Identification of Valid Housekeeping Genes for Real-Time Quantitative PCR Analysis of Collapsed Lung Tissues of Neonatal Somatic Cell Nuclear Transfer-Derived Cattle

Identification of valid housekeeping genes for real-time quantitative PCR analysis of collapsed lung tissues of neonatal somatic cell nuclear transfer-derived cattle

Liu, Yan; Zhang, Yunhai; Jiang, Qiuling; Rao, Man; Sheng, Zheya; Zhang, Yu; Du, Weihua; Hao, Haisheng; Zhao, Xueming; Xu, Zhe; Liu, Jianning; Zhu, Huabin

Published in: Cellular Reprogramming

DOI: 10.1089/cell.2015.0024

Publication date: 2015

Document version Publisher's PDF, also known as Version of record

Citation for published version (APA): Liu, Y., Zhang, Y., Jiang, Q., Rao, M., Sheng, Z., Zhang, Y., Du, W., Hao, H., Zhao, X., Xu, Z., Liu, J., & Zhu, H. (2015). Identification of valid housekeeping genes for real-time quantitative PCR analysis of collapsed lung tissues of neonatal somatic cell nuclear transfer-derived cattle. Cellular Reprogramming, 17(5), 360-367. https://doi.org/10.1089/cell.2015.0024

Download date: 26. sep.. 2021 CELLULAR REPROGRAMMING Volume 17, Number 5, 2015 ª Mary Ann Liebert, Inc. DOI: 10.1089/cell.2015.0024

Identiﬁcation of Valid Housekeeping Genes for Real-Time Quantitative PCR Analysis of Collapsed Lung Tissues of Neonatal Somatic Cell Nuclear Transfer–Derived Cattle

Yan Liu,1,4 Yunhai Zhang,2,4 Qiuling Jiang,1 Man Rao,3 Zheya Sheng,3 Yu Zhang,2 Weihua Du,1 Haisheng Hao,1 Xueming Zhao,1 Zhe Xu,1 Jianning Liu,1 and Huabin Zhu1

Abstract

Cloned calves produced by somatic cell nuclear transfer frequently suffer alveolar collapse as newborns. To study the underlying pathophysiological mechanisms responsible for this phenomenon, the expression proﬁles of numerous genes involved in lung development need to be investigated. Quantitative real-time PCR is commonly adopted in gene expression analysis. However, selection of an appropriate reference gene for normalization is critical for obtaining reliable and accurate results. Seven housekeeping genes—b-glucuronidase (GUSB), phosphoglycerate kinase 1 (PGK1), b-2-microglobolin (B2M), peptidylprolyl isomerase A (PPIA), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), TATA-box binding protein (TBP), and 5.8S ribosomal RNA (5.8S rRNA)—were selected and evaluated as candidates. Their gene expression levels in the collapsed lungs of deceased neonate cloned calves and normal lung derived from normal calves were assessed. The ranking of gene expression stability was estimated by the geNorm, NormFinder, and BestKeeper programs. 5.8S rRNA and PPIA were determined to be the most stable reference genes by geNorm and BestKeeper, whereas the combination of GAPDH and TBP was suggested as reference genes by NormFinder. Taking these results into account, we conclude that 5.8S rRNA and PPIA could be the most reliable reference genes for studying the genes involved in alveolar collapse. Moreover, 5.8S rRNA could be represented as a uniform reference gene in similar cases.

Introduction alveolar wall shows deteleotasis under microscopy, and consolidation of lung tissue can be found in dissection. omatic cell nuclear transfer (SCNT) is a powerful In humans, a similar condition called neonatal respiratory Smethod for producing transgenic animals from the same distress syndrome occurs in premature infants and is caused genetic background. The technique involves in vitro manip- by developmental insufficiency of surfactant production ulation and embryonic culture that may lead to inadequate and structural immaturity in the lungs (Avery and Mead, nuclei reprogramming and therefore result in abnormal gene 1959). expression in embryos and animals (Everts et al., 2008; There have been several approaches for investigating Jafarpour et al., 2011; Lin et al., 2008). Developmental de- mechanisms of alveoli collapse in neonatal cloned cattle, ficiency leads to a low survival rate of newborns, especially in most of which focus on imprinted genes and the methylation cloned animals. Respiratory failure resulting from alveoli status of certain transcription factors (Lin et al., 2008). collapse is a high-frequency event and one major cause of However, the underlying pathophysiological mechanisms death in newborn cloned cattle (Hill et al., 1999; Lanza et al., that account for this condition are yet to be elucidated. 2001). Alveoli collapse in neonatal cloned cattle is a condi- Numerous genes are involved in lung development, thus tion accompanied by symptoms such as ventilation defi- identification of the genes responsible for lung collapse in ciency and disturbance of air exchange in the lung tissue, newborn cattle and evaluation of gene expression profiles of leading to dyspnea and even respiratory failure. The pulmonary target genes in collapsed lung tissues are prerequisites.

