Reproductive and Developmental Medicine

: 2020  |  Volume : 4  |  Issue : 4  |  Page : 212--218

Selection and evaluation of optimal reference genes for quantitative reverse transcription-polymerase chain reaction analyses of gene expression in human spermatozoa

Chun-Hai Luo1, Yun-Ge Tang2, Shi-Hao Hong1, Yuan Tang2, Ying Zhang3, Fei Sun1,  
1 Institute of Reproductive Medicine, School of Medicine, Nantong University, Nantong 226001, China, China
2 Key Laboratory of Male Reproductive and Genetics, National Health and Family Planning Commission, Guangzhou 510060, China
3 Institute of Reproductive Medicine, School of Medicine,Nantong University, Nantong 226001; Key Laboratory of Male Reproductive and Genetics, National Health and Family Planning Commission, Guangzhou 510060, China, Chinaa

Correspondence Address:
Fei Sun
Institute of Reproductive Medicine, School of Medicine, Nantong University, No. 19, Qixiu Road, Nantong, Jiangsu 226001
Ying Zhang
Key Laboratory of Male Reproductive and Genetics, National Health and Family Planning Commission, No. 17, Meidong Road, Guangzhou, Guangdong 510060


Objective: Optimal reference genes are critical for accurate normalization and reliable interpretation of gene expression quantification data. Recently, several strategies have been utilized for validating reference genes in different human tissues. However, no universal reference genes have been described that accurately summarize transcriptional activity in human spermatozoa. Methods: Using quantitative reverse transcription-polymerase chain reaction (RT-qPCR), we evaluated ten commonly used candidate reference genes between two groups of human cryopreserved donor sperm with different pregnancy rates. We assessed the stability of reference genes using three different algorithms, namely geNorm, NormFinder, and BestKeeper. We then identified the most stable reference genes. Results: Male-enhanced antigen 1 (MEA1) was identified as the most stably expressed reference gene, followed by testis-enhanced gene transcript (TEGT). Conclusions: We comprehensively identified MEA1 and TEGT as the most stably expressed reference genes for the normalization of gene expression data in human spermatozoa.

How to cite this article:
Luo CH, Tang YG, Hong SH, Tang Y, Zhang Y, Sun F. Selection and evaluation of optimal reference genes for quantitative reverse transcription-polymerase chain reaction analyses of gene expression in human spermatozoa.Reprod Dev Med 2020;4:212-218

How to cite this URL:
Luo CH, Tang YG, Hong SH, Tang Y, Zhang Y, Sun F. Selection and evaluation of optimal reference genes for quantitative reverse transcription-polymerase chain reaction analyses of gene expression in human spermatozoa. Reprod Dev Med [serial online] 2020 [cited 2021 Apr 14 ];4:212-218
Available from:

Full Text


Globally, infertility affects an estimated 10%–15% of couples of reproductive age, and male-factor infertility contributes to approximately 40%–50% of overall cases.[1] Male infertility, an alarming public health problem, refers to the inability of a man to impregnate a fertile woman, and it is characterized by alteration in sperm count, concentration, motility, or morphology in at least one of two sperm sample analyses performed between 1 and 4 weeks apart. Upon fertilization, a sperm delivers not only the paternal haploid genome but also specific functional RNAs, which may play crucial roles following fertilization, into the oocyte, thereby influencing early embryo development.[2] Recent studies using next-generation sequencing (NGS) on transcripts that remain within the spermatozoa have provided a comprehensive snapshot of the complex population of sperm-specific RNAs, including both coding and noncoding RNAs, which may serve as potential biomarkers for sperm quality control, particularly for sperm donors enlisted in sperm banks.[3],[4],[5],[6] Moreover, spermatozoal RNAs may provide new insights into the developmental history of each sperm, thereby facilitating the identification of new biomarkers of fertility and pregnancy outcomes. In this context, quantitative reverse transcription-polymerase chain reaction (RT-qPCR) is a powerful tool that has been extensively employed for sperm-specific transcript profiling of RNAs due to its high sensitivity, specificity, and reproducibility.[7],[8] However, multiple factors, including the initial amount, quality, and integrity of RNA and differences in the expression levels of target RNAs between tissues and cells, can significantly influence the quantitative measures obtained from RT-qPCR analyses. Thus, to ensure the accuracy and reproducibility of such results, appropriate stably expressed reference (housekeeping) genes for the analysis of human spermatozoa are required.

