Figures
Abstract
The amplification of RNA is becoming increasingly important, as often only limited amounts of cells are available for gene expression analysis. In this study, the gene expression profile of the 39 human homeobox (HOX) genes was analyzed in bone marrow-derived multipotent stromal cells (BM-MSCs) by reverse transcription (RT-) and quantitative polymerase chain reaction (qPCR). For further unlimited gene expression analysis, Whole Transcriptome Amplification (WTA) was used to amplify RNA from human BM-MSCs. However, WTA led to biased RT- and qPCR results, and even non-detectability of HOX transcripts compared to non-amplified BM-MSC samples which instead revealed transcription. It is important to note that the same RNA of the respective human BM-MSC line was used for normal cDNA synthesis by standard reverse transcription (non-amplified RT samples) and for cDNA synthesis by WTA (amplified WTA samples). On this account, the different RT- and qPCR results were unexpected applying WTA. The significantly reduced detection of HOX transcripts after WTA has been demonstrated for numerous BM-MSC lines (n = 26) by RT-PCR analysis. Furthermore, undetectable HOX transcripts meaning HOX transcripts of human BM-MSCs that were detected after normal cDNA synthesis, but were no longer detectable after WTA, were consistently observed by qPCR analysis. Finally, qPCR experiments revealed a possible explanation for the differences between amplified and non-amplified BM-MSC samples: an inverse correlation between the biased qPCR results and the low expression level of the respective HOX gene. The PCR analysis of high-copy transcripts like GAPDH or RPL13A revealed unchanged qPCR results after WTA compared to corresponding non-amplified BM-MSC samples. In contrast, WTA led to biased qPCR results for medium-copy HOX transcripts, and even non-detectability of low-copy HOX transcripts of human BM-MSCs resulting in false negative RT- and qPCR data applying WTA.
Citation: Coenen C, Liedtke S, Kogler G (2015) RNA Amplification Protocol Leads to Biased Polymerase Chain Reaction Results Especially for Low-Copy Transcripts of Human Bone Marrow-Derived Stromal Cells. PLoS ONE 10(10): e0141070. https://doi.org/10.1371/journal.pone.0141070
Editor: Sakamuri V. Reddy, Charles P. Darby Children's Research Institute, 173 Ashley Avenue, Charleston, SC 29425, UNITED STATES
Received: July 22, 2015; Accepted: October 5, 2015; Published: October 20, 2015
Copyright: © 2015 Coenen et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: The authors have no support or funding to report.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The analysis of transcriptome profiling provides understanding of important biological processes in cancer or other diseases. However, often only limited amounts of cells and correlating RNA are available for molecular biological analysis like cDNA microarrays [1], next generation sequencing [2] or polymerase chain reactions [3]. RNA amplification offers unlimited gene expression analysis of limited amounts of raw material. The challenge of RNA amplification is to represent the original transcriptome.
In this study, Whole Transcriptome Amplification (WTA) method developed by Qiagen [4] was used to amplify RNA from human bone marrow-derived multipotent stromal cells (BM-MSCs). The WTA method is modified from the Whole Genome Amplification (WGA) method based on Multiple Displacement Amplification (MDA) [5, 6]. The special feature of MDA compared to other PCR-based methods is that the DNA amplification proceeds at a constant temperature of 30°C. The DNA amplification without thermal cycling is based on the extraordinary processibility of the enzyme phi29 polymerase synthesizing DNA fragments up to a length of 100 kb [7]. After phi29 polymerase tightly binds to random primers, the enzyme replicates the DNA template without dissociating from the DNA template strand. Instead, phi29 polymerase displaces complementary DNA strands and resolves secondary structures. The hybridization of further primers at the displaced DNA strands leads to additional replication resulting in DNA products larger than 10 kb. The long and branched DNA products synthesized by phi29 polymerase ensure that MDA enables the representative amplification of human genomic DNA [5]. An advantage of phi29 polymerase compared to thermostable Taq polymerase usually used in PCR-based methods, is their 3’− 5’ exonuclease proofreading activity resulting in significantly lower error rate of 1 in 106−107 [8] in contrast to normal Taq polymerase with an error rate of 3 in 104 [9].
However, MDA cannot be used for the amplification of RNA or more precisely of transcribed cDNA, because the enzyme phi29 polymerase only recognizes cDNA fragments longer than 2 kb. Therefore, a modified MDA approach for transcriptomic analysis has been developed by Qiagen [4]. To amplify also cDNA fragments smaller than 2 kb, a ligation step was added prior to MDA. In principle, the WTA method consists of three steps: 1. reverse transcription to generate cDNA, 2. ligation of short cDNA fragments to obtain template cDNA longer than 2 kb, 3. amplification of ligated cDNA by phi29 polymerase.
