neurosphere”标签，那不是这篇的内容——看文章先看它到底做了什么，别被标签带着走。
3. BrdU 那一段值得抄进实验本。打两次、间隔 3 小时；注射后 1 天取材；2 mol/L HCl 37°C 30 分钟变性 DNA；硼酸（pH 8.0）中和三次。这几步是 BrdU 免疫荧光成败的分水岭。你下次问师姐，可以问得很具体：“咱们 HCl 打孔是 30 分钟还是 10 分钟？中和用硼酸还是 PBS？”——这是内行话。剂量原文写 5 μg/mg，按字面换算不合理，实操量按组里惯例核一遍再用。
4. 你正在跑的 WB，这篇的 Fig 8 就是模板。内参 β-actin，目的蛋白一抗 1:20 000 过夜，HRP 二抗 1:1 000 两小时，ECL 2 分钟曝光。它给的“相对表达量”算法也很朴素：目标条带吸光度 ÷ 内参条带吸光度。你现在做的每一张膜，最后都是这个比值。
5. 一个可以直接问出去的问题。“组里现在有没有做 BrdU／DCX 双标？这篇用的抗体是 DCX #4604（CST）、Prox1 AB5475（Chemicon）、BrdU ZM-0013（中杉金桥），要不要按这个清单备货？”——问完你就知道组里下一步往哪个方向走，也顺便让师姐知道你真的读了方法。
▲ 本部分翻译难点（对照学习用）
附录 · 英文原文
Ceramide is involved in alcohol-induced neural proliferation
Abstract
Keywords: neural regeneration, brain injury, ceramide, neural cells, proliferation, prenatal alcohol exposure, sphingomyelin synthase 2 knockout mice, sphingomyelin, sphingomyelin synthase, cera-mide-1-phosphate, fetal alcohol syndrome, grants-supported paper, neuroregeneration
INTRODUCTION
Acute alcohol exposure can cause a large amount of nerve cell apoptosis, but can it cause nerve cell proliferation? Which mechanisms are involved in the regulation of this phenomenon? This paper focuses on nerve cell proliferation in the hippocampal dentate gyrus, because this region belongs to the ancient cortex and it is the storage area for memories and learning. The hippocampus is also a relatively active area of nerve cell proliferation[[1]].
Long-term excessive consumption of alcohol can cause damage to multiple organs and systems of the body, including the digestive, cardiovascular, immune, and nervous systems[[2]]. Alcohol can cross through the blood-brain barrier and directly cause neurotoxicity to the central nervous system, leading to the apoptosis of nerve cells and a decrease in synapses[[3]]. Brain functions can even be altered, with the development of cognitive impairments and neurodegenerative diseases, such as Alzheimer's disease and Parkinson's diseases[[4],[5],[6]]. Clinical literature and animal experiments have shown that prenatal alcohol exposure in utero, especially during the early stages of pregnancy, can cause stillbirth, miscarriage, premature birth, and even fetal alcohol syndrome[[6],[7]]. Children with fetal alcohol syndrome may have body deformities, abnormalities in organ development, mental retardation, and functional impairments in the central nervous system[[6]].
The mechanisms underlying alcohol neurotoxicology are not yet clear. Some scholars believe that alcohol causes its effects by inhibiting NMDA receptors[[8],[9],[10]] and activating GABA receptors[[11],[12],[13]]. Alcohol can also induce cellular apoptosis by influencing the activity of Na+-K+-ATPase and adenylate cyclase, as well as by modulating the opening of Ca2+ channels[[14]]. Other scholars believe that the pharmacological and toxicological mechanisms involve ceramide[[15],[16]]. As an important signal transduction molecule, ceramide participates in a variety of cellular transduction pathways and can modulate cell cycle, cellular differentiation, proliferation, and apoptosis (including autophagy). Ceramide also participates in many cellular physiological functions, such as stress, immune response and inflammation[[17]]. Ceramide is an intermediate-type sphingolipid, whose phosphorylation, glycosylation and acylation products participate in different cellular physiological functions[[18]]. The main metabolic product of ceramide is to transform into sphingomyelin through sphingomyelin synthase, which plays a central role in the metabolic process of sphingolipids and phospholipids. This enzyme also participates in the modulation of cellular ceramide and diacylglycerol levels[[19],[20],[21]]. In the mammalian genome, sphingomyelin synthase (SMS) has two isozymes, SMS1 and SMS2. The loss of SMS1 or SMS2 genes through RNA interference would decrease the synthesis of sphingomyelin, resulting in the accumulation of ceramide and a subsequent increase in its cellular effects[[22]]. In this project, we constructed a prenatal alcohol exposure model using SMS2 knockout (SMS2-/-) mice to investigate the role of ceramide in the process of alcohol-induced nerve cell proliferation.
Our study has the following characteristics: (1) We used SMS2-/- mice to establish an alcohol exposure model. (2) Experimental design and subgroup analyzes were random. (3) The analysis aimed to elucidate the relationships among alcohol exposure, ceramide, and neural proliferation.