Reference genes are internal reaction controls that exhibit sequences different from that of the target, and these are used for the normalization of relative mRNA expression levels by RT-qPCR.[9],[10],[11] Notably, an ideal reference gene must possess the following characteristics: high abundance; stable expression levels in different tissues and cell types; and showing minimal changes due to experimental, biological, or other treatment-related factors.[10] Housekeeping genes are involved in maintaining fundamental biological processes and expressed at constant levels in various tissues and cell types, such as glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and actin (ACTB). Therefore, housekeeping genes are extensively used as reference genes for the normalization of gene expression.[12],[13] However, an increasing number of studies have suggested that the endogenous expression levels of various housekeeping genes may differ under certain physiological and pathological conditions.[14],[15] Conceivably, as the final-stage spermatogenic cell, transcriptional activity is repressed in the mature sperm due to the packaging of chromatin that occurs during DNA compaction in spermiogenesis.[4] Intriguingly, the transcripts retained in sperm represent a complex repertoire of “residual” RNA products expressed in various spermatogenic cells; the retained RNAs are characterized by short length, multiple isoforms, and low abundance.[3],[4],[16] Moreover, housekeeping genes from other tissues may not necessarily meet the prerequisites for an optimal reference gene for sperm; thus, no stable reference genes are known that can normalize gene expression in sperm.

Therefore, in this study, using the ten most widely applied or putative reference genes (ACTB, ATP5B, GAPDH, PRM1, PRM2, DAZAP2, MEA1, SPAG7, SRM, and TEGT), we analyzed gene expression in donor sperm samples with different pregnancy rates that were obtained from a sperm bank.[17] Furthermore, well-established algorithms, including geNorm (qbaseplus software), NormFinder, and BestKeeper, were used to evaluate the stability of the aforementioned ten reference genes.[9],[11],[18] Based on the findings from these algorithms, a comprehensive ranking of all genes was derived to assess their suitability as reference genes.


Sample collection and grouping

Cryopreserved sperm samples were obtained from the human sperm bank of Guangdong located in Guangzhou, China. Sperm were collected from volunteer donors aged 26–42 years according to the 5th edition of the World Health Organization Laboratory Manual Examination and Processing of Human Semen (2010). No sperm donors had a genetic or medical history that might affect semen parameters, such as high fever, malignant tumors, endocrine abnormalities, metabolic diseases or other chronic conditions, autoimmune diseases, sexually transmitted diseases, scrotal trauma, and inflammation of the reproductive organs. Donors who were occupationally exposed to extreme heat conditions and toxic chemical reagents or radioactive materials were excluded from the study. Quality control analyses of donor sperm standard parameters, such as sperm count, motility, and morphology, were strictly conducted daily to ensure consistency; however, some of the samples still demonstrated low fertility or even thorough sterility. Accordingly, sperm samples were divided into the following two groups: one with high pregnancy rate (HPR, 5 pregnancies out of <20 attempts, >25%; n = 8) and the other with low pregnancy rate (LPR, <1 pregnancy out of 30 attempts, <3.3%; n = 8). This study was approved by the Ethics Committee of Nantong University Medicine School (Nantong, China). All samples were procured with written informed consent from the parents.

RNA isolation and complementary DNA synthesis

Previous studies have indicated that in mature spermatozoa, the sperm head is highly condensed, making it challenging to lyse the sperm heads completely.[4] Thus, we adopted a slightly modified protocol with a strong lysate to completely lyse the sperm head and release RNA.[19] First, 300 μL of cryopreserved donor sperm samples was incubated at 37°C for 10 min. Spermatozoa were then selected and separated from the somatic cells by centrifuging thawed semen through a two-layer (1-mL 45% and 1-mL 90%) gradient of Percoll at ×400 g for 20 min. The fraction in the 90% layer was washed with 1 mL of Tyrode, and the tube was centrifuged at ×400 g for 10 min. The pellet was collected with extreme care to avoid any somatic cell contamination and subjected to the mini-swim-up modified technique. Total RNA was extracted from the spermatozoa isolated by mini-swim-up using the mirVana RNA Isolation Kit (Thermo Scientific, Waltham, USA) following the manufacturer's protocol with slight modifications. The sperm pellet was resuspended in 500 μL of lysis buffer, followed by 15 min of incubation at 70°C to lyse the condensed sperm head completely. The lysate was then passed through the first column to remove genomic DNA, and the filtrate was saved for RNA purification. After phenol extraction, RNA was isolated from the aqueous phase using the mirVana Kit according to the manufacturer's instructions. RNA was recovered as total RNA and cleaned using an ethanol precipitation protocol.[19] Total RNA was eluted from the membrane of the column and resuspended in 30 μL of nuclease-free water. The quantity and quality of the purified sperm RNA samples were assessed using a NanoDrop 2000 Spectrophotometer (Thermo Fisher Scientific) and an Agilent 2100 Electrophoresis Bioanalyzer (Agilent Technologies, Santa Clara, USA).

