Altered molecular signatures during kidney development after intrauterine growth restriction of different origins. Altered molecular signatures during kidney development after intrauterine growth restriction of different origins

Intrauterine growth restriction (IUGR) has been linked to predisposition towards an unfavorable course of glomerulopathies and early loss of kidney function. This study was performed to identify transcriptional alterations in male IUGR rats during and at the end of nephrogenesis in order to generate hypotheses which molecular mechanisms contribute to adverse kidney programming. IUGR was induced by low protein (LP) diet throughout pregnancy, bilateral uterine vessel ligation (LIG), or intrauterine stress (IUS) by sham operation. Offspring of unimpaired dams served as controls. Significant acute kidney damage was ruled out by negative results for proteins indicative of ER-stress, autophagy, apoptosis or infiltration with macrophages. Renal gene expression was examined by transcriptome microarrays, demonstrating 53 (LP, n=12; LIG, n=32; IUS, n=9) and 134 (LP, n=10; LIG, n=41; IUS, n=83) differentially expressed transcripts on postnatal days (PND) 1 and 7, respectively. Reduced Pilra (all IUGR groups, PND 7), Nupr1 (LP and LIG, PND 7) and Kap (LIG, PND 1) as well as increased Ccl20, S100a8/a9 (LIG, PND 1), Ifna4 and Ltb4r2 (IUS, PND 7) indicated that inflammation-related molecular dysregulation could be a “common” feature after IUGR of different origins. Network analyses of transcripts and predicted upstream regulators hinted at proinflammatory adaptions mainly in LIG (arachidonic acid-binding, neutrophil aggregation, toll-like-receptor, NF-kappa B and TNF signaling) and dysregulation of AMPK and PPAR signaling in LP pups. The latter may increase susceptibility towards obesity-associated kidney damage. Western blots of the most prominent predicted upstream regulators confirmed significant dysregulation of RICTOR in LP (PND 7) and LIG pups (PND 1), suggesting that mTOR-related processes could further modulate kidney programming in these groups of IUGR pups. Overall design: Direct comparison of kidneys during early postnatal life from IUGR offspring after (1) low protein (LP) diet throughout pregnancy, (2) bilateral uterine vessel ligation (LIG) during terminal pregnancy and (3) intrauterine stress (IUS) by sham operation in late pregnancy with a common control (C) group offers the unique opportunity to differentiate the details of early kidney programming after reduced uterine blood flow from alterations induced by intrauterine stress or nutritional deficiency in a critical window of kidney development. The models have been thoroughly validated in our laboratory with a special focus on translational relevance. In detail, Wistar rat dams (Janvier Labs, France) were time-mated [presence of sperm plug = gestational day (GD) 0] and either fed a defined diet (no. C1000, Altromin, Germany) containing 17.6% protein throughout pregnancy in groups LIG, IUS and C, or a diet containing 8.1% protein (no. C1003, Altromin) in group LP as described before [16,19]. In LIG dams (n=19), fetuses were counted intraoperatively after midline laparotomy and careful extra-abdominal positioning of both uterine horns on GD 18 under isoflurane anesthesia and metamizol analgesia. Consecutively, bilateral uterine vessel ligation was performed caudally of the most caudal vessel branch running to the most caudal feto-placental unit. The duration of the complete intervention was 15 – 20 minutes. The size of LIG-litters was 9 – 15 living fetuses, all uterine horns (n=38) contained 4 – 9 living fetuses. After ligation, blood flow to the LIG fetuses is reduced, as blood now is supplied from cranially via the ovarian arteries only. In IUS dams (n=13), the suture material was not fixed but removed after identical procedures. The size of IUS-litters was 10 – 16 living fetuses, all uterine horns (n=26) contained 3 – 10 living fetuses. After surgery, all dams started to drink water (containing tramadol for 24h after surgery) after 5 – 10 minutes, and their behavior recovered fully within 1 hour. Dams of groups C (n=19) and LP (n=12) did not undergo surgery. All dams delivered spontaneously after approximately 22.5 gestational days within a time frame of 12h. Two C litters (litter size was <9, but had to be in the range of 9 – 16 to reduce bias due to varying litter size) and one LIG litter (11 out of 13 fetuses were resorbed after LIG) were excluded on PND 1. Thus, we finally included 17 C litters, 18 LIG litters, 13 IUS litters and 12 LP litters (n=60 litters in total). All offspring [C, n=232; LP, n=152; LIG, n=167 born alive out of n=209 counted intraoperatively, i.e. n=42 (20%) resorbed; IUS, n=174 born alive out of 174 counted intraoperatively) were weighed and measured on PND 1, and sexed visually. The error of this sexing method is ~1% in our laboratory on PND 1. In LIG, only the six smallest pups per litter (n=108 out of 167) were selected since we know from thorough intrauterine evaluation that especially the fetuses near to the ligation suffer from insufficient blood and nutrient supply and display intrauterine growth restriction. As there is sex-specific transcriptional regulation and male rodents seem to be more susceptible to renal programming than females, we focused on male pups in all further analyses (PND 1 male pups: C, n=105; LP, n=70; LIG, n=52 small; IUS, n=99; PND 1 kidneys: C, n=27; LP, n=17; LIG, n=21; IUS, n=24; PND 7 pups and kidneys: C, n=24; LP, n=16; LIG, n=14; IUS, n=17). For PND 1 studies, newborn male rats were narcotized with pentobarbital and decapitated. Then, whole kidneys as well as further organs for studies on brain and metabolism were obtained and weighed. Right kidneys were shock frozen in liquid nitrogen and stored at 80°C, left kidneys dissected longitudinally in the middle, directly fixed in 4% PFA for 24h, transferred in 70% isopropyl alcohol for 24h, and processed for paraffin embedding. For PND 7 studies, postnatal environmental conditions were standardized in all groups on the