For decades, biochemists and molecular biologists have utilized reverse transcription of mRNA coupled with polymerase chain reaction (RT-PCR) to analyze differential gene expression in biological samples. While this approach is quite specific for target genes of interest, it is limited in its ability to sensitively and quantitatively measure gene expression. The introduction of “real-time” quantitative PCR (qPCR) in recent years has allowed researchers to better quantify gene expression in biological systems [1, 2]. Consequently, qPCR has widely replaced traditional PCR in gene expression protocols .
Given the current emphasis on gene expression analysis by RT-qPCR in research and clinical settings, we aimed to expose chemistry majors in our biochemistry track to this modern technique in the context of a meaningful biological system. While numerous published laboratory experiments introduce undergraduate students to “real time” qPCR (for examples, see Refs. [4-6]), a recent search of the literature found only one published undergraduate experiment employing RT-qPCR to analyze the differential expression of a target gene . In this RT-qPCR project described by Hancock and coworkers, students measure the expression of four target sequences during the DMSO-induced differentiation of cultured erythrocytes. Unfortunately, cultured erythrocytes are not routinely maintained on many smaller campuses. We developed an RT-qPCR project for upper-level undergraduate students using the model laboratory plant Arabidopsis thaliana (thale cress). Arabidopsis is an ideal system for gene expression experiments in an undergraduate setting; Arabidopsis seeds are readily available, and they are quickly and easily grown , without the need to maintain cell culture lines from year-to-year. In addition, Arabidopsis plants can be grown just about anywhere under inexpensive fluorescent lamps.
In this novel laboratory project, students use RT-qPCR to assess the environmental regulation of the FLOWERING LOCUS C (FLC) gene in Arabidopsis. FLC encodes the FLC protein, a transcription factor that plays a key role in the timing of the developmental transition from vegetative growth to reproductive growth (flowering) in Arabidopsis . FLC delays flowering by directly inhibiting the expression of genes that promote the floral transition . In many Arabidopsis populations (ecotypes), the cellular concentration of FLC transcripts (mRNA) declines after vernalization (long-term exposure to cold that is characteristic of winter), thereby enabling the plants to flower rapidly in the springtime (Fig. 1) . While different ecotypes of Arabidopsis vary in their vernalization responses for the onset of flowering, vernalization produces a robust and reproducible reduction in FLC transcripts in many Arabidopsis ecotypes [9, 11-13]. In the project described here, students use RT-qPCR to determine the relative difference in FLC expression between vernalized (vern.) and nonvernalized (nonvern.) Arabidopsis plants of the Columbia ecotype, the variety most commonly used in Arabidopsis research.
An additional goal for our project was to provide an opportunity for students to calculate relative gene expression values from raw data and propagate experimental errors, giving students a greater understanding of their experimental results and an appreciation for the assumptions and limitations inherent with this analytical method . The emergence of qPCR as a mainstream laboratory technique in recent years has prompted manufacturers of qPCR systems to develop software that is widely accessible. Indeed, current qPCR software programs allow researchers to obtain gene expression results with a few simple clicks of a mouse. As a result, many researchers today lack a thorough understanding of the data analysis underlying RT-qPCR experiments and the appropriate application of statistics to RT-qPCR results. Indeed, results of a recent survey published in Science points to an alarming reality in the use of scientific software: “Most people, in some form, ‘trust’ software without knowing everything about how it works .”
After completing this laboratory experiment, students should be able to
1. explain how RT-qPCR works,
2. follow established protocols to successfully extract RNA from plant tissues and reverse transcribe total RNA to cDNA,
3. follow an established protocol to successfully carry out qPCR reactions with appropriate control reactions,
4. use the method  to analyze qPCR data and determine the relative quantities of a target transcript in different biological samples with appropriate consideration of experimental errors, and
5. design a relative gene expression experiment using RT-qPCR, with the appropriate controls.
How “Real Time” qPCR Works
In traditional PCR, amplification products are qualitatively analyzed by gel electrophoresis during the “plateau” phase of the PCR process that follows 20–40 amplification cycles. In qPCR, amplification products are detected fluorometrically in real time, during the exponential phase of the PCR process (Fig. 2), using commercially available fluorescent dyes . The resulting fluorescence signal is proportional to the amount of double-stranded DNA product present after each PCR cycle, allowing for a quantitative analysis of the reaction products. The analytical parameter used to quantify DNA in a sample is the CT (cycle threshold) value. The CT value is determined from a log-linear plot of fluorescence signal versus cycle number and is the cycle number at which the fluorescence signal exceeds a threshold value [1, 16]. The more abundant the target DNA is in a sample, the earlier the fluorescence signal will cross the threshold and the lower the CT value will be. Using a series of cDNA standards, RT-qPCR enables the determination of the absolute number of transcript copies in a sample (absolute quantification). However, it is often sufficient to use RT-qPCR to determine the relative differences in gene expression between two or more biological samples (relative quantification) . In this project, students use RT-qPCR to determine the effect of vernalization on the relative levels of FLC transcripts in young Arabidopsis seedlings.
Teaching and Laboratory Procedures
This laboratory project requires four 3-hr laboratory sessions, which includes in-class time for data analysis. Detailed Instructor Set-up Sheets and a Student Handout with complete experimental protocols are provided in the Supporting Information. In laboratory Period 1, students extract RNA from Arabidopsis seedlings grown from vern. and nonvern. seeds. In laboratory Periods 2 and 3, students synthesize cDNA from their total extracted RNA and set-up their collective qPCR reactions. The final laboratory period is spent on validation of amplification products by gel electrophoresis and data analysis. Finally, to better understand the theoretical basis of relative gene expression analysis in RT-qPCR, student pairs perform their own fold-expression calculations using the method , complete with propagation of experimental errors.
In the project, each student has the opportunity to manipulate one plant sample (vern. or nonvern.), allowing all students to gain confidence in the laboratory techniques. However, students are paired up (vern. with nonvern.) for the entire project so they can, as a team, work toward a common experimental goal: to compare transcript levels between their two plant samples and determine the effect of vernalization on the expression of FLC. Pairs of students collaboratively set up their qPCR reactions, validate their reaction products, and carry out data analysis. Student pairs work together to follow lengthy experimental protocols, troubleshoot experiments as necessary, and interpret results from several experimental techniques (gel electrophoresis, melting temperature analysis, qPCR). This facilitates peer learning  and encourages communication of scientific concepts. In addition, students learn valuable teamwork skills, which is a desired outcome of the American Society for Biochemistry and Molecular Biology's recommended undergraduate curriculum in biochemistry and molecular biology .
Laboratory Period 1
Students harvest 10- to 14-day-old seedlings grown from vern. and nonvern. Arabidopsis (ecotype Columbia) seeds. Seedlings are grown ahead of time by the instructor or teaching assistant. Each pair of students analyzes one vern. sample and one nonvern. sample. Students add aerial (above ground) parts of seedlings to 1.7-mL microfuge tubes with locking caps and freeze the tubes in liquid nitrogen. Next, students grind the plant tissue in the microfuge tube by hand with a disposable plastic pellet pestle and extract total RNA from each sample using the Qiagen RNeasy Mini Kit. 1 To minimize contamination from genomic DNA, an additional