For first-strand complementary DNA (cDNA) synthesis, 500 ng of purified RNA from each sample was reverse transcribed to cDNA using the Prime Script RT-PCR kit (Takara, Otsu, Japan) according to the manufacturer's protocol.

Selection of candidate reference genes and primer design

Ten candidate reference genes were selected for this study, and the National Center for Biotechnology Information Primer-BLAST software ( = BlastHome) was used to design RT-qPCR assay primers. All primers were carefully designed to hybridize to exon–exon junctions to distinguish and quantify the expression of a specific gene using the following parameters: target length between 90 bp and 160 bp, primer size between 18 bp and 22 bp, and melting temperature between 57°C and 63°C [Table 1]. Potential secondary structures of the primers were also examined as previously reported.[20] Each primer set was also assessed for self- and cross-PCR dimer formation using a corresponding online oligonucleotide analysis tool ( To calculate the amplification efficiency, cDNA templates were diluted with nuclease-free water to × 1, ×2, ×5, ×10, and ×100 and then amplified to generate standard curves for each reference gene. The characteristics of the RT-qPCR primers, such as the corresponding amplicon length and amplification efficiency, are summarized in [Table 1].{Table 1}

Quantitative reverse transcription-polymerase chain reaction

RT-qPCR was performed according to the manufacturer's instructions using TB Green™ Premix Ex TaqTM (Takara) on a LightCycler96 Real-Time PCR platform (Roche, Basel, Switzerland) for analyzing the candidate reference genes.[21] Each reaction mixture contained 10-μL TB Green™ Premix Ex Taq, 1-μL diluted cDNA, 10 μM of the forward and reverse primers, and 5-μL double-distilled H2O in a total volume of 20 μL. The reaction mixture was subjected to an initial denaturation at 95°C for 30 s, followed by 40 cycles of denaturation at 95°C for 5 s, annealing at 60°C for 20 s (PCR), and melting curve analysis in the range of 65°C–95°C with 0.5°C/5 s increments to determine product specificity (melting curve). The cycle threshold (Ct) is defined as the number of PCR cycles required for the fluorescent signal to reach a specific threshold of detection and is inversely correlated to the amount of the target cDNA template.[22]

Statistical analysis

PCR amplification efficiencies were estimated using standard curves. The distribution of the reference gene expression between the HPR and LPR groups was assessed using a t-test. All statistical analyses were performed using SPSS software package version 20.0. (IBM, Armonk, USA). P < 0.01 was considered statistically significant in this study.

Classical and the most extensively used algorithms, including geNorm, NormFinder, and BestKeeper, were applied to analyze the stability of the ten reference genes in 16 samples from the HPR and LPR groups. The geNorm algorithm ranks the stability of reference genes based on their expression stability (M value) and introduces the appropriate stable reference gene number among those tested. Candidate reference genes were ranked by NormFinder and BestKeeper tools. Subsequently, comprehensive gene stability ranks were selected based on a previously reported tool.[23]


Purity and size distribution of sperm RNA

Total RNA extracted from the HPR and LPR groups was assessed for purity using NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific). The 260/280 absorbance (A260/A280) ratios of all the samples ranged from 1.90 to 2.05. Therefore, the purity of the total RNA from the sperm samples met the requirements for the subsequent studies. We also analyzed total RNA using the Agilent 2100 Bioanalyzer (Agilent Technologies) for size distribution. Although mature spermatozoa may harbor a distinct population of RNAs, gene transcription patterns remain elusive and exhibit no new RNA transcriptions.[3] Therefore, some of the transcripts retained in mature sperm are products of early transcription from various spermatogenic cells and are present as highly fragmented pieces of RNA whose RNA integrity numbers are <3 [Supplementary Table 1]. Thus, the sperm RNA size distribution recorded in our study was different from that of total RNA isolated from normal tissues, such as the testis and liver. These unique sperm RNA fragments were thus referred to as “RNA elements” [Figure 1].[INLINE:1]{Figure 1}