1Institute of Animal Sciences, Chinese Academy of Agricultural Sciences, Beijing, 100193, China. 2Anhui Provincial Laboratory of Local Livestock and Poultry Genetical Resource Conservation and Breeding, College of Animal Science and Technology, Anhui Agricultural University, Hefei 230036, China. 3Geno-Ming Bioscience, Beijing, 101101, China. 4These authors contributed equally to this work.

360 REFERENCE GENE SELECTION IN COLLAPSED LUNG 361

Real-time quantitative PCR (qPCR) is a powerful tool for RNA extraction, DNase treatment, monitoring changes in gene expression and for understanding and cDNA synthesis biological and molecular mechanisms due to its sensitivity, Total RNA was extracted from lung tissues using the accuracy, specificity, and broad quantification range. To ac- RNeasy Mini Kit (Qiagen, Hilden, Germany) according to quire accurate and meaningful gene expression quantification, the manufacturer’s protocol with DNase treatment during the magnitude of expression normalized to a reference gene as purification of RNA. RNA concentration and purity were an internal control is required. The reference gene is often determined by measuring the absorbance at A260 and A280 referred to as a housekeeping gene (HKG), which should be using a NanoDrop ND-1000 Spectrophotometer (NanoDrop expressed constitutively at relative constant levels under either Technologies Inc., USA). RNA integrity was checked by different experimental conditions or different disease pro- electrophoresis of 500 ng of RNA on 0.8% agarose ethidium cesses. However, commonly used HKGs cannot be universal bromide–stained gels. A mean ratio of A260/A280 &2.0 is under all experimental conditions; indeed, previous studies generally accepted as pure RNA. have shown that some HKGs can vary considerably either in cDNA was synthesized using the QuantiTect Reverse different cell types or under different disease processes (Bustin, Transcription Kit (Qiagen, Hilden, Germany). Briefly, 2 lg 2000; Suzuki et al., 2000; Thellin et al., 1999; Vandesompele of total RNA was used to produce first-strand cDNA with an et al., 2002; Warrington et al., 2000). Therefore, it is crucial to oligo(dT) primer and random primers. seek stably expressed HKGs through a systematic study to interpret the qPCR data more effectively. Primer design and real-time quantitative PCR Few articles in the literature could be found concerning alveolar collapse in cloned cattle; therefore, comparison Oligonucleotide primer pairs were generated by Primer data of suitable HKGs expressed stably in collapsed lung tissue Express 3.0 (Applied Biosystems, Foster City, CA, USA) for quantification is limited. The aim of this study was to according to the published bovine sequences in GenBank identify candidate reference genes for qPCR in the collapsed and synthesized by the Beijing Invitrogen Company (Beijing, lung tissue of cloned cattle. The expression levels of seven China). The primer efficiency was tested through a serial commonly used HKGs—b-glucuronidase (GUSB), phospho- dilution curves with the equation: PCR efficiency = (10[-1/slope] glycerate kinase 1 (PGK1), b-2-microglobolin (B2M), pepti- - 1) $ 100. dylprolyl isomerase A (PPIA), glyceraldehyde-3-phosphate qPCR amplification was performed using the 7500 Fast dehydrogenase (GAPDH), TATA-box binding protein (TBP), Real-time PCR system (Applied Biosystems, Foster City, and 5.8S ribosomal RNA (5.8S rRNA)—were assessed in the CA, USA) following the manufacturer’s instructions and collapsedlungtissuesofcloned cattle and normal lungs of age- with SYBR Green Real-Time PCR with ROX correction.