Increasing evidence has confirmed that alcohol can induce nerve cell apoptosis, but few studies have observed whether alcohol can promote nerve cell proliferation. Our research is the first in China to establish an alcohol exposure model in gene knockout mice. Our aims in this study are: to determine if alcohol can induce nerve cell proliferation, and the mechanisms underlying this process; to investigate whether SMS2 gene knockout affects alcohol-induced neural proliferation, and to clarify the role of ceramide and pathways involved in the above process.
RESULTS
Quantitative analysis of experimental animals
Thirty wild-type female mice and thirty SMS2-/- mice (pregnant) were randomly divided into three groups: low-dose alcohol group (intragastric administration of 25% alcohol 2.0 g/kg per day), high-dose alcohol group (intragastric administration of 25% alcohol 4.0 g/kg per day), and control group (intragastric administration of distilled water per day). Fifteen pups were selected at postnatal days 0, 7, 14, 30 (P0, P7, P14, P30). A total of 360 pups were used for experiments. Among them, ten pups in each group were applied for immunohistochemical staining while 5 pups for western blot assay. The present study is an in vivo experiment and the pups were processed immediately after selection. There was no loss of animals.
Effect of prenatal alcohol exposure on serum sphingomyelin levels
The SMS2-/- mice could not transform ceramide into sphingomyelin due to the deletion of the SMS2 gene; therefore, ceramide accumulated in the body, including central nervous system and blood. Since sphingomyelin level was inversely related to the concentration of ceramide, serum sphingomyelin levels could roughly reflect the level of ceramide accumulation in the animals. In our study, the serum sphingomyelin levels were measured in P0 pups, and the comparisons were made between SMS2+/+ and SMS2-/- pups (). The serum sphingomyelin level of SMS2-/- pups were significantly lower than that of wild-type pups (P < 0.01). Prenatal alcohol exposure could decrease the sphingomyelin levels in both SMS2+/+ and SMS2-/- pups with in a dose dependent manner (P < 0.05). This evidence suggests that alcohol exposure could decrease the activity of SMS2, resulting in the large accumulation of ceramide in cells following the suppression of the ceramide pathway.
[Figure 1] Effect of prenatal alcohol exposure on serum sphingomyelin levels in pups at postnatal day 0.The serum sphingomyelin levels in SMS2-/- pups were significantly lower than wild-type pups. Serum sphingomyelin level decreased with increasing ethanol dose in the two groups of pups. aP < 0.01, vs. wild-type group; bP < 0.01, vs. control group (from the the same genome); cP < 0.05, vs. M-EtOH group (from the same genome). Sphingomyelin levels are shown as mean ± SD in 15 pups for each group. One-way analysis of variance and Student-Newman-Keuls test were used for statistical analysis.M-EtOH: Moderate-dose ethyl alcohol treatment (2.0 g/kg per day); H-EtOH: high-dose ethyl alcohol treatment (4.0 g/kg per day); SMS2-/-: sphingomyelin synthase 2 knockout.
Effect of prenatal alcohol exposure on neural proliferation in the hippocampal dentate gyrus
Normally, there are stem cell pools in the central nervous system even in adulthood, for example in the subgranular zone of the dentate gyrus, and these neural stem cells have long-term proliferative capabilities[[23]]. Understanding the normal development of neural proliferation in the dentate gyrus would be the prerequisite for understanding the effects of alcohol on neural proliferation. First, we observed proliferative cells in the dentate gyrus of wild-type mice at P0, P7, P14, and P30. The general pattern was that neural proliferation decreased with increasing age. At P0, the proliferative cells in wild-type pups were distributed widely throughout the hippocampus and dentate gyrus. At P7, the proliferative cells were mainly located in the dentate gyrus, including the hilus, subgranular zone, granular layer and molecular layer. At P14, the proliferative cells were located in the subgranular zone and granular layer. However, at P30 (adulthood), the proliferative cells were mainly limited to the subgranular zone, suggesting that neural proliferation in the dentate gyrus of wild-type pups gradually decreased with as age increased (). This principle of neural proliferation also was demonstrated in SMS2-/- mice (). Statistical analysis () indicated that neural proliferation in the dentate gyrus gradually decreased with increasing age in wild-type (F = 483.84, P < 0.001) or SMS2-/- pups (F = 873.92, P < 0.001).
[Figure 2] Effect of prenatal alcohol exposure on neural proliferation in the hippocampal dentate gyrus of wild-type pups at postnatal days 7, 14, 30 (laser confocal microscope, × 20).The granular cells were immunocytochemically labeled with Prox 1 (green), and proliferative nerve cells were labeled with BrdU (red). Prenatal alcohol exposure can induce neural proliferation in a dose-dependent manner.WT: Wild-type group; Cont: control group; Mod: moderate-dose alcohol (2.0 g/kg per day) group; High: high-dose alcohol (4.0 g/kg per day) group. Scale bar: 80 μm.