The HOX genes belong to the family of homeodomain-containing transcription factors acting as master transcriptional regulators [10]. During embryonic development, HOX genes determine the positional identity along the anterior-posterior body axis [11]. However, HOX genes are not only expressed during early development of vertebrates, but also in normal adult cells indicating tissue specific expression [12–14]. In human, 39 HOX genes are distributed among four HOX-clusters (A-D) on four different chromosomes (7p15, 17p21, 12q13 and 2q31). The term HOX-code describes the specific expression of functional active HOX genes. Our research group revealed the HOX-code as a “biological fingerprint” to distinguish functionally distinct stem cell populations derived from cord blood [15] and showed the adaption of native HOX-negative unrestricted somatic stromal cells (USSCs) to a positive HOX-code after co-culturing with HOX-positive cells [16].
To analyze whether the original transcriptome is still represented after WTA, the same RNA of the respective human BM-MSC line was applied for normal cDNA synthesis by standard reverse transcription (non-amplified RT samples) and for cDNA synthesis by WTA (amplified WTA samples). As readout, the gene expression profile of the 39 human homeobox (HOX) genes was analyzed for amplified and non-amplified human BM-MSC samples by RT- and qPCR analysis.
Material and Methods
Generation and Expansion of BM-MSCs
The human bone marrow-derived multipotent stromal cell (BM-MSC) lines were isolated using BM aspirated from the iliac crest of adult donors. The participants provide their written content. The ethical approval to isolate the stromal cells was obtained from the ethical review board of the Medical Faculty, University of Duesseldorf (study 3484). Cells were cultivated in medium consisting of DMEM, 30% fetal bovine serum (Invitrogen) and 1% Penicillin/Streptomycin/L-Glutamine (Lonza) under the standard conditions at 37°C, 21% O2 and 5% CO2 until reaching 80% confluence. The stromal cells were detached with 0.2% trypsin/EDTA solution.
To evaluate the cumulative population doublings (CPD), the total number of cells was calculated applying Eq 1:
(1)
N1: number of plated cells
N0: number of harvested cells
Total RNA Extraction
Total RNA was extracted from cell samples at passage 3 or 4, applying the RNeasy Mini Kit (Qiagen) according to manufacturer’s instructions. The isolation was performed with the optional use of QIAshredder columns to homogenize the cell-lysate and RNase-free DNase I to digest remaining DNA. The RNA concentration and purity was measured using a Nanodrop ND-100 device (Thermo Fisher Scientific). The ratio of 260/280 nm and 260/230 nm was in a confident range of 2.
Quality of RNA
The quality of RNA was examined by Agilent 2100 Bioanalyzer at the Genomics and Transcriptomics Laboratory of Biological Medical Research Center Düsseldorf allowing the calculation of RNA integrity number (RIN). RIN values can range from ten to one. A RIN value of ten represents intact RNA, while complete degraded RNA has a RIN value of one. In this study, only RNA samples with RIN values ≥ 9.8 were used representing very high RNA quality.
Reverse Transcription
Reverse transcription was performed for 60 min at 50°C using the First-strand cDNA synthesis Kit (Invitrogen) and the enclosed oligo(dT)20 primer. Up to 1000 ng total RNA were converted into first-strand cDNA in a 20 μl reaction. All control reactions provided with this system were carried out to monitor the efficiency of cDNA-synthesis. Prior to PCR, the completed first-strand reaction was heat-inactivated at 85°C for at least 5 min. Finally, cDNA was treated with RNaseH according to the manufacturer´s protocol.
Whole Transcriptome Amplification
For unlimited gene expression analysis, the QuantiTect Whole Transcriptome Kit (Qiagen) was used according to manufacturer’s instructions. This method synthesizes cDNA from total RNA by Whole Transcriptome Amplification and comprises three steps: reverse transcription, cDNA ligation and amplification. About 10 up to 100 ng of total RNA were added to RT mix. Furthermore, the high-yield reaction consisting of an 8 h amplification step was performed to receive high cDNA yield. Corresponding to the QuantiTect Whole Transcriptome Handbook, the amplified cDNA was diluted 1:250 with water, and directly used for RT- and qPCR analysis.
RT-PCR
Reverse transcription (RT)-PCR was carried out with gene specific and mainly intron-spanning primers (Thermo Scientific). The primer sequences (S1 Table) were carefully examined and checked for their specificity by applying BLASTn. GAPDH was used as reference gene. Approximately, 50 ng of cDNA or 1 μl of 1:250 diluted amplified cDNA were used for subsequent RT-PCR analysis in a total volume of 25 μl containing 1x PCR-buffer, 0.2 μM of each primer, 1.5 mM MgCl2, 0.2 mM of each dNTP and 1 U Taq DNA Polymerase (Invitrogen) at the following conditions: (1) 2 min at 95°C for initial denaturation, (2) 30 sec at 95°C, 30 sec at 56°C, 30 sec at 72°C for 35 cycles, (3) 5 min at 72°C for final extension of PCR products. As negative control, RT-PCR was performed without cDNA template. PCR was performed on a Mastercycler ep gradient S (Eppendorf). Subsequently, 20 μl aliquots of the RT-PCR products and related controls were analyzed by electrophoresis on a 2% agarose gel.