matched normal cattle, and the comparison data were analyzed Briefly, 1 lL of 1:10 diluted cDNA, 1 lL of primer pairs through three different statistical methods. (10 pmol/lL for each primer), 7.5 lLof2· Power SYBR Green Real-Time PCR Master Mix (Applied Biosystems, Materials and Methods Foster City), and 5.5 lL of double-distilled water was added to a final reaction volume to 15 lL. The holding conditions for Animals qPCR were as follows: Elimination of carryover contamina- Collapsed lung tissues from four neonatal cloned cattle that tion by uracil N-glycosylase (UNG) at 50C for 2 min and died of respiratory failure and normal lung tissues from four activation of Taq polymerase and template denaturation at age-matched normal cattle were collected for this study. All 95C for 10 min. The cycling conditions included template procedures were performed at the Institute of Animal Sciences, denaturation at 95C for 15 sec and annealing and elongation Chinese Academy of Agricultural Sciences (Beijing, China) at 60C and 72C for 1 min, respectively. To ensure a single and were in accordance with the Guiding Principles for the product for each reaction, melt curve analysis was added after Care and Use of Laboratory Animals. thermocycling ranging from 60Cto95C by temperature increases of 0.5C per cycle. All assays were performed in Tissue collection triplicate with no template control (NTC). Both the neonatal cloned calves died of respiratory failure within 2 days of age and were dissected immediately after Statistical analysis death. Normal calves within 2 days of age were killed as The expression stability of HKGs was calculated by geN- control samples. Lung tissue samples of all calves were orm (http://medgen.ugent.be/*jvdesomp/genorm/index.php), immediately collected and frozen in liquid nitrogen or NormFinder (http://moma.dk/normfinder-software), and Best- soaked in 4% paraformaldehyde (PFA) for later analysis. Keeper software (http://www.gene-quantification.de/bestkeeper The samples for histopathological study were gathered from .html) packages. more than five sites within each lung tissue sample to make sure they represented the entirety. Result Necropsy and histopathological study Alveolar collapse Tissues were autopsied by veterinarians from the Chinese The collapsed lung tissues from deceased neonatal cloned Academy of Agricultural Sciences. Histopathological diag- cattle were dissected, and pathological sections were analyzed nosis was performed in the animal pathogenesis lab in between cloned cattle and age-matched normal controls. The the College of Veterinary Medicine at China Agricultural lung tissue from cloned cattle showed a series pathogenic University. symptoms compared to normal calves (Fig. 1), such as 362 LIU ET AL.

Levels of candidate reference genes mRNA The cycle threshold (Ct) values were plotted directly to characterize the expression profiles of the HKGs (Fig. 2). The median expression level of raw Ct values spanned from 13.39 cycles for B2M to 24.48 cycles for GAPDH (Sup- plementary Table S1). By t-test, all the reference genes observed did not significantly differentially express between collapsed lungs and normal lungs. The corresponding p values ranged from 0.094 for B2M to 0.922 for PGK FIG. 1. Atelectasis of neonatal died cloned calf showed by (Supplementary Table S2; Supplementary Data are available histopathological analysis. (a) (Black dots) nuclei; (light at www.liebertpub.com/cell/). gray dots) cells in which inflammation occurred; (white areas) alveolar space. This photograph demonstrates alveolar Expression stability within HKGs invagination and narrow alveolar space in collapsed lung tissue found in a deceased neonatal cloned calf. (b)(Black To identify the least variable HKGs between collapsed dots) nuclei; (white areas) alveolar space in normal lung lungs from cloned cattle and normal lungs from age- tissue derived from a normal neonatal calf. No inflammation matched normal cattle, three Excel-based programs were (light gray dots) was observed. These results indicate that used—geNorm, NormFinder and BestKeeper. lung tissue derived from a deceased neonatal cloned calf was severely collapsed compared to normal lung. Bar, 100 lm. geNorm analysis. The principle of geNorm is that two ideally internal control genes should main a constant ratio between the genes in all samples (Vandesompele et al., atelectasis, broader alveolar septum, mesenchymocyte hyper- 2002). The geNorm applet provides the average pairwise plasia, inflamed effusion in the alveolar space, generation of variation of a certain HKGs with all other tested HKGs as hyaline membrane, transformation to interstitial pneumonia, the expression stability value (M). The lower M value re- and lymphocyte and neutrophil infiltration as a focus. These flects the higher stability of gene expression. The M values phenotypes indicated that lung tissue derived from deceased of all tested HKGs were below the default limit of 1.5, neonatal cloned calves was severely collapsed compared to showing the high expression stability of tested HKGs in our normal lung tissues. samples. The ranking M values of tested HKGs were as follows: GUSB > PGK > B2M > TBP > GAPDH > PPIA and 5.8S rRNA (Fig. 3). This order indicates that PPIA and 5.8S

HKG selection and primer identification rRNA could be determined as the most two stable HKGs For the purpose of finding suitable endogenous control between the two sets of samples. genes for qPCR data normalization in collapsed lung tissue The optimal number of HKGs for normalization was re- of SCNT-derived cattle, seven commonly used HGKs (B2M, vealed by geNorm software for accurate normalization of GAPDH, GUSB, PPIA, PGK, TBP, and rRNA 5.8S) be- qPCR data. The average pairwise variation Vn/Vn+1 was longing to different functional classes were chosen that calculated for two sequential normalization factors with could significantly reduce the chance that the genes might the cutoff threshold of Vn/Vn+1 = 0.15, below which the in- be co-regulated (Vandesompele et al., 2002) (Table 1). All clusion of an additional control gene (n + 1) would be not of the primers selected showing the single peak of the dis- required (Vandesompele et al., 2002). In our assay groups, sociation curve and each single PCR product were verified the pairwise variation V2/V3 was 0.119, indicating that there by sequencing, which indicated good specificity of primers. is no need to include a third reference gene for normaliza- The primer efficiencies ranged from 99.214% to 101.802% tion. As recommended by geNorm, PPIA and 5.8S rRNA with a correlation coefficient (R2) varying from 0.992 to would be the most reliable reference genes for normalization 0.999 (Table 2). (Fig. 4).