[Figure 3] Neural proliferation in wild-type (WT) versus sphingomyelin synthase 2 knockout (SMS2-/-) pups at postnatal day 14 (laser confocal microscope).Alcohol exposure could upregulate neural proliferation in the dentate gyrus of both WT and SMS2-/- pups in a dose-dependent manner. For both the control and alcohol groups, neural proliferation in the dentate gyrus of SMS2-/- pups was significantly higher than that of the WT pups. The granular cells (green) and proliferative nerve cells (red) were labeled using immunocytochemistry.WT: Wild-type group; Cont: control group; Mod: moderate-dose alcohol (2.0 g/kg per day) group; High: high-dose alcohol (4.0 g/kg per day) group. Scale bar: 80 μm.
[Figure 4] Effect of prenatal alcohol exposure on neural proliferation in the dentate gyrus from pups of differing ages.BrdU positive cells are quantified in ten pups for each group (data were expressed as mean ± SD). One-way analysis of variance followed by Student-Newman-Keuls test and least significant difference test was performed. Neural proliferation in the dentate gyrus gradually decreased with increasing age in both wild-type or SMS2-/- pups. Ethanol could increase neural proliferation (in a long-lasting manner) in a dose-dependent manner. Moreover, neural proliferation in the SMS2-/- pups was higher than that of the WT pups. aP < 0.01, vs. control group (from the the same genome); bP < 0.05, vs. M-EtOH group (from the same genome); cP < 0.05 vs. WT group.WT: Wild-type; SMS2-/-: sphingomyelin synthase 2 knockout; M-EtOH: moderate-dose ethanol (2.0 g/kg per day); H-EtOH: high-dose ethanol (4.0 g/kg per day).
Prenatal alcohol exposure could significantly increase neural proliferation in the dentate gyrus in both the wild-type () and SMS2-/- pups () in a dose-dependent manner (Figures,). Statistical analysis further confirmed our observation (). Neural proliferation in the dentate gyrus of alcohol-treated groups was significantly higher than that of the control group (P < 0.01), with the high-dose alcohol group being significantly more than the low-dose group (P < 0.05).
The same pattern was observed in SMS2-/- pups (P < 0.05), suggesting that prenatal alcohol exposure could dose-dependently induce neural proliferation in a long-lasting manner. To understand the roles of ceramide in alcohol-induced neural proliferation, neural proliferation was compared in wild-type and SMS2-/- pups with and without alcohol exposure. Our findings showed that neural proliferation in the dentate gyrus of SMS2-/- pups was greater than that of the wild-type pups (). Statistical analysis () also confirmed that the level of neural proliferation in the dentate gyrus of SMS2-/- knockouts was higher than that of the wild-types (P < 0.05), suggesting that the accumulation of ceramide upregulated alcohol-induced neural proliferation.
Effect of prenatal alcohol exposure on newborn neurons in the dentate gyrus of wild-type and SMS2 knockouts
To further demonstrate whether ceramide is involved in ethanol-induced neural proliferation in the dentate gyrus, newborn neurons were labeled in the different treatment groups. Doublecortin is a microtubule-associated protein used to label migrating and differentiating newborn neurons[[24]]. In mammals, doublecortin positive neurons were distributed in a similar pattern to that of the proliferative cells in the dentate gyrus[[24]]. In this study, the newborn neurons in the dentate gyrus were specifically labeled with doublecortin to further observe whether alcohol-induced ceramide accumulation can influence the formation of newborn neurons. In our experiments, the development and distribution pattern of newborn neurons were similar to those of the proliferative neural cells. Mainly distributed in the subgranular zone of the dentate gyrus, newborn neurons presented with the morphological characters of granular cells (). Although the number of newborn neurons in the dentate gyrus decreased with age increase, there was a quantitative difference between SMS2 knockouts and wild-types (P < 0.05). At the same age, the SMS2 knockout pups displayed more newborn neurons than wild-type pups.
[Figure 5] Effect of prenatal alcohol exposure on newborn neurons in the hippocampal dentate gyrus of pups at postnatal day 30 (laser confocal microscope).Doublecortin (DCX, green) positive newborn neurons were labeled using immunocytochemistry. The number of newborn neurons in the dentate gyrus of SMS2-/- pups was much higher than that of the wild-type pups, and alcohol could upregulate the formation of newborn neurons.WT: Wild-type; SMS2-/-: sphingomyelin synthase 2 knockout; Cont: control group; High: high-dose alcohol group. Scale bar: 80 μm.
Moreover, prenatal alcohol exposure could induce the upregulation of newborn neurons in both the wild-type and SMS2 knockout pups, and the newborn neurons in the high-dose alcohol group were densely concentrated with long extensions (). The number of newborn neurons in both wild-type and SMS2 knockout pups after prenatal alcohol exposure was significantly higher than that of the control groups (P < 0.01). This increase occurred in a dose-dependent manner (P < 0.05;), suggesting that prenatal alcohol exposure could induce the formation of newborn neurons and that ceramide played an important role in this process.