Quantitative PCR
Quantitative PCR analysis was performed with the suitable primers already applied for RT-PCR analysis (S1 Table). GAPDH was used as reference gene for normalization in all experiments. Quantitative PCR was carried out with SYBR® Green PCR Mastermix (Applied Biosystems) using 10 ng cDNA or 2.5 μl of amplified cDNA. All reactions were run in technical triplicates, on a Step One Plus device (Applied Biosystems). PCR reactions were performed in a total volume of 25 μl containing 12.5 μl SYBR® Green PCR Mastermix, 6.0 μl dH2O, 2.5 μl cDNA template and 2 μl (0.2 μM) of each primer at the following condition: (1) 10 min at 95°C, (2) 15 sec at 95°C and 1 min at 60°C for 40 cycles. Specificity of the PCR product was confirmed by analyzing the melting curve, respectively. The Step One Software v2.3 was used to run and analyze the comparative Ct experiments. The threshold was kept at 0.2 for all experiments. Eq 2 was applied to evaluate the relative gene expression:
(2)
Differential gene expression was calculated by 2-ΔCt. Quantitative PCR results were documented as mean values of technical triplicates and biological replicates with standard error of the mean (SEM).
Statistical Analysis
The data are presented as arithmetic means with a standard error of mean (SEM) of at least three independent experiments. Two-tailed unpaired t-tests were conducted with GraphPad Prism (v. 5.01) to determine significance. A Welch’s correction was applied in case of significantly different variances. P values lower than 0.05 were considered as significant (*, p = 0.01–0.05; **, p = 0.001–0.01; ***, p < 0.001).
Results
Expression Level of HOX Genes
The analysis of qPCR revealed that the expression level of HOX genes comprise a wide range in human BM-MSCs (Fig 1). Therefore, HOX genes are appropriate candidates to test WTA for unlimited PCR analysis of human BM-MSCs. In principle, HOX genes are weakly expressed in BM-MSCs compared to high-copy transcripts like RPL13A (Fig 1A) or GAPDH (data not shown). The high-copy RPL13A transcript revealed a threshold cycle (Ct) value of about 16. On basis of their expression level, HOX genes were graduated in groups of medium -, and low-copy transcripts. HOX genes with a Ct value between 22 and 25 were classified into the group of medium-copy transcripts. For example HOXA9, HOXC10, HOXB7, HOXA10, HOXC8, HOXB4 and HOXD8 represent medium-copy transcripts (Fig 1B). HOX genes revealing a Ct value between 26 and 29 were associated as low-copy transcripts. For instance, low-copy transcripts are HOXA5, HOXD9, HOXB6, HOXA1, HOXC5, HOXA4, and HOXD3 (Fig 1C).
The expression level of seven medium-, and low copy HOX transcripts is documented for biological and technical replicates of BM-MSC lines (n = 3) by 2-ΔCT value related to GAPDH. (A) HOX genes are weakly expressed in human BM-MSCs compared to high-copy transcripts like RPL13A showing an expression level of about 10−1 related to GAPDH. (B) The medium-copy HOX transcripts present an expression level of about 10−3 related to GAPDH. (C) The expression level of low-copy HOX transcripts decreases to 10−4 related to GAPDH.
Non-Detectability of HOX Transcripts after WTA
The same RNA of the respective human BM-MSC line was used for normal cDNA synthesis by standard reverse transcription (non-amplified RT samples) or for cDNA synthesis by Whole Transcriptome Amplification (amplified WTA samples). Following RT- and qPCR analysis were performed and the PCR results of non-amplified and amplified BM-MSC samples were compared with each other (Fig 2).
Human bone marrow-derived multipotent stromal cell (BM-MSC) lines were cultivated and expanded until passage 3 or 4. Following the RNA was isolated and the cDNA preparation was performed by standard reverse transcription and by Whole Transcriptome Amplification (WTA). The same RNA of the respective human BM-MSC line was applied for standard reverse transcription and WTA to compare RT- and qPCR results of the non-amplified and amplified cDNA samples. To exclude RNA degradation, the RNA integrity number (RIN) was determined by Agilent Bioanalyzer and only RNA samples with RIN values ≥ 9.8 were applied. Additionally, various amounts of RNA and cDNA were tested as starting material for WTA or for RT-PCR analysis.