Table 1. Candidate Reference Genes Evaluated in Our Study GenBank Gene symbol Gene name Gene function accession no. B2M b-2-microglobolin b-chain of major histocompatibility complex NM_173893 class I molecules GAPDH Glyceraldehyde-3-phosphate Oxidoreductase in glycolysis and gluconeogenesis NM_001034034 dehydrogenase GUSB b-glucuronidase Hydrolase that degrades glycosaminoglycans NM_001083436 PPIA Peptidylprolyl isomerase A Catalyzes the cis-trans isomerization of proline NM_178320 imidic peptide bonds in oligopeptides, accelerating folding TBP TATA-box binding protein General RNA polymerase II transcription factor NM_001075742 PGK1 Phosphoglycerate kinase 1 Glycolytic enzyme NM_001034299 5.8S rRNA 5.8S ribosomal RNA Component of large ribosomal subunit NR_036643 REFERENCE GENE SELECTION IN COLLAPSED LUNG 363

Table 2. Primer Sequences and Efﬁciency for qPCR Amplicon Efﬁciency Name Forward primer 5¢ to 3¢ Reverse primer 5¢ to 3¢ (bp) Slope R2 (100%) PGK1 GCTGCTGTTCCAAGCATCAA ACATCCTTGCCCAGCAGAGA 155 -3.341 0.998 99.214 GUSB TCCGCAGGGACAAGAATCAC GGGCAATCAGCGTCTTGAAG 105 -3.324 0.999 99.905 PPIA CGCGTCTCTTTTGAGCTGTTT CCACCCTGGCACATAAATCC 140 -3.307 0.996 100.627 TBP TGCACAGGAGCCAAGAGTGA TCACAGCTCCCCACCATGTT 128 -3.28 0.993 101.802 B2M TCCAGCGTCCTCCAAAGATT TCTTCTCCCCATTCTTCAGCAA 135 -3.294 0.992 101.175 GAPDH CGACTTCAACAGCGAC CCCTGTTGCTGTAGCCAAATTC 140 -3.319 0.999 100.131 ACTCAC 5.8S GCGGTGGATCACTCGGCT GACAGGCATAGTCATGTG 129 -3.306 0.996 100.661 rRNA AGGAAC

NormFinder analysis. NormFinder is a Microsoft Excel- BestKeeper index, which is a combined parameter computed based program designed to estimate the HKGs expression from all the candidates. Consequently, the result demon- stability by calculating the expression variation both within strated that the HKGs stability order generated from Best- groups and between groups. It is more suitable for studies Keeper was: PPIA > 5.8S rRNA > PGK > TBP > GAPDH > containing different disease groups. The candidate HKGs GUSB > B2M (Table 4) with less expression variation infer lower stability values, therefore indicating more stablility for normalization. In Discussion collapsed lungs and normal lungs, GAPDH showed the lowest stability value of 0.122 followed by 5.8S rRNA and Lung collapse is one of dominant causes of death in neo- TBP, which were both 0.15 (Fig. 5). The best combination natal SCNT cattle. The measurement of transcript abun- of two genes as candidate references was GAPDH and TBP dance with real-time qPCR is a sensitive and effective (stability value = 0.098) (Table 3). method for providing evidence for interpreting the underlying pathophysiological mechanisms of this condition. A BestKeeper analysis. BestKeeper, Excel-based soft- stably expressed reference gene is vital for the normalization ware, ranks the candidate stabilities by a pairwise correla- of gene expression data from qPCR, because an inappro- tion analysis with all of the HKGs pairs. The geometric priate reference gene could lead to inaccurate calculation of average of the ‘‘best’’ HKGs was calculated by raw Ct gene expression and incorrect interpretation of data. It is of values from qPCR. The seven candidate genes were corre- great importance to validate internal control/reference genes lated in a pairwise manner with each other and with the for normalization prior to gene expression studies (Thellin

FIG. 2. Expression levels of candidate reference genes in collapsed lungs of cloned calves and normal lungs of healthy calves. Raw Ct values of each gene from qPCR are given by boxplot. Data from collapsed lung and normal lung are plotted in black and gray, respectively. Boxes indicate the upper and lower quartiles with medians, and the error bars represent the 90th and 10th percentiles. 364 LIU ET AL.