[Figure 6] Effect of prenatal alcohol exposure on newborn neurons in the hippocampal dentate gyrus from pups of differing ages.The number of newborn neurons in the dentate gyrus of SMS2-/- pups was much higher than that of the WT pups. Alcohol could upregulate the formation of newborn neurons. The number of doublecortin positive cells is shown as mean ± SD in ten pups for each group. One-way analysis of variance followed by Student-Newman-Keuls test and least significant difference test was performed. aP < 0.01, vs. control group (from the same genome); bP < 0.05, vs. M-EtOH group (from the same genome); cP < 0.05, vs. WT group.WT: Wild-type; SMS2-/-: sphingomyelin synthase 2 knockout; M-EtOH: moderate-dose alcohol (2.0 g/kg per day); H-EtOH: high-dose alcohol (4.0 g/kg per day).
Role of protein kinase C α in the ceramide 1-phosphate signal pathway
Protein kinase C α is a key protein and an important marker for the evaluation of the activity of the ceramide pathway. Inhibition of protein kinase C α activity could block the ceramide pathway and ceramide-1-phosphate-induced cell mitosis[[25]]. In our study, protein kinase C α-positive cells in the hippocampus were visualized with immunocytochemistry. Protein kinase C α was mainly expressed in the cytoplasm of cells in the neocortex, subventricular zone, hippocampus, dentate gyrus, hilus and brain stem, suggesting that protein kinase C α may participate in the majority of signal transduction processes in nerve cells.
As shown in, protein kinase C α (green) was expressed in either NeuN (red) positive mature neurons or newborn neurons, which were mainly located in the subgranular layer. To prove the effects of protein kinase C α expression on alcohol-induced neural proliferation, we used western blot assay to semi-quantitatively detect the relative expression level of protein kinase C α in the hippocampus of different groups at P7 (). Compared with wild-type pups, the expression of protein kinase C α in the hippocampal cells of SMS2 knockouts was relatively low. After alcohol exposure, protein kinase C α expression in the two genotypic groups increased. In other words, the level of protein kinase C α expression in the alcohol-treated groups was significantly more than that of the control group. This expression increase occurred in a dose-dependent manner. Semi-quantitative analysis results () indicated that the relative expression level of protein kinase C α in the hippocampal cells of wild-type pups was 0.32 ± 0.03, but it was decreased to 0.24 ± 0.02 in SMS2 knockouts (P < 0.05). After prenatal alcohol exposure, the relative protein kinase C α expressions in two genotypic groups increased (P < 0.01) in a dose-dependent manner.
[Figure 7] Effect of prenatal alcohol exposure on the expression of protein kinase C (PKC) α in wild-type pups at postnatal day 7.Protein kinase C α was expressed in the cytoplasm of nerve cells in the neocortex, subventricular zone, hippocampus (hip), dentate gyrus and the hilus. Inset in the right upper corner shows a high magnification image of the suprapyramidal blade (SB) of the granular layer. Protein kinase C α could be expressed in both neural stem cells of the subgranular layer, and granular cells co-expressing neuronal nuclei NeuN; arrows). The infrapyramidal blade (IB) of the granular layer is labeled as well. Scale bar: 200 μm.
[Figure 8] Expression of protein kinase C (PKC) α in the hippocampus of different genotypic pups at postnatal day 7.(A) Western blot analysis of the relative expression of PKC α in the hippocampus of pups. (B) Semi-quantitative western blot analysis of PKC α expression in the hippocampus of pups. PKC α expression levels are expressed as mean ± SD in five pups per group. One-way analysis of variance followed by Student-Newman-Keuls test was performed. aP < 0.01, vs. control group; bP < 0.05, vs. M-EtOH group; cP < 0.05, vs. WT group.WT: Wild-type; SMS2-/-: sphingomyelin synthase 2 knockout; WT/C: control group of WT; WT/M: WT moderate-dose alcohol group; WT/H: WT high-dose group; SMS2-/-/C: control group of SMS2-/-; SMS2-/-/M: moderate-dose alcohol group of SMS2-/-; SMS2-/-/H: high-dose group of SMS2-/-; M-EtOH: moderate-dose alcohol treatment (2.0 g/kg per day); H-EtOH: high-dose alcohol (4.0 g/kg per day).
DISCUSSION
As an effective toxin to the developing nervous system, prenatal alcohol exposure in utero can cause wide-ranging neuropathological damage, and even fetal alcohol syndrome, which seriously affects the development of multiple organ systems in the fetus, especially the central nervous system[[26],[27]]. The influences of alcohol on the central nervous system mainly occur through certain neurotransmitter receptors[[28]], such as the glutamate (GluR2/3, NMDAR), γ-aminobutyric acid, neural acetylcholine[[29]], and the dopamine receptors[[30]]. Alcohol can also activate downstream signal pathways, such as the c-Jun N-terminal kinase pathway[[31]] to induce cell apoptosis[[32]].