RT-PCR analysis of human BM-MSC lines revealed a characteristic HOX gene expression, the so-called HOX-code (Fig 3). The HOX-code is characterized by predominant expression of genes from the HOXA-, HOXB- and HOXC-cluster (Fig 3A) The majority of BM-MSC lines expressed the genes HOXA1, HOXA2, HOXA4, HOXA5, HOXA6, HOXA7, HOXA9, HOXA10 and HOXA11 from HOXA-cluster. The consistently expressed genes from HOXB-cluster were HOXB4, HOXB5, HOXB6 and HOXB7. The constantly expressed genes from HOXC-cluster were HOXC6, HOXC8, HOXC9 and HOXC10. The smallest number of HOX genes was expressed in HOXD-cluster. All BM-MSC lines revealed the expression of HOXD8 and HOXD9, while the expression of HOXD3 and HOXD4 varied individually between the different BM-MSC lines. Generally, the BM-MSCs showed a reduced expression of the 5’-HOX genes from the HOXB-, HOXC- and HOXD-cluster. The majority of HOX genes, including the 5’-HOX genes, were expressed in the HOXA-cluster. The BM-MSCs expressed by mean 24 out of the 39 tested human HOX genes. But only about 10 HOX genes were still detectable after WTA (Fig 3B). It is important to note that the same RNA of the respective human BM-MSC line was used for normal cDNA synthesis by standard reverse transcription (non-amplified RT samples) and for cDNA synthesis by WTA (amplified WTA samples). On this account, the different RT-PCR results were unexpected applying WTA. The significantly reduced detection of HOX transcripts after WTA has been demonstrated for numerous BM-MSC lines (n = 26) by RT-PCR analysis. Evaluating the RT-PCR results revealed that mainly HOX genes like HOXA5, HOXA7, HOXA9, HOXA10, HOXB5, HOXB7, HOXC6, HOXC10 and HOXD8 were still detectable after WTA. The HOX transcripts of human BM-MSCs that were no more detectable after WTA are called undetectable HOX transcripts hereafter.
In this experiment, the same RNA of the respective human bone marrow-derived multipotent stromal cell (BM-MSC) line (biological triplicates, n = 9) was applied for preparation of cDNA by standard reverse transcription (RT samples) and by Whole Transcriptome Amplification (WTA samples). (A) The so-called HOX-code of BM-MSCs is characterized by predominant gene expression of HOXA-, HOXB- and HOXC-cluster compared to HOXD-cluster. (B) After WTA, a significantly reduced detection of HOX transcripts was observed for all tested BM-MSC lines.
The RNA quality of human BM-MSC lines was tested to exclude RNA degradation. The RIN values ranged between 9.8 and 10.0 (data not shown) representing very high RNA quality. Additionally, various amounts of RNA as starting material for WTA were analyzed to test, if low-copy HOX transcripts are potentially underrepresented in the recommended amount of 10 ng RNA (Fig 4). Further on, RT-PCR analysis revealed significantly reduced detection of HOX transcripts after WTA for numerous BM-MSC lines (n = 16), even though gradually higher amounts of RNA template were used for WTA reaction.
To exclude underrepresentation of low-copy HOX transcripts in the recommended amount of 10 ng RNA, between 10 ng up to 100 ng RNA were inserted to WTA reaction. RT-PCR analysis revealed consistently reduced detection of HOX transcripts for numerous BM-MSC lines (n = 16), despite gradually increasing amounts of RNA.
Furthermore, different dilutions of amplified cDNA as starting material for RT-PCR analysis were analyzed by dilution series (Fig 5). But lower (1:100, 1:10) and higher (1:500, 1:1000, 1:2000) cDNA dilutions led to a higher amount of undetectable HOX transcripts in RT-PCR analysis compared to the recommended 1:250 cDNA dilution. According to these results, the 1:250 cDNA dilution of the respective human BM-MSC line was used for further PCR analysis. In summary, neither the RNA quality nor the amount of RNA or cDNA seemed to be responsible for the undetectable HOX transcripts after WTA.
The same cDNA from human bone marrow-derived multipotent stromal cell (BM-MSC) line 1 and 3 was constantly diluted for RT-PCR analysis. In brackets, the dilution of amplified cDNA is outlined after high-yield Whole Transcriptome Amplification (WTA) reaction (8 h amplification). (A) Lower cDNA dilutions (1:100, 1:10) led to more undetectable HOX transcripts like HOXA9 or HOXA10 compared to the recommended 1:250 cDNA dilution. (B) Higher cDNA dilutions (1:500, 1:1000, 1:2000) also revealed a higher amount of undetectable HOX transcripts like HOXA9, HOXA10 or HOXB9 compared to the recommended 1:250 cDNA dilution.