FIG. 3. geNorm analysis of average expression stability of the M value. The M value was identified as the average pairwise variation of a certain reference gene with all other reference genes; thus, a lower M value reflects a more stable gene. geNorm analysis showed that the most stable genes were PPIA and 5.8S rRNA and the least stable gene was GUSB, comparing gene expression between collapsed lung and normal lung. et al., 1999). This study systematically analyzed stability of The geNorm applet uses the pairwise variation analysis reference genes for normalization of qPCR data in collapsed method to estimate the best combination of two reference lung tissues of newborn cloned cattle with respiratory failure genes based on the geometric mean expression levels. The to provide validated reference genes in similar cases. The principle is that two ideal putative reference genes should data analysis for selecting the most stable HKG was per- display constant expression ratios in all samples whether in formed by three Microsoft Excel-based programs—geNorm, different cell types or different disease types. BestKeeper NormFinder, BestKeeper. These programs were applied software employs the same pairwise idea as geNorm, but widely for reference gene selection (Bae et al., 2015; Mariot BestKeeper adopts a measure of the linear correlation called et al., 2015; Park et al., 2015; Velada et al., 2014). the Pearson correlation coefficient to characterize the pairwise correlation between genes across samples (Pfaffl et al., 2004). The pairwise analysis idea neglects the co-regulation of reference genes in analyzed samples, which would cause sampling errors. Therefore, an integrated parameter named the BestKeeper index was developed, and genes were cor- related pairwise with each other and then with the Best- Keeper index to revise the co-regulation effect. The NormFinder software uses a model-based variance estimation approach to provide a more precise method to seek internal control genes suitable for normalization of qPCR data by calculating the expression variation both

Table 3. Determination of Housekeeping Genes Expression Stability by NormFinder

Gene name Stability value CV GUSB 0.298 0.035 PGK 0.269 0.035 FIG. 4. Optimal number of HKGs for data normalization B2M 0.255 0.027 in qPCR by geNorm. The optimal number of HKGs was PPIA 0.214 0.034 inferred by normalization factors (NF) of n and n + 1genes TBP 0.150 0.017 (Vn/n+1), in which the NF calculation was based on the geo- 5.8S rRNA 0.150 0.021 metric mean of multiple control genes given by geNorm GAPDH 0.122 0.013 software. When Vn/n+1 is below 1.5, then the additional n + 1 GAPDH and TBP 0.098 gene is not necessarily needed. Our data showed that V2/3 was smaller than 0.15, suggesting that two reference genes CV, coefﬁcient of variance (CV = SD/mean · 100%); SD, stan- should be adequately suitable for gene expression in our case. dard deviation. REFERENCE GENE SELECTION IN COLLAPSED LUNG 365

Table 4. Repeated Pairwise Correlation Analysis Table 5. Stability Ranking of Candidate Genes of HKGs and Correlation Analysis of HKGs from Three Software Programs with BestKeeper Index geNorm NormFinder BestKeeper vs. B2M GUSB GAPDH PPIA PGK TBP 5.8S 5.8S rRNA GAPDH PPIA GUSB -0.356 — — — — — — PPIA 5.8S rRNA 5.8S rRNA GAPDH -0.495 0.001 — — — — — GAPDH TBP PGK PPIA 0.000 0.186 0.721 — — — — TBP PPIA TBP PGK -0.261 0.432 0.628 0.551 — — — B2M B2M GAPDH TBP 0.240 0.063 0.370 0.544 0.550 — — PGK PGK GUSB 5.8S rRNA -0.172 0.272 0.803 0.934 0.576 0.372 — GUSB GUSB B2M BestKeeper 0.00 0.48 0.62 0.87 0.82 0.68 0.84 p value 0.992 0.229 0.105 0.005 0.013 0.065 0.009

HKGs, housekeeping genes. derived from patients suffering from chronic obstructive pulmonary disease, and GNB2L1, HPRT1, and RPL32 were inferred to be valid HKGs (Ishii et al., 2006). between groups and in groups (Andersen et al., 2004). In our The reference gene stability of embryos with different data, the coefficient of variance of each gene less than 0.05 culture conditions of in vitro production was still of signifi- confirmed the minor variance of intragroup and intergroup cant interest. For example, GAPDH, HMBS,andEEF1A2 computed by NormFinder. The model-based approach sup- were shown to be the best combination of reference genes for plied by NormFinder is less affected by gene co-regulation normalizing gene expression data between bovine blastocysts and could make up for the deficiency of pairwise method as produced by in vitro fertilization (IVF), intracytoplasmic genes co-regulation. sperm injection (ICSI), and SCNT. Both the culture condi- Many studies revealed that HKG expression could be tions of embryos and embryo production method gave rise to variable and prone to directional shifts induced by experi- the unstable expression patterns of reference genes (Luch- mental conditions, consequently resulting in inaccurate singer et al., 2014). Coincidentally, RPS15, RPS18,and normalization (Dheda et al., 2004; Kumar et al., 2012). GAPDH were considered to be suitable for normalization There was an influx of reference gene selection studies in of real-time PCR data from buffalo oocytes/IVF embryos lung diseases and development (Bouhaddioui et al., 2014; (Kumar et al., 2012). Ishii et al., 2006; Jiang et al., 2009; Krasnov et al., 2011; According to the ranking of tested HKG expression sta-