Recently, Saito et al[[33]] have found that alcohol could induce apoptosis of nerve cells through the de novo synthesis of ceramide, eventually leading to pathological neurodegeneration and compensatory neural proliferation. In the ceramide pathway, ceramide-1-phosphate (produced through the phosphorylation of ceramide by ceramide kinase) and extracellular signal-regulated protein kinase are key elements to induce cell apoptosis and compensatory proliferation[[25]]. But it is not yet certain whether alcohol-induced neural proliferation was specifically achieved through the ceramide pathway. In this study we constructed a prenatal alcohol exposure model using SMS2 knockout pups and sought to prove our hypothesis that alcohol-induced ceramide accumulation would upregulate neural proliferation in the hippocampus.
Decreased blood sphingomyelin levels in both wild-type and SMS2 knockout pups following prenatal alcohol exposure
Since the SMS2 gene was deleted, ceramide could not be metabolized into sphingomyelin smoothly and would subsequently accumulate in cells, tissues and blood[[34]]. However, the involvement of ceramide in ethanol toxicity is not clear. To study this involvement, blood sphingomyelin level which indirectly reflects the level of ceramide was measured in different alcohol treatment groups of both SMS2 knockout and wild-type pups at P0. Our findings showed that blood sphingomyelin level of SMS2 knockouts was significantly lower than that of wild-type pups, and alcohol could reduce blood sphingomyelin level in a dose-dependent manner. Generally, the level of sphingomyelin is closely related to ceramide metabolism, because the synthesis of sphingomyelin requires the participation of SMS. In mammals, SMS is present as two isozymes: SMS1 (mainly concentrated in golgi bodies) and SMS2 (mainly distributed to cell membrane)[35],[36]]. Ceramide is mainly transported extracellularly for SMS2-catalyzed synthesis of sphingomyelin. There is evidence that the loss or inactivation of SMS2 can block the synthesis of sphingomyelin, causing a decrease sphingomyelin level in blood and the accumulation of intracellular ceramide, which will cause different cellular effects through downstream signaling cascade reactions[[22],[37]]. The alcohol-induced decrease of sphingomyelin levels suggested that ceramide was involved in the mechanisms of alcohol toxicity.
Neural proliferation caused by alcohol-induced ceramide accumulation
The regulatory mechanisms of neural proliferation during development are not understood well. Signal pathways thought to be involved in neural proliferation include MAPK[[38]], Wnt[[39]], tumor growth factor-β[[40]] and Shh[[41]]. Recently, the ceramide-1-phosphate pathway was shown to be involved in the upregulation of neural proliferation[[42],[43]]. In this study, we wanted to know specifically whether the ceramide/ceramide-1-phosphate pathway participated in the upregulation of alcohol-induced neural proliferation. SMS2 knockout mice served as a good model to study the relationships between alcohol-induced neural proliferation and ceramide modulation. During adulthood, only some areas in the central nervous system, such as the subventricular zone, the subgranular layer of the dentate gyrus and the rostral migratory stream in the olfactory bulb retain the capacity of cell proliferation[[44]]. In the present study, the dentate gyrus was used as a target to study the alterations in neural proliferation after alcohol exposure. Our studies indicated that alcohol could induce compensatory neural proliferation and the formation of newborn neurons in a long-term, dose dependent manner. We found that neural proliferation of the alcohol-treated groups, for both the wild-type and SMS2 knockouts, were higher than the control groups. Neural proliferation and the number of newborn neurons increased in a dose-dependent manner with higher proliferation resulting from higher dosage of prenatal alcohol exposure. Neural proliferation was probably the result of a compensatory mechanism for neuroapoptosis induced by alcohol[[45]]. The proliferative stem cells responded to this apoptosis with mitotic divisions and modulations of the cell cycle, thus causing compensatory neural proliferation. The differentiated polymorphic neurons would migrate into the hippocampus or the cortex, to replace some apoptotic cells and maintain functional completeness of the nervous system[[32],[46]]. To further understand how alcohol-induced ceramide accumulation upregulated neural proliferation and to explore the neurotoxicological functions and mechanisms of alcohol toxicity, neural proliferation in SMS2 knockout pups after prenatal alcohol exposure was observed. In our experiments, neural proliferation in the dentate gyrus of SMS2 knockout pups was significantly higher than the wild-type pups at the same age, resulting in more newborn neurons. This is probably due to the accumulation of intracellular ceramide following the loss of the SMS2 gene leading to subsequent compensatory cell proliferation. As one type of biologically active sphingolipid, ceramide is important for signal transduction functions in most cells. The ceramide/ceramide-1-phosphate pathway mainly enhanced cell mitosis and cell migration[[47],[48]] through the activation of phosphatidylinositol 3 kinase/protein kinase B and protein kinase C α signaling[[49]], which is most likely the case in our experiments.