Biased qPCR Results for Medium-Copy HOX Transcripts after WTA
Similar to RT-PCR analysis, the same cDNA of the respective human BM-MSC line synthesized by standard reverse transcription (non-amplified RT samples) and Whole Transcriptome Amplification (amplified WTA samples) was used for q PCR analysis. Exemplary, three medium- and low-copy HOX transcripts were examined in qPCR analysis. As a result, the amplified WTA samples of human BM-MSCs revealed biased expression level for medium-copy HOX transcripts in comparison to RT samples of the same BM-MSC lines (Fig 6). For example, the expression level of HOXA9 reached a negative fold change of -10.9 after WTA (Fig 6A). Quantitative PCR analysis of HOXB7 revealed also a significantly reduced expression level after WTA with a negative fold change of -15.8 compared to the same non-amplified human BM-MSC lines (Fig 6B). The expression level of HOXD8 showed a negative fold change of -7.4 after WTA (Fig 6C). As control, the qPCR analysis was also performed with high-copy transcripts like RPL13A (Fig 6D) and GAPDH (data not shown). The expression level of RPL13A and GAPDH was not significantly different for human BM-MSCs with and without WTA.
Similar to RT-PCR analysis, the same cDNA of the respective human bone marrow-derived multipotent stromal cell (BM-MSC) line (biological triplicates, n = 3) synthesized by standard reverse transcription (RT samples) and Whole Transcriptome Amplification (WTA samples) was used for qPCR analysis. (A-C) The WTA samples revealed a significantly decreased expression level of the medium-copy HOX genes like HOXA9, HOXB7 and HOXD8 in comparison to RT samples. (D) The reference gene RPL13A is expressed on a significantly higher level compared to the HOX genes resulting in no significantly different qPCR results after WTA.
Non-Detectability of Low-Copy HOX Transcripts in qPCR Analysis after WTA
Quantitative PCR analysis of medium-copy HOX transcripts of amplified human BM-MSC lines (WTA samples) revealed on average a negative fold change of -11.4 in comparison to same non-amplified BM-MSC lines (RT samples). In contrast, qPCR analysis of the low-copy HOXA4 and HOXD9 transcripts of amplified human BM-MSC lines showed higher negative fold changes (Fig 7). The expression level of HOXA4 reached a negative fold change of -20.1 after WTA. The expression level of HOXD9 showed a negative fold change of -83.7 after amplification. In contrast, the expression level of the low-copy HOXD3 transcript was not significantly different with and without WTA. However, a special feature was observed for low-copy HOX transcripts: The low-copy HOXA4, HOXD3 and HOXD9 transcripts could not always be detected in all tested WTA samples resulting in artificial qPCR data. For example, the expression of HOXA4 transcript could only be detected in 7 out of 9 (7/9) tested WTA samples (Fig 7A). The low-copy HOXD3 transcript was only established in 3/9 WTA samples (Fig 7B). The HOXD9 expression was only detected in 5/9 WTA samples (Fig 7C). In comparison, the low-copy HOXA4, HOXD3 and HOXD9 transcripts of human BM-MSCs were detected in each of the 9 tested RT samples. Agarose electrophoresis was additionally performed with qPCR products to identify the specific PCR products, not only by their characteristic melt curves, but also by their characteristic product length (data not shown).
Similar to RT-PCR analysis, the same cDNA of the respective human bone marrow-derived multipotent stromal cell (BM-MSC) line (biological replicates, n = 9) synthesized by standard reverse transcription (RT samples) or Whole Transcriptome Amplification (WTA samples) was used for qPCR analysis. (A-C) The number of samples out of the 9 tested samples is shown in brackets, where the respective low-copy HOX transcript could be detected in qPCR analysis. The low-copy HOX transcripts were undetectable in the remaining samples. In comparison to WTA samples, all RT samples of BM-MSC lines revealed HOXA4, HOXD3 and HOXD9 gene expression.
It can be concluded that the PCR analysis of high-copy transcripts revealed unchanged qPCR results for human BM-MSCs after WTA. In contrast, WTA led to biased qPCR results for medium-copy transcripts, and even non-detectability of low-copy transcripts of human BM-MSCs. Moreover, undetectable HOX transcripts of human BM-MSCs after WTA were consistently observed by RT- and qPCR analysis.
Discussion
RNA amplification provides a tool to perform PCR analysis of limited amount of cells, and associated RNA. In this study, the gene expression profile of the 39 HOX genes was analyzed for human BM-MSCs by RT- and qPCR analysis. For further unlimited gene expression analysis, the isolated RNA was amplified by WTA method based on MDA.