Verification, amplification efficiency, and expression levels of candidate reference genes

According to Taylor et al., amplification efficiencies between 1.8 and 2.0 are considered ideal for RT-qPCR applications, while R2 values >0.98 suggest that the efficiencies have been determined reliably. In the current study, ATP5B, SPAG7, and SRM did not meet these standards.[24] Thus, these three genes were excluded and not discussed in the present study. The amplification efficiencies of the seven remaining candidate reference genes ranged from 1.897 to 1.997 [Table 1]; therefore, these genes were considered ideal for RT-qPCR analysis. Moreover, these seven paired primers generated a single peak, and melting curve analyses showed no primer dimer formation. Overall, a wide spectrum of Ct values was observed across samples, ranging from 24.78 to 36.87 in seven candidate reference genes [Figure 2] and [Supplementary Table 2]. The t-test was a fundamental statistical analysis applied before the evaluation of the stability of reference genes. This is because the algorithms used assume that there is no difference in the expression pattern of candidate genes between experimental groups. All P values in our study were >0.01 [Supplementary Table 2], which indicated that the observed Ct values did not exhibit any significant differences between the HPR and LPR groups.{Figure 2}[INLINE:2]

Stability analysis using different algorithms

The geNorm algorithm ranks genes based on their M values. A lower M value indicates a more stable reference gene expression. In our study, the two most stable genes were MEA1 and TEGT, which exhibited the lowest M values of 0.856 and 0.872, respectively, followed by GAPDH, DAZAP2, and ACTB with M values of 0.875, 0.948, and 0.988, respectively. The least stable reference genes were PRM2 and PRM1 with M values of 1.257 and 1.336, respectively [Figure 3]. Moreover, pair-wise variation (V), contributing to the ultimate determination of the appropriate number of reference genes, was calculated as Vn/Vn + 1 (n = appropriate number of reference genes). Vn/Vn + 1 <0.15 indicated that the number of reference genes should be n; otherwise, n + 1 reference genes should be selected as internal controls. Our results showed a V3/V4 value of 0.141, which was below the threshold of 0.15 [Figure 4]. Therefore, according to geNorm, the three most stable genes (MEA1, TEGT, and GAPDH) were determined to be suitable as reference genes.{Figure 3}{Figure 4}

According to NormFinder analysis, the most stable reference gene was MEA1, followed by TEGT, GAPDH, DAZAP2, ACTB, and PRM2, whereas the least stable gene was PRM1 [Figure 5]. Based on the BestKeeper analysis criterion, a stable reference gene should exhibit Ct values with a low standard deviation (SD), and genes with an SD >1 should be considered unacceptable.[18] Therefore, PRM2 and PRM1 with a Ct value with SD>1 excluded from the study. Thus, the BestKeeper analysis results indicated that the most stable gene was ACTB (SD = 0.515), followed by TEGT, GAPDH, DAZAP2, and MEA1 [Figure 6].{Figure 5}{Figure 6}

Comprehensive ranking

Results were obtained using the three algorithms which differed for several genes. Therefore, a previously reported tool was considered to combine the three results for a comprehensive and optimal candidate reference gene ranking.[23] Briefly, the ranking of each reference gene in all of the algorithms was listed; for instance, PRM2 ranked as the 6th, 6th, and 7th most stably expressed reference gene according to geNorm, NormFinder, and BestKeeper, respectively. Next, the geometric mean of the three ranking numbers was calculated; thus, for PRM2, the result was 6.3 (6 × 6 × 7) 1/3. Finally, all the geometric means were ranked in descending order, with the most stable reference gene being the gene with the smallest geometric mean. Based on this analysis, MEA1 and TEGT were selected as the two most stable candidate reference genes [Table 2] and [Figure 7].{Figure 7}{Table 2}