Activation of protein kinase C α in alcohol-induced neural proliferation
Ceramide is a bioactive sphingolipid that regulates critical biological functions, including apoptosis and cell proliferation[[50],[51]]. Neural proliferation is mainly achieved through its phosphorylated metabolite, ceramide-1-phosphate. In this process, protein kinase C α plays an important role in the activity of ceramide/ceramide-1-phosphate signaling[[52]]. First, the phosphorylation of protein kinase C α needs to be activated by SMS, and then the phosphorylated protein kinase C α can subsequently enhance the activation of phosphatidylinositol 3-kinase/protein kinase B for the downstream modulation of ceramide-1-phosphate activity directly. Eventually, ceramide-1-phosphate would cause the influx of Ca2+ into cells to stimulate cell apoptosis and proliferation[[25],[53],[54]]. Therefore, a major factor leading to the stimulation of cell proliferation is the activation of protein kinase C α isoforms, whose activation requires their recruitment to the cell membranes[[55]]. Ceramide-1-phosphate was able to induce the phosphorylation of this important site of protein kinase C α. In this way, measuring protein kinase C α expression would be helpful to understand how ceramide regulates alcohol-induced neural proliferation. Immunocytochemistry showed that protein kinase C α was expressed in most nerve cells, especially in proliferative cells, suggesting that protein kinase C α is necessary for neuronal activity and neural proliferation. The semi-quantitative analysis indicated that the expression level of protein kinase C α in the hippocampal neurons of SMS2 knockout pups was relatively low. This is because the lack of SMS in SMS2 knockout pups could decrease the activity of protein kinase C α, thereby slowing the ceramide/ceramide-1-phosphate cascade process[[25]]. By contrast, alcohol could induce the expression of protein kinase C α in a dose-dependent manner. This suggests that alcohol-induced accumulation of ceramide could upregulate neural proliferation, and that protein kinase C α phosphorylation is a key step for the activation of the ceramide pathway.
In summary, serum sphingomyelin level in SMS2 knockout pups was significantly lower than that of wild-types. Alcohol exposure could decrease sphingomyelin levels in both wild-type and SMS2 knockouts in a dose-dependent manner. Interestingly, SMS2 knockouts revealed a greater amount of neural proliferation in the dentate gyrus compared with wild-type pups, and alcohol could further induce neural proliferation in a long-term and dose-dependent manner. Protein kinase C α phosphorylation was probably a key step for the activation of the ceramide pathway. Alcohol-induced expression of protein kinase C α could activate the ceramide-1-phosphate signaling cascade to enhance neural proliferation and the formation of newborn neurons.
MATERIALS AND METHODS
Design
A randomized, controlled animal experiment.
Time and setting
The experiment was performed at the Institute of Neurobiology, College of Life Science, Henan University, China, between March 2009 and March 2011.
Materials
SMS2 heterozygous mice (one male and two female, average weight 21.2 ± 1.7 g, clean level, animal license number: NSJXC-79J14537) were donated by Professor Xiancheng Jiang from the State University of New York Downstate Medical Center, USA. Ten male and twenty female wild-type (C57BL/6J) mice (average weight 20.7 ± 2.1 g, clean level, animal license No. HNKF36790131) were provided by the Experimental Animal Center of Henan University, China. A total of 360 pups from P0–P30 were used. Offspring were produced from the heterozygotes, and their genotypes were identified by PCR assay. Primer information is listed in. SMS2 genotype identification was then performed as previously described[[56]]. Two possible bands could be identified at 250 and 145 (bp) on the gel image. The wild-type sample had only the 145 (bp) band, while the heterozygous sample had both 250- and 145-bp bands. The knockout sample had only the 250 bp band. All experiments were conducted in accordance with the Guidance Suggestions for the Care and Use of Laboratory Animals, issued by the Ministry of Science and Technology of China[[57]], to ensure animal welfare during experiments.
[Table 1] The information about primers
Methods
Establishment of the alcohol exposure model
SMS2 knockout females were checked each morning for the presence of a vaginal plug; a positive plug was defined as E0. Embryonic (E = day of conception, E0 = vaginal plug found in mated females) or postnatal offspring (P = postnatal days, P0 = the first 24 hours after birth) were produced from timed pregnancies. Postnatal pups were genotyped using PCR, and the wild-type and SMS2 knockout pups were assigned to one of two groups. (1) Alcohol treatment groups: females received a daily intragastric gavage of 25% (w/v) alcohol [Kaifeng Civilized (Group) Co., Ltd., Henan Province, China] at a dose of either 2.0 g/kg or 4.0 g/kg, which began at E5 and continued to parturition. To allow the stomach to empty and to facilitate the absorption of alcohol, food access was removed 2 hours before alcohol dosing. Animals were weighed and gavaged at the same time each day. (2) Control groups: chow fed females were allowed free access to food and water except in the morning when food was removed for 2 hours prior to weighing. The pair fed females were intubated with the same quantity of isocaloric and isovolumetric maltose-dextrin as the alcohol treatment groups. Because there were no significant differences in neural proliferation between the pair fed and chow fed groups, the data from the two control groups were combined to increase the statistical power of the analysis[[9]]. To understand the long-term effect of ethanol toxicity on the development of the central nervous system, pups were harvested at P0, P7, P14 and P30. At least 15 pups were used in each treatment group at P0, P7, P14, and P30, with a total of 230 pups used. Ten pups of each group were used for immunohistochemistry, and five pups were used for western blot analysis.