Kriegova et al., 2008; Mehta et al., 2015; Pinhu et al., 2008; bility, we suggest that 5.8S rRNA and PPIA could be applied Saviozzi et al., 2006; Yin et al., 2010) that indicated that the as reference genes for gene expression analysis in bovine reference genes would not be constantly stably expressed in collapsed lung as well as respiratory diseases resulting from treatment group and control group, or within different de- similar processes. Ultimately, we presented 5.8S rRNA as velopmental stages of the lung. For instance, ﬁve HKGs the uniform reference gene (Table 5). 5.8S rRNA is a well- (TUBA1A, ACTB, GAPDH, 18S rRNA, and HIST4H4) were documented HKG and is the most conserved and most assessed in which TUBA1A was proposed as an appropriate widely used reference gene. Raw expression data of 5.8S reference gene for gene expression analysis during mouse rRNA between collapsed lung and normal lung did not ex- lung development (Mehta et al., 2015). Ten HKGs (ABL1, hibit great differences; indeed, all three software programs ACTB, B2M, GAPDH, GNB2L1, HPRT1, PBGD, RPL21, suggested that 5.8S rRNA was the most stable gene among TBP, and TUBB) were evaluated in alveolar macrophages all of the candidate references gene we tested.

FIG. 5. Ranking of HKGs through gene expression stability calculated by NormFinder software. NormFinder evaluated the reference gene for qPCR normalization through gene expression stability; a gene with a lower stability value was more stable for normalization. The ranking order and corresponding stability value given by NormFinder is shown along the x axis, from which GAPDH and TBP were determined to be the best combination of reference genes for normalization. 366 LIU ET AL.

Atelectasis is a phenotype resulting from pulmonary sur- Dunbar, A.E., 3rd, Wert, S.E., Ikegami, M., Whitsett, J.A., factant deficiency. In humans, several genes have been found Hamvas, A., White, F.V., Piedboeuf, B., Jobin, C., Guttentag, to be associated with lung collapse (Bullard and Nogee, 2007; S., and Nogee, L.M. (2000). Prolonged survival in hereditary Cameron et al., 2005; Dunbar et al., 2000; Klein et al., 1998; surfactant protein B (SP-B) deficiency associated with a novel Nogee et al., 2000; Thomas et al., 2002). However, to in- splicing mutation. Pediatr. Res. 48, 275–282. vestigate the mechanism explaining atelectasis in neonatal Everts, R.E., Chavatte-Palmer, P., Razzak, A., Hue, I., Green, cloned cattle and the impact of nuclear transfer leading to C.A., Oliveira, R., Vignon, X., Rodriguez-Zas, S.L., Tian, lung collapse requires more gene expression data; on the X.C., Yang, X., Renard, J.P., and Lewin, H.A. (2008). other hand, this could provide comparative data to enrich Aberrant gene expression patterns in placentomes are asso- more evidence for human respiratory disease research. ciated with phenotypically normal and abnormal cattle Therefore, the combination of PPIA and 5.8S rRNA could be cloned by somatic cell nuclear transfer. Physiol. Genomics 33, 65–77. recommended as a valuable source for gene expression nor- Hill, J.R., Roussel, A.J., Cibelli, J.B., Edwards, J.F., Hooper, malization in pulmonary surfactant–related research. N.L., Miller, M.W., Thompson, J.A., Looney, C.R., Wes- thusin, M.E., Robl, J.M., and Stice, S.L. (1999). Clinical and Acknowledgments pathologic features of cloned transgenic calves and fetuses This work was supported by the National Natural Science (13 case studies). Theriogenology 51, 1451–1465. Foundation of China (31301977), Chinese Academy of Ishii, T., Wallace, A.M., Zhang, X., Gosselink, J., Abboud, Agricultural Sciences Foundation (2011cj-1), and the R.T., English, J.C., Pare, P.D., and Sandford, A.J. (2006). Agricultural Science and Technology Innovation Program Stability of housekeeping genes in alveolar macrophages (ASTIP-IAS06). We also thank Prof. Jifeng Zhang, Dr. from COPD patients. Eur. Respir .J 27, 300–306. Jinming Huang, Dr. Changfa Wang, and Dr. Hongjun Yang Jafarpour, F., Hosseini, S.M., Hajian, M., Forouzanfar, M., Os- tadhosseini, S., Abedi, P., Gholami, S., Ghaedi, K., Gourabi, H., from the Dairy Cattle Research Center, Shandong Academy Shahverdi, A.H., Vosough, A.D., and Nasr-Esfahani, M.H. of Agricultural Sciences for providing some samples of (2011). Somatic cell-induced hyperacetylation, but not hypo- cloned cattle. methylation, positively and reversibly affects the efficiency of in vitro cloned blastocyst production in cattle. Cell. Reprogram. Author Disclosure Statement 13, 483–493. The authors declare that no conflicting financial interests Jiang, C., Meng, L., Zhu, W., Shahzad, M., Yang, X., and Lu, S. exist. (2009). Housekeeping gene stability in pristane-induced ar- thritis and antigen-induced pulmonary inflammation of rats. Inflamm. Res. 58, 601–609.