Infertility, a disease of the reproductive system, is characterized by the failure to achieve a clinical pregnancy after 1 year of regular unprotected sexual intercourse. Infertility has become the subject of an increasing number of studies in the past several decades. Advances in scientific research and clinical technologies have led to significant breakthroughs in human reproduction.[25],[26] Although aberrant gene expression and epigenetic alterations have been associated with infertility, further studies on epigenetics, multi-omics, and RNA expression are needed to understand their functional significance in male fertility.[5],[6],[27] Spermatozoal RNAs are a recently discovered type of RNA; however, unequivocal evidence of precise sperm RNA function requires further investigation. Recently, RT-qPCR has been extensively used to analyze gene expression in male reproduction studies; however, the selection of suitable reference genes for quantitative spermatozoal transcript analyses remains to be precisely elucidated.

RT-qPCR is often used to validate gene expression data obtained from high-throughput NGS platforms. However, the interpretation of RT-qPCR data largely depends on the reference gene selected for normalization. Thus, to minimize systematic biases and accurately normalize target gene expression, the expression stability of the reference gene is crucial. Although several strategies have been proposed for the normalization of RT-qPCR data, including normalization to sample size or volume and total RNA, the use of one or more internal reference genes remains the most universal and precise method of normalization for such studies.[9] Sub-optimal selection of reference genes usually leads to bias or false results. GAPDH and ACTB are the two universally stable and the most commonly used reference genes in RT-qPCR studies of human tissues or animal models.[12],[13] However, these two genes are not robust and reliable for all gene expression analyses.[14] Moreover, due to the low yield and extreme complexity of RNAs in human spermatozoa, the evaluation of eligible reference genes for studies on human spermatozoa has not yet been extensively conducted.[3],[4],[15] Therefore, in the present study, using two groups of human sperm samples with different clinical pregnancy rates, we reanalyzed ten existing or putative candidate reference genes for human spermatozoa. We assessed the expression stabilities of these genes using three programs, namely geNorm, NormFinder, and BestKeeper.

geNorm analysis revealed MEA1 as the most stable gene, followed by TEGT and GAPDH. This result corresponds with a previous study showing that MEA1 is an appropriate internal reference gene for mRNA expression studies in human spermatozoa.[28] In addition, this method showed that the optimal number of reference genes was 3 (V3/V4 = 0.141). Although the ratio of V4/V5 (0.128) was below the cutoff value of 0.15, the three most stable genes are usually appropriate for accurate normalization, and numerous reference genes are impractical to some extent.[9],[29] Furthermore, MEA1 was identified as the reference gene with the highest stability using NormFinder, which was consistent with the results of geNorm. BestKeeper analysis determined that ACTB and TEGT were the reference genes with the least overall variation, which differed from the results obtained using geNorm and NormFinder. To overcome these discrepancies and obtain a comprehensive ranking, the results of the three different algorithms were combined, following which MEA1, TEGT, and ACTB were selected as the three most stable reference genes.[23] Because of the complexity of sperm RNA, ACTB usually does not present reliable normalization results; therefore, we selected MEA1 and TEGT as reference genes for the accurate normalization of RT-qPCR data for human spermatozoa.[3],[15]

Because these ten candidate reference genes were selected based on previous RT-qPCR studies, MEA1 and TEGT were considered the relative optimal reference genes in the current study. Herein, we did not study the relative quantification of small RNAs; thus, it is necessary to examine whether these two reference genes are appropriate for small RNA quantification. However, in recent years, advances in NGS, long-read RNA sequencing (RNA-seq), and direct RNA-seq methods have enabled the exploration of new transcripts and identification of novel reference genes.[30],[31] In addition, the characterization of RNAs retained in sperm by NGS has recently been reported.[2],[32] Moreover, RNA-seq has revealed a rich and complex population of unique coding and noncoding transcripts, including sperm-specific isoforms, intronic-retained and otherwise unannotated elements, and long or small noncoding RNAs.[32],[33] These newly discovered findings may help to provide insights into the most suitable and novel reference genes for sperm RNAs, including mRNAs, small RNAs, and other noncoding RNAs in future.

In summary, using three different statistical algorithms, we analyzed ten potential reference genes for gene expression studies in human sperm using clinical sperm samples, and we comprehensively identified MEA1 and TEGT as the most stably expressed reference genes. We believe that the most stable reference genes identified in the present study will improve the accuracy and standardization of gene expression studies using RT-qPCR analysis in human spermatozoa and facilitate further exploration of the mechanism underlying male infertility.