Measurements of blood alcohol concentration
On E10 and E15, the tail tips of pregnant mice treated with alcohol were nicked with scissors at 60 and 120 minutes after gavage, and 10 μL of blood was collected into heparinized hematocrit tubes. Approximately 5 μL serum was prepared by centrifugation and blood alcohol concentration was determined using an ANALOX model GL5 Analyser (Germany). Four blood alcohol concentration values were obtained from each dam (two readings on E10 and E15, respectively). The highest of the four values was defined as the peak blood alcohol concentration value of the dam and her offspring. The peak blood alcohol concentration value of 4.0 g/kg group was in the range of 30–50 mg/L, while that of 2 g/kg group was in the range of 15–30 mg/L. These groups were accordingly designated as the high and moderate alcohol treatment groups.
Measurements of serum sphingomyelin levels
P0 pups of each group were obtained, and blood was collected through cardiac puncture. The sphingomyelin level was analyzed using enzymatic assays. Briefly, the SMase hydrolyzed sphingomyelin was added into the phosphorylcholine and n-acylsphingosine, then, alkaline phosphatase generated choline from phosphorylcholine. Choline could then be used to generate hydrogen peroxide in a reaction catalyzed by choline oxidase. Finally hydrogen peroxide was used together with DAOS, 4-aminoantipyrine, and peroxidase, as a catalyst, to generate a blue dye with an optimal absorption at 595 nm. The concentration of blue dye, which reflected the concentration of sphingomyelin, was analyzed using a Bio-Rad 680 automatic microplate reader (Bio-Rad, CA, USA) at 595 nm. The sphingomyelin concentration in blood could be calculated based on the standard curve[[58]].
Immunohistochemistry
Pups at various ages (P0, P7, P14, P30) were anesthetized with sodium pentobarbital (20 mg/kg) and perfused transcardially with 4% paraformaldehyde. Brains were removed, and immersion-fixation was continued at 4°C for 1–2 days. Coronal brain slices (50 μm) were then sectioned using a vibratome. After washing three times with 0.1 mol/L PBS, slices were treated with different primary antibodies and the corresponding secondary antibodies for immunofluorescent single- or double-labeling. The primary antibodies include rabbit anti-doublecortin (DCX) polyclonal antibody (1:200; #4604, Cell Signaling, Danver, MA, USA), rabbit anti-Prox 1 polyclonal antibody (1:400; AB5475, Chemicon, Billerica, MA, USA), rabbit anti-protein kinase C α polyclonal antibody (1:50; P4334, Sigma, St. Louis, MO, USA), and mouse monoclonal anti-NeuN antibody (1:500; MAB377B, Millipore, Billerica, MA, USA). After incubating at 4°C overnight, slices were washed three times with 0.1 mol/L PBS and then treated with secondary antibodies. The corresponding secondary antibodies were Alexa Fluor 568 donkey anti-mouse IgG (1:300; A10037, Invitrogen Life Technologies, Carlsbad, CA, USA) and Alexa Fluor 488 donkey anti-rabbit IgG (1:600; Invitrogen, A21206). After incubating with secondary antibody at room temperature for 3 hours, slices were washed three times with 0.1 mol/L PBS and were subsequently mounted with fluorescent mounting medium. With the excitation of rhodamine and FITC (Jianglaibio, Shanghai, China), the slices were observed and photographed under a fluorescent microscope (BX61, Olympus, Tokyo, Japan). Well labeled slices were then observed and imaged under laser scanning confocal microscopy (Olympus).