References Klein, J.M., Thompson, M.W., Snyder, J.M., George, T.N., Andersen, C.L., Jensen, J.L., and Orntoft, T.F. (2004). Nor- Whitsett, J.A., Bell, E.F., McCray, P.B., Jr., and Nogee, L.M. malization of real-time quantitative reverse transcription-PCR (1998). Transient surfactant protein B deficiency in a term in- data: A model-based variance estimation approach to identify fant with severe respiratory failure. J. Pediatr. 132, 244–248. genes suited for normalization, applied to bladder and colon Krasnov, G.S., Oparina, N., Dmitriev, A.A., Kudriavtsev, A.V., cancer data sets. Cancer Res. 64, 5245–5250. Anedchenko, E.A., Kondrat’eva, T.T., Zabarovskii, E.R., and Avery, M.E., and Mead, J. (1959). Surface properties in relation Senchenko, V.N. (2011). [Novel reference gene RPN1 for to atelectasis and hyaline membrane disease. AMA J. Dis. normalization of quantitative data in lung and kidney cancer]. Child. 97, 517–523. Mol. Biol. (Mosk.) 45, 238–248. Bae, I.S., Chung, K.Y., Yi, J., Kim, T.I., Choi, H.S., Cho, Y.M., Kriegova, E., Arakelyan, A., Fillerova, R., Zatloukal, J., Mra- Choi, I., and Kim, S.H. (2015). Identification of reference zek, F., Navratilova, Z., Kolek, V., du Bois, R.M., and Petrek, genes for relative quantification of circulating microRNAs in M. (2008). PSMB2 and RPL32 are suitable denominators to bovine serum. PLoS One 10, e0122554. normalize gene expression profiles in bronchoalveolar cells. Bouhaddioui, W., Provost, P.R., and Tremblay, Y. (2014). BMC Mol. Biol. 9, 69. Identification of most stable endogenous control genes for Kumar, P., Yadav, P., Verma, A., Singh, D., De, S., and Datta, microRNA quantification in the developing mouse lung. T.K. (2012). Identification of stable reference genes for gene PLoS One 9, e111855. expression studies using quantitative real time PCR in buffalo Bullard, J.E., and Nogee, L.M. (2007). Heterozygosity for oocytes and embryos. Reprod. Domest. Anim. 47, e88–e91. ABCA3 mutations modifies the severity of lung disease as- Lanza, R.P., Cibelli, J.B., Faber, D., Sweeney, R.W., Henderson, sociated with a surfactant protein C gene (SFTPC) mutation. B., Nevala, W., West, M.D., and Wettstein, P.J. (2001). Cloned Pediatr Res. 62, 176–179. cattle can be healthy and normal. Science 294, 1893–1894. Bustin, S.A. (2000). Absolute quantification of mRNA using Lin, L., Li, Q., Zhang, L., Zhao, D., Dai, Y., and Li, N. (2008). real-time reverse transcription polymerase chain reaction as- Aberrant epigenetic changes and gene expression in cloned says. J. Mol. Endocrinol. 25, 169–193. cattle dying around birth. BMC Dev. Biol. 8, 14. Cameron, H.S., Somaschini, M., Carrera, P., Hamvas, A., Luchsinger, C., Arias, M.E., Vargas, T., Paredes, M., Sanchez, Whitsett, J.A., Wert, S.E., Deutsch, G., and Nogee, L.M. R., and Felmer, R. (2014). Stability of reference genes for (2005). A common mutation in the surfactant protein C gene normalization of reverse transcription quantitative real-time associated with lung disease. J. Pediatr. 146, 370–375. PCR (RT-qPCR) data in bovine blastocysts produced by IVF, Dheda, K., Huggett, J.F., Bustin, S.A., Johnson, M.A., Rook, ICSI and SCNT. Zygote 22, 505–512. G., and Zumla, A. (2004). Validation of housekeeping genes Mariot, R.F., de Oliveira, L.A., Voorhuijzen, M.M., Staats, M., for normalizing RNA expression in real-time PCR. Bio- Hutten, R.C., Van Dijk, J.P., Kok, E., and Frazzon, J. (2015). techniques 37, 112–114, 116, 118–119. Selection of reference genes for transcriptional analysis of REFERENCE GENE SELECTION IN COLLAPSED LUNG 367