One property of HOX genes is their wide range of different expression levels in human BM-MSCs. Furthermore, the majority of HOX genes only comprise one intron. Therefore, the HOX genes reveal similar structure and full-length transcripts between 800–5000 nucleotides, whereby the most HOX genes have transcript lengths of about 2000 nucleotides. As a result, the location of primers is selected very similar, nevertheless specific for each HOX gene. Their comparability and simultaneously their specificity and different expression levels in human BM-MSCs enable HOX genes as suitable candidates for the evaluation of WTA.
As result, WTA led to biased RT- and qPCR results, and even non-detectability of HOX transcripts compared to non-amplified BM-MSC samples. It is important to note that the same RNA of the respective BM-MSC line was used for normal cDNA synthesis by standard reverse transcription (non-amplified RT samples) and for cDNA synthesis by WTA (amplified WTA samples). On this account, the different RT- and qPCR results were unexpected applying WTA.
RT-PCR results revealed significantly reduced detection of HOX transcripts after WTA for numerous BM-MSC lines (n = 26). It was conspicuous that mainly HOX genes like HOXA5, HOXA7, HOXA9, HOXA10, HOXB5, HOXB7, HOXC6, HOXC10 and HOXD8 were still detectable after amplification. Based on these results, it might be suggested that preferential more strongly expressed HOX genes, identifiable by strong DNA bands in RT-PCR gels, are still detectable after WTA. Accordingly, undetectable HOX genes meaning HOX transcripts of human BM-MSCs that were detected after normal cDNA synthesis, but were no more detectable after WTA, seemed to be more weakly expressed HOX genes. This assumption was confirmed by qPCR analysis (Fig 8). The medium-copy HOX transcripts HOXA9, HOXB7 and HOXD8 could be detected in all tested human BM-MSC lines (n = 9) after WTA by qPCR analysis. In contrast, the low-copy HOX transcripts HOXA4, HOXD3 and HOXD9 could only be detected in some of the BM-MSC lines after WTA, despite the expression of these low-copy HOX transcripts was established in the corresponding non-amplified BM-MSC samples by qPCR analysis. Correlating to qPCR analysis, the majority of medium-copy HOX transcripts could also be detected in RT-PCR analysis after WTA showing a DNA band in the RT-PCR gel of WTA samples. Likewise, the low-copy HOX transcripts HOXA4, HOXD3 and HOXD9 could not be detected after WTA by RT-PCR, despite the expression of these low-copy HOX genes was established in the corresponding non-amplified BM-MSC samples. Considering that qPCR is a more sensitive method compared to RT-PCR requiring a certain amount of DNA molecules to result in a visible DNA band on the RT-PCR gel, it is not surprising that some low-copy HOX transcripts could be detected by qPCR. Nevertheless, low-copy HOX transcripts got lost after WTA for human BM-MSCs. Moreover, qPCR analysis of high-, medium-, and low-copy transcripts revealed an intergradation reaching from not significantly different to biased and even undetectable HOX transcripts.
When cDNA was synthesized by standard cDNA synthesis, the tested medium- and low-copy HOX transcripts could be detected by qPCR. In contrast, when cDNA was synthesized and amplified by WTA, the low-copy HOX transcripts were not fully converted to cDNA. n.d. = not determined.
The represented qPCR data of human BM-MSCs can lead to the conclusion that the gene expression level may influence the efficiency of WTA. This assumption is confirmed by the observation that qPCR analysis of high-copy transcripts RPL13A and GAPDH revealed no significantly different PCR results after WTA. In contrast, WTA led to biased qPCR results for medium-copy HOX transcripts based on significantly reduced expression level of medium-copy HOX transcripts after WTA compared to the same non-amplified BM-MSC samples. Finally, WTA led to non-detectability of low-copy HOX transcripts for human BM-MSCs resulting in false negative RT- and qPCR data applying WTA.
Simultaneously, the identification of high- and medium-copy transcripts by RT- and qPCR analysis of human BM-MSCs after WTA gave evidence for functioning of WTA reaction and the correct technical application. But nevertheless, the significantly reduced expression level of medium-copy transcripts might be negligible, if only amplified cDNA samples are compared with each other. The analysis of up-regulated or overexpressed genes of amplified cDNA samples is also possible [17, 18]. However, the use of WTA is getting difficult, when the gene expression profile of amplified and non-amplified cDNA samples is analyzed and low-copy transcripts are no more detectable after WTA resulting in false negative RT- and qPCR data.