Supplementary information is linked to the online version of the paper on the Reproductive and Developmental Medicine website.

Financial support and sponsorship

This work was supported by grants from the National Key Research and Development Program of China (2018YFC1003500) and Medical Scientific Research Foundation of Guangdong Province of China (A2017531).

Conflicts of interest

The authors declare no conflicts of interest.


1Matzuk MM, Lamb DJ. The biology of infertility: Research advances and clinical challenges. Nat Med 2008;14:1197-213. doi: 10.1038/nm.f. 1895.
2Ostermeier GC, Miller D, Huntriss JD, Diamond MP, Krawetz SA. Reproductive biology: Delivering spermatozoan RNA to the oocyte. Nature 2004;429:154. doi: 10.1038/429154a.
3Miller D, Ostermeier GC. Towards a better understanding of RNA carriage by ejaculate spermatozoa. Hum Reprod Update 2006;12:757-67. doi: 10.1093/humupd/dml037.
4Pessot CA, Brito M, Figueroa J, Concha II, Yañez A, Burzio LO. Presence of RNA in the sperm nucleus. Biochem Biophys Res Commun 1989;158:272-8. doi: 10.1016/s0006-291x(89)80208-6.
5Legrand JM, Hobbs RM. RNA processing in the male germline: Mechanisms and implications for fertility. Semin Cell Dev Biol 2018;79:80-91. doi: 10.1016/j.semcdb.2017.10.006.
6Dadoune JP. Spermatozoal RNAs: What about their functions? Microsc Res Tech 2009;72:536-51. doi: 10.1002/jemt.20697.
7Bustin SA, Benes V, Nolan T, Pfaffl MW. Quantitative real-time RT-PCR – A perspective. J Mol Endocrinol 2005;34:597-601. doi: 10.1677/jme.1.01755.
8Arya M, Shergill IS, Williamson M, Gommersall L, Arya N, Patel HR. Basic principles of real-time quantitative PCR. Expert Rev Mol Diagn 2005;5:209-19. doi: 10.1586/14737159.5.2.209.
9Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, De Paepe A, et al. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 2002;3:RESEARCH0034. doi: 10.1186/gb-2002-3-7-research0034.
10Thellin O, Zorzi W, Lakaye B, De Borman B, Coumans B, Hennen G, et al. Housekeeping genes as internal standards: Use and limits. J Biotechnol 1999;75:291-5. doi: 10.1016/s0168-1656(99)00163-7.
11Andersen CL, Jensen JL, Ørntoft TF. Normalization of real-time quantitative reverse transcription-PCR data: A model-based variance estimation approach to identify genes suited for normalization, applied to bladder and colon cancer data sets. Cancer Res 2004;64:5245-50. doi: 10.1158/0008-5472.Can-04-0496.
12Bas A, Forsberg G, Hammarström S, Hammarström ML. Utility of the housekeeping genes 18S rRNA, beta-actin and glyceraldehyde-3-phosphate-dehydrogenase for normalization in real-time quantitative reverse transcriptase-polymerase chain reaction analysis of gene expression in human T lymphocytes. Scand J Immunol 2004;59:566-73. doi: 10.1111/j.0300-9475.2004.01440.x.
13Barber RD, Harmer DW, Coleman RA, Clark BJ. GAPDH as a housekeeping gene: Analysis of GAPDH mRNA expression in a panel of 72 human tissues. Physiol Genomics 2005;21:389-95. doi: 10.1152/physiolgenomics.00025.2005.
14Jung M, Ramankulov A, Roigas J, Johannsen M, Ringsdorf M, Kristiansen G, et al. In search of suitable reference genes for gene expression studies of human renal cell carcinoma by real-time PCR. BMC Mol Biol 2007;8:47. doi: 10.1186/1471-2199-8-47.
15Jodar M, Sendler E, Moskovtsev SI, Librach CL, Goodrich R, Swanson S, et al. Absence of sperm RNA elements correlates with idiopathic male infertility. Sci Transl Med 2015;7:295re6. doi: 10.1126/scitranslmed.aab1287.