edible tubers of potato (Solanum tuberosum L.). PLoS One Thomas, A.Q., Lane, K., Phillips, J., 3rd, Prince, M., Markin, 10, e0120854. C., Speer, M., Schwartz, D.A., Gaddipati, R., Marney, A., Mehta, A., Dobersch, S., Dammann, R.H., Bellusci, S., Ilins- Johnson, J., Roberts, R., Haines, J., Stahlman, M., and Loyd, kaya, O.N., Braun, T., and Barreto, G. (2015). Validation of J.E. (2002). Heterozygosity for a surfactant protein C gene Tuba1a as appropriate internal control for normalization of mutation associated with usual interstitial pneumonitis and gene expression analysis during mouse lung development. Int. cellular nonspecific interstitial pneumonitis in one kindred. J. Mol. Sci. 16, 4492–4511. Am. J. Respir. Crit. Care Med. 165, 1322–1328. Nogee, L.M., Wert, S.E., Proffit, S.A., Hull, W.M., and Whit- Vandesompele, J., De Preter, K., Pattyn, F., Poppe, B., Van sett, J.A. (2000). Allelic heterogeneity in hereditary surfactant Roy, N., De Paepe, A., and Speleman, F. (2002). Accurate protein B (SP-B) deficiency. Am. J. Respir. Crit. Care Med. normalization of real-time quantitative RT-PCR data by 161, 973–981. geometric averaging of multiple internal control genes. Park, S.J., Kwon, S.G., Hwang, J.H., Park da, H., Kim, T.W., Genome Biol. 3, RESEARCH0034. Epub 2002 Jun 18. and Kim, C.W. (2015). Selection of appropriate reference Velada, I., Ragonezi, C., Arnholdt-Schmitt, B., and Cardoso, H. genes for RT-qPCR analysis in Berkshire, Duroc, Landrace, (2014). Reference genes selection and normalization of oxi- and Yorkshire pigs. Gene 558, 152–158. dative stress responsive genes upon different temperature Pfaffl, M.W., Tichopad, A., Prgomet, C., and Neuvians, stress conditions in Hypericum perforatum L. PLoS One 9, T.P. (2004). Determination of stable housekeeping genes, e115206. differentially regulated target genes and sample integrity: Warrington, J.A., Nair, A., Mahadevappa, M., and Tsyganskaya, BestKeeper—Excel-based tool using pair-wise correlations. M. (2000). Comparison of human adult and fetal expression and Biotechnol. Lett. 26, 509–515. identification of 535 housekeeping/maintenance genes. Physiol. Pinhu, L., Park, J.E., Yao, W., and Griffiths, M.J. (2008). Re- Genomics 2, 143–147. ference gene selection for real-time polymerase chain reac- Yin, R., Tian, F., Frankenberger, B., de Angelis, M.H., and tion in human lung cells subjected to cyclic mechanical strain. Stoeger, T. (2010). Selection and evaluation of stable Respirology 13, 990–999. housekeeping genes for gene expression normalization in Saviozzi, S., Cordero, F., Lo Iacono, M., Novello, S., Scagliotti, carbon nanoparticle-induced acute pulmonary inflammation G.V., and Calogero, R.A. (2006). Selection of suitable reference in mice. Biochem. Biophys. Res. Commun. 399, 531–536. genes for accurate normalization of gene expression profile studies in non-small cell lung cancer. BMC Cancer 6, 200. Address correspondence to: Suzuki, T., Higgins, P.J., and Crawford, D.R. (2000). Control Huabin Zhu selection for RNA quantitation. Biotechniques 29, 332–337. Institute of Animal Sciences Thellin, O., Zorzi, W., Lakaye, B., De Borman, B., Coumans, Chinese Academy of Agricultural Sciences B., Hennen, G., Grisar, T., Igout, A., and Heinen, E. (1999).

Beijing, 100193, China Housekeeping genes as internal standards: use and limits. J. Biotechnol. 75, 291–295. E-mail: [email protected]