The quality and quantity of RNA as starting material for WTA reaction was tested to exclude that these factors limit the efficiency of WTA reaction. The Agilent analysis revealed high RNA quality of human BM-MSCs represented by RIN values between 9.8 and 10.0. To exclude underrepresentation of low-copy transcripts, larger amounts of RNA were inserted to WTA reaction. Despite higher starting amounts of RNA from 10 ng up to 100 ng, the reduced detection of HOX transcripts was consistently observed for human BM-MSCs indicating that not the absolute amount of RNA, but rather the expression level of high-, medium-, and low-copy transcripts seems to determine the efficiency of WTA reaction. Different cDNA dilutions were additionally tested to determine the suitable amount of cDNA template for RT-PCR reactions. As a result, the recommended 1:250 cDNA dilution was considered as adequate for PCR analysis. Another possible source of errors could be the unequal distribution of low-copy transcripts when RNA isolation is performed with less than 1000 cells resulting in stochastic problems. However, the number of cells inserted to RNA isolation reached from minimal 40.000 up to 1.000.000 cells in this study. In the first step of WTA reaction, a mixture of random and oligo-dT primers were used for cDNA generation. Based on the primers, it is recommended to analyze no full-length transcripts after WTA. In this study no full-length transcripts were analyzed by RT- and qPCR and the following example revealed that the length of PCR products could also not be the explanation for undetectable HOX transcripts after WTA reaction. The product length of GAPDH is 238 bp. In qPCR analysis, GAPDH was stable detected in 9 out of 9 tested WTA samples. The PCR product length of HOXD3 is 187 bp. Although the PCR product length of HOXD3 is smaller compared to GAPDH, HOXD3 was only detected in 3 out of 9 WTA samples by qPCR. For SYBR Green assays, the amplicon for PCR products should be less than 300 bp [3]. This aspect was also taken into account for qPCR analysis since the PCR products exhibit a length between 112–238 bp. In summary, neither the RNA quality nor the quantity of RNA template for WTA reaction, the amount of cDNA template for PCR reaction, the number of cells for RNA isolation or the length of PCR products seemed to be responsible for the undetectable low-copy HOX transcripts in RT- and qPCR analysis after WTA.
The undetectable low-copy HOX transcripts may be a result of unspecific amplification during WTA reaction, perhaps of random extension of primer-dimers. In case of unspecific amplification, at first the low-copy transcripts get lost, then the number of medium-copy transcripts is impacted, while the influence on high-copy transcripts is unaffected. However, the customer has no influence on this effect, provided that RNA degradation and contamination can be excluded. Additionally it cannot be excluded that the different primers used for normal cDNA synthesis by standard reverse transcription (oligo-dT primers) and for cDNA synthesis by WTA (mixture of random and oligo-dT primers) contribute to the significantly reduced detection of HOX transcripts after WTA compared to non-amplified BM-MSC samples. The mixture of random and oligo-dT primers was probably used for cDNA synthesis by WTA to also recognize transcripts without poly(A) tail. Therefore, it is contradictory that less HOX transcripts could be detected after WTA compared to standard reverse transcription using only oligo-dT primers for cDNA synthesis.
In the second step of WTA reaction, the generated cDNA is ligated before amplification. The ligation is required because the amplification enzyme phi29 DNA polymerase only recognizes cDNA fragments longer than 2 kb [4]. Another research group modified the ligation and amplification time of WTA reaction and identified the ligation step as key factor [19]. Their experiments revealed a correlation between the ligation time and the amplification efficiency. Higher amplification efficiencies were obtained with increasing ligation time. The efficiency of ligation may also be an explanation for significantly reduced detection of HOX transcripts after WTA in line with their expression level.
The generation of cDNA in the first step of WTA reaction, mainly led to cDNA fragments smaller than 2 kb. If these cDNA fragments are not ligated during the ligation step of WTA reaction, they cannot be amplified by phi29 DNA polymerase afterwards. Therefore, the representation of original transcriptome may depend on both, the amplification and ligation efficiency of WTA reaction.
In the study presented here, the WTA efficiency was not equal for medium- and low copy HOX transcripts of human BM-MSCs. Moreover, the low-copy HOX transcripts were not fully converted into cDNA by WTA reaction resulting in false negative RT- and qPCR data. The represented qPCR data of human BM-MSCs can lead to the conclusion that the WTA efficiency may depend on the expression level of the respective gene. Furthermore, it can be suggested that this is not a specific effect of HOX transcripts and most likely also applies to other genes tested in other cell types. The use of RNA amplification is becoming more popular, as often only limited amounts of cells are available for gene expression analysis, but representing the complete original transcriptome, especially with regard to low-copy transcripts, is error-prone.
Author Contributions
Conceived and designed the experiments: CC SL GK. Performed the experiments: CC. Analyzed the data: CC. Contributed reagents/materials/analysis tools: CC. Wrote the paper: CC. Revising the manuscript for important intellectual content: SL. Final approval of manuscript: SL GK.