BrdU immunofluorescent labeling
BrdU (Sigma, B-5002) dissolved at sterile 7 mol/L in NaOH (7.5 mg/500 μL) in physiological saline was injected at a dose of 5 μg/mg into pregnant mice at E18 or into pups at P6, P13 and P29. To intensify labeling, a second identical dose was given 3 hours later. Animals were sacrificed 1 day after injection (at P0, P7, P14 and P30 respectively). The postnatal pups were transcardially perfused and brains were post-fixed in 4% paraformaldehyde in 0.1 mol/L PBS overnight at 4°C. After fixation, coronal brain sections (50 μm) were cut using a vibratome. To visualize nuclear BrdU incorporation, sections were incubated in 2 mol/L HCl at 37°C for 30 minutes to denature the DNA. Sections were then thoroughly rinsed with boric acid (pH 8.0) three times (10 minutes each time) and then with 0.1 mol/L PBS three times (10 minutes each time) to fully remove HCl within the slices. To visualize the neural proliferative cells and granular cells in the dentate gyrus, rabbit anti-Prox 1 polyclonal antibody (1:400; Chemicon, AB5475) and mouse anti-BrdU monoclonal antibody (1:100; ZSGB-BIO, Beijing, China, ZM-0013) were used for immunofluorescent double-labeling. The slices were incubated with all primary antibodies at 4°C overnight. After washing three times with 0.1 mol/L PBS, slices were incubated with secondary antibodies for 3 hours at room temperature. The secondary antibodies were Alexa Fluor 568 donkey anti-mouse IgG (1:300; Invitrogen, A10037) and Alexa Fluor 488 donkey anti-rabbit IgG (1:600; Invitrogen, A21206). After rinsing and mounting, the laser confocal microscope (Olympus, FV10) was used to observe and photograph Prox 1 and BrdU positive cells.
Western blot analysis of protein kinase C α expression in the hippocampi
Because ceramide possibly affects neural proliferation through ceramide-1-phosphate and protein kinase C α, we analyzed the level of protein kinase C α of the ceramide-1-phosphate loop using western blot analysis. P7 pups of each group were collected and sacrificed by cervical dislocation. After opening the skull, brains were quickly dissected away. The hippocampi were separated for homogenization. The cytoplasmic proteins of hippocampus were extracted using a protein extraction kit (Shanghai Beyotime Institute of Biotechnology, Shanghai, China, P0027). The collected cytoplasmic proteins were quantified using the Coomassie brilliant blue G250 method[[59]]. After standard curves were created, the sample protein concentration in different treatment groups was measured. Then, the proteins were electrophorized and subsequently electro-transfer red onto membranes. Rabbit anti protein kinase C α polyclonal antibody (1:20 000; Sigma, P4334) was then added to the membrane and incubated at 4°C overnight. Horseradish peroxidase-labeled goat anti-rabbit IgG secondary antibody (1:1 000; Beijing Solarbio Science & Technology) was then added to the membrane for a 2-hour incubation at room temperature. Prior to each incubation, PBS washing was performed three times, for 5 minutes each. After the enhanced chemiluminescence reagent (ZSGB-BIO, Beijing) reaction for 2 minutes, the protein bands were exposed to film for subsequent development and analysis. β-actin, detected with mouse anti monoclonal antibody (1:1 000; Shanghai Beyotime Institute of Biotechnology, Shanghai, China, #AA128), was used as the internal control. The absorbance ratio of positive bands to internal standard (β-actin) bands represented the relative expression level of the protein of interest.
Data analysis
Therefore, we chose to observe the dentate gyrus as our target area[[24]]. The measurement area of the dentate gyrus was targeted to the suprapyramidal blade, the part of the granule cell layer that is close to hippocampal area CA1 and located above the pyramidal cell layer in CA3. The number of proliferated cells (BrdU positive cells) and newborn neurons (doublecortin positive cells) in suprapyramidal blade (cells/mm2, 20 × magnification, from the sub-granular layer to the super-granular layer) was measured and analyzed statistically by ImageJ software (NIH, Bethesda, MD, USA).
Statistical analysis
Data were represented as mean ± SD. SPSS 17.0 software (SPSS, Chicago, IL, USA) was used for statistical analysis. One-way analysis of variance followed by Student-Newman-Keuls test and/or least significant difference test was performed. P < 0.05 was accepted for an indication of statistically significant difference.
Research background: Extensive clinical data and animal experiments have demonstrated that prenatal alcohol exposure, especially early pregnancy, can cause stillbirth, miscarriage, premature birth, and even lead to fetal alcohol syndrome. The neurotoxicology mechanisms of alcohol are not very clear.
Research frontiers: Some scholars believe that alcohol can induce cellular apoptosis by influencing the activity of Na+/K+-ATPase and adenylate cyclase, as well as by modulating the opening of Ca+ channels. The pharmacological and toxicological mechanisms of alcohol may, however, depend on ceramide.
Clinical significance: Our findings may provide evidence for the pathogenesis of fetal alcohol syndrome and can help further in the understanding of the physiological function and significance of ceramide in nerve cells.
Academic terminology: Fetal alcohol syndrome: Mothers consuming large amounts of alcohol during pregnancy cause growth and development disorders in the fetus and postnatally.
Peer review: This is the first study in China to use SMS2 knockout mice to establish an alcohol exposure model. This model enriches the knowledge surrounding the toxicological mechanisms of alcohol, and further reveals a new strategy for the prevention and treatment of fetal alcohol syndrome.