first day of life, because litter size and postnatal nutrition are known to affect the final stages of renal development. In detail, pups from C, LIG and IUS dams were transferred to non-operated dams from group C receiving control diet (C1000, Altromin) and the size of all foster litters was adjusted to four male and four female pups to ensure similar postnatal nutrition. LP pups (four male, four female) were transferred to LP foster dams as postnatal renal programming effects are in part mediated by altered milk quality in this model. On PND 7, kidneys and further organs were harvested and stored as described above.

宫内生长受限（Intrauterine growth restriction, IUGR）已被证实与肾小球疾病病程不良及肾功能早期丧失的易感性相关。本研究旨在明确雄性宫内生长受限大鼠在肾发生期间及结束时的转录组改变，以阐明不良肾脏编程的潜在分子机制。
IUGR造模方式包括妊娠全程饲喂低蛋白（low protein, LP）日粮、妊娠末期双侧子宫血管结扎（uterine vessel ligation, LIG），或假手术诱导宫内应激（intrauterine stress, IUS）；以未接受干预的孕鼠子代作为对照。
通过检测内质网应激（ER-stress）、自噬（autophagy）、凋亡（apoptosis）及巨噬细胞浸润相关蛋白均为阴性，排除了严重急性肾损伤的可能。
采用转录组微阵列分析肾脏基因表达，结果显示，在出生后第1天（postnatal day, PND1）与出生后第7天（PND7）分别筛选到53个（LP组n=12；LIG组n=32；IUS组n=9）与134个（LP组n=10；LIG组n=41；IUS组n=83）差异表达转录本。
Pilra在所有IUGR组PND7时表达下调，Nupr1在LP与LIG组PND7时表达下调，Kap在LIG组PND1时表达下调；同时Ccl20、S100a8/a9在LIG组PND1时表达上调，Ifna4与Ltb4r2在IUS组PND7时表达上调，提示炎症相关分子调控异常可能是不同病因IUGR后肾脏的共性特征。
对差异转录本及预测上游调控因子的网络分析显示，LIG组主要出现促炎适应通路异常（如花生四烯酸结合、中性粒细胞聚集、Toll样受体、NF-κB及TNF信号通路），而LP组则存在AMPK与PPAR信号通路失调，后者可能增加肥胖相关性肾损伤的易感性。
对最显著的预测上游调控因子进行免疫印迹（Western blot）验证，证实LP组（PND7）与LIG组（PND1）的RICTOR表达存在显著异常，提示mTOR相关过程可进一步调控此类IUGR子代的肾脏编程。
本研究整体设计为：以通用对照组（C）为参照，直接比较妊娠期间（1）饲喂低蛋白日粮、（2）妊娠末期双侧子宫血管结扎、（3）妊娠末期假手术诱导宫内应激的IUGR子代在出生早期的肾脏组织，从而得以区分子宫血流减少与宫内应激或营养缺乏在肾脏发育关键窗口内诱导的早期肾脏编程差异。
本模型已在本实验室完成全面验证，尤其关注其转化医学相关性。具体如下：选用法国Janvier Labs提供的Wistar孕鼠，通过阴道栓确认妊娠第0天（gestational day, GD0）。其中LIG、IUS及C组孕鼠饲喂含17.6%蛋白质的标准日粮（货号C1000，德国Altromin），LP组孕鼠饲喂含8.1%蛋白质的低蛋白日粮（货号C1003，德国Altromin），方法参照既往研究[16,19]。
LIG组孕鼠（n=19）于GD18在异氟烷麻醉及metamizol镇痛下，行正中剖腹术并小心将双侧子宫角移出腹腔以计数胎鼠。随后在连接最尾端胎-胎盘单位的血管分支尾侧行双侧子宫血管结扎，手术全程耗时15~20分钟。LIG组每窝存活胎鼠数为9~15只，双侧子宫角（n=38）每侧含4~9只存活胎鼠。结扎后，LIG组胎鼠仅通过颅侧卵巢动脉获取血流，因此其子宫血流供应减少。
IUS组孕鼠（n=13）行相同手术操作，但仅放置缝线后移除，不进行血管结扎。IUS组每窝存活胎鼠数为10~16只，双侧子宫角（n=26）每侧含3~10只存活胎鼠。术后所有孕鼠均可在5~10分钟内饮用含曲马多的饮水（术后镇痛24小时），并于1小时内恢复正常行为。C组（n=19）与LP组（n=12）孕鼠不接受任何手术操作。所有孕鼠均于妊娠约22.5天时自发分娩，分娩时间跨度为12小时。
PND1时剔除不合格窝别：2窝C组（窝仔数<9，需维持9~16只以减少窝仔数差异带来的偏倚）及1窝LIG组（13只胎鼠中11只被吸收）。最终纳入分析的窝别为：C组17窝、LIG组18窝、IUS组13窝、LP组12窝，共计60窝。所有子代（C组n=232；LP组n=152；LIG组共计数209只胎鼠，出生存活167只，即42只（20%）被吸收；IUS组计数174只胎鼠，全部存活）均于PND1称重、测量体长，并通过外观鉴定性别。本实验室PND1时外观性别鉴定的误差率约为1%。
LIG组仅选取每窝中体型最小的6只幼鼠（n=108/167），因为前期宫内评估显示，靠近结扎处的胎鼠更易出现血流及营养供应不足，从而表现出宫内生长受限。由于存在性别特异性转录调控，且雄性啮齿类动物对肾脏编程的易感性高于雌性，因此后续分析仅纳入雄性幼鼠：PND1雄性幼鼠：C组n=105；LP组n=70；LIG组n=52（小型幼鼠）；IUS组n=99；PND1肾脏样本：C组n=27；LP组n=17；LIG组n=21；IUS组n=24；PND7幼鼠及肾脏样本：C组n=24；LP组n=16；LIG组n=14；IUS组n=17。
PND1样本采集：新生雄性大鼠经戊巴比妥麻醉后断头处死，摘取全肾及其他用于脑和代谢研究的器官并称重。右肾置于液氮中速冻后于-80℃保存；左肾沿纵轴正中剖开，立即置于4%多聚甲醛（PFA）中固定24小时，随后转移至70%异丙醇中24小时，之后进行石蜡包埋。
PND7样本采集：出生首日即统一所有组别的产后环境，因为窝仔数及产后营养可影响肾脏发育的终末阶段。具体操作：将C组、LIG组及IUS组的幼鼠转移至饲喂标准日粮（C1000，德国Altromin）的未手术C组孕鼠处寄养，并将每窝寄养幼鼠调整为4雄4雌，以保证产后营养水平一致。LP组幼鼠（4雄4雌）则寄养至饲喂低蛋白日粮的LP组孕鼠处，因为该模型中产后肾脏编程效应部分受乳汁质量改变的影响。于PND7时采集肾脏及其他器官，保存方式同上。