References
- 1. Xiang CC, Chen Y. cDNA microarray technology and its applications. Biotechnol Adv. 2000 Mar;18(1):35–46. pmid:14538118
- 2. van Dijk EL, Auger H, Jaszczyszyn Y, Thermes C. Ten years of next-generation sequencing technology. Trends Genet. 2014 Sep;30(9):418–26. pmid:25108476
- 3. Ginzinger DG. Gene quantification using real-time quantitative PCR: an emerging technology hits the mainstream. Exp Hematol. 2002 Jun;30(6):503–12. pmid:12063017
- 4. Berthet N, Reinhardt AK, Leclercq I, van Ooyen S, Batejat C, Dickinson P, et al. Phi29 polymerase based random amplification of viral RNA as an alternative to random RT-PCR. BMC Mol Biol. 2008;9:77. pmid:18771595
- 5. Dean FB, Hosono S, Fang L, Wu X, Faruqi AF, Bray-Ward P, et al. Comprehensive human genome amplification using multiple displacement amplification. Proc Natl Acad Sci U S A. 2002 Apr 16;99(8):5261–6. pmid:11959976
- 6. Hosono S, Faruqi AF, Dean FB, Du Y, Sun Z, Wu X, et al. Unbiased whole-genome amplification directly from clinical samples. Genome Res. 2003 May;13(5):954–64. pmid:12695328
- 7. Blanco L, Bernad A, Lazaro JM, Martin G, Garmendia C, Salas M. Highly efficient DNA synthesis by the phage phi 29 DNA polymerase. Symmetrical mode of DNA replication. J Biol Chem. 1989 May 25;264(15):8935–40. pmid:2498321
- 8. Esteban JA, Salas M, Blanco L. Fidelity of phi 29 DNA polymerase. Comparison between protein-primed initiation and DNA polymerization. J Biol Chem. 1993 Feb 5;268(4):2719–26. pmid:8428945
- 9. Eckert KA, Kunkel TA. DNA polymerase fidelity and the polymerase chain reaction. PCR Methods Appl. 1991 Aug;1(1):17–24. pmid:1842916
- 10. Gehring WJ, Affolter M, Burglin T. Homeodomain proteins. Annu Rev Biochem. 1994;63:487–526. pmid:7979246
- 11. Krumlauf R. Hox genes in vertebrate development. Cell. 1994 Jul 29;78(2):191–201. pmid:7913880
- 12. Yamamoto M, Takai D, Yamamoto F, Yamamoto F. Comprehensive expression profiling of highly homologous 39 hox genes in 26 different human adult tissues by the modified systematic multiplex RT-pCR method reveals tissue-specific expression pattern that suggests an important role of chromosomal structure in the regulation of hox gene expression in adult tissues. Gene Expr. 2003;11(3–4):199–210. pmid:14686791
- 13. Takahashi Y, Hamada J, Murakawa K, Takada M, Tada M, Nogami I, et al. Expression profiles of 39 HOX genes in normal human adult organs and anaplastic thyroid cancer cell lines by quantitative real-time RT-PCR system. Exp Cell Res. 2004 Feb 1;293(1):144–53. pmid:14729064
- 14. Ackema KB, Charite J. Mesenchymal stem cells from different organs are characterized by distinct topographic Hox codes. Stem Cells Dev. 2008 Oct;17(5):979–91. pmid:18533811
- 15. Liedtke S, Buchheiser A, Bosch J, Bosse F, Kruse F, Zhao X, et al. The HOX Code as a "biological fingerprint" to distinguish functionally distinct stem cell populations derived from cord blood. Stem Cell Res. 2010 Jul;5(1):40–50. pmid:20434420
- 16. Liedtke S, Freytag EM, Bosch J, Houben AP, Radke TF, Deenen R, et al. Neonatal mesenchymal-like cells adapt to surrounding cells. Stem Cell Res. 2013 Jul;11(1):634–46. pmid:23660338
- 17. Hwang Y, Kim J, Shin JY, Kim JI, Seo JS, Webster MJ, et al. Gene expression profiling by mRNA sequencing reveals increased expression of immune/inflammation-related genes in the hippocampus of individuals with schizophrenia. Transl Psychiatry. 2013;3:e321. pmid:24169640
- 18. Reinart N, Nguyen PH, Boucas J, Rosen N, Kvasnicka HM, Heukamp L, et al. Delayed development of chronic lymphocytic leukemia in the absence of macrophage migration inhibitory factor. Blood. 2015 Jan 31;121(5):812–21.
- 19. Wu J, Gao W, Zhang W, Meldrum DR. Optimization of whole-transcriptome amplification from low cell density deep-sea microbial samples for metatranscriptomic analysis. J Microbiol Methods. 2011 Jan;84(1):88–93. pmid:21044647