16Muscatelli F, Walker AP, De Plaen E, Stafford AN, Monaco AP. Isolation and characterization of a MAGE gene family in the Xp21.3 region. Proc Natl Acad Sci U S A 1995;92:4987-91. doi: 10.1073/pnas.92.11.4987.
17Barragán M, MartÍnez A, Llonch S, Pujol A, Vernaeve V, Vassena R. Effect of ribonucleic acid (RNA) isolation methods on putative reference genes messenger RNA abundance in human spermatozoa. Andrology 2015;3:797-804. doi: 10.1111/andr.12053.
18Pfaffl MW, Tichopad A, Prgomet C, Neuvians TP. Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper – Excel-based tool using pair-wise correlations. Biotechnol Lett 2004;26:509-15. doi: 10.1023/b: bile. 0000019559.84305.47.
19Schuster A, Tang C, Xie Y, Ortogero N, Yuan S, Yan W. SpermBase: A database for Sperm-Borne RNA contents. Biol Reprod 2016;95:99. doi: 10.1095/biolreprod.116.142190.
20Zuker M. Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res 2003;31:3406-15. doi: 10.1093/nar/gkg595.
21Morrison TB, Weis JJ, Wittwer CT. Quantification of low-copy transcripts by continuous SYBR Green I monitoring during amplification. Biotechniques 1998;24:954-62.
22Walker NJ. Tech.Sight. A technique whose time has come. Science 2002;296:557-9. doi: 10.1126/science.296.5567.557.
23Chen D, Pan X, Xiao P, Farwell MA, Zhang B. Evaluation and identification of reliable reference genes for pharmacogenomics, toxicogenomics, and small RNA expression analysis. J Cell Physiol 2011;226:2469-77. doi: 10.1002/jcp.22725.
24Taylor CM, Jost R, Erskine W, Nelson MN. Identifying stable reference genes for qRT-PCR normalization in gene expression studies of Narrow-Leafed Lupin (Lupinus angustifolius L.). PLoS One 2016;11:e0148300. doi: 10.1371/journal.pone.0148300.
25Coutton C, Satre V, Arnoult C, Ray P. Genetics of male infertility: The new players. Med Sci 2012;28:497-502. doi: 10.1051/medsci/2012285014.
26Zegers-Hochschild F, Adamson GD, de Mouzon J, Ishihara O, Mansour R, Nygren K, et al. The international committee for monitoring assisted reproductive technology (icmart) and the world health organization (WHO) revised glossary on ART terminology, 2009. Hum Reprod 2009;24:2683-7. doi: 10.1093/humrep/dep343.
27Yan W. Male infertility caused by spermiogenic defects: Lessons from gene knockouts. Mol Cell Endocrinol 2009;306:24-32. doi: 10.1016/j.mce.2009.03.003.
28Ortega-Lázaro JC, del Mazo J. Expression of the B56delta subunit of protein phosphatase 2A and Mea1 in mouse spermatogenesis. Identification of a new B56gamma subunit (B56gamma4) specifically expressed in testis. Cytogenet Genome Res 2003;103:345-51. doi: 10.1159/000076823.
29Condori J, Nopo-Olazabal C, Medrano G, Medina-Bolivar F. Selection of reference genes for qPCR in hairy root cultures of peanut. BMC Res Notes 2011;4:392. doi: 10.1186/1756-0500-4-392.
30Garalde DR, Snell EA, Jachimowicz D, Sipos B, Lloyd JH, Bruce M, et al. Highly parallel direct RNA sequencing on an array of nanopores. Nat Methods 2018;15:201-6. doi: 10.1038/nmeth.4577.
31Byrne A, Beaudin AE, Olsen HE, Jain M, Cole C, Palmer T, et al. Nanopore long-read RNAseq reveals widespread transcriptional variation among the surface receptors of individual B cells. Nat Commun 2017;8:16027. doi: 10.1038/ncomms16027.
32Krawetz SA, Kruger A, Lalancette C, Tagett R, Anton E, Draghici S, et al. A survey of small RNAs in human sperm. Hum Reprod 2011;26:3401-12. doi: 10.1093/humrep/der329.
33Sendler E, Johnson GD, Mao S, Goodrich RJ, Diamond MP, Hauser R, et al. Stability, delivery and functions of human sperm RNAs at fertilization. Nucleic Acids Res 2013;41:4104-17. doi: 10.1093/nar/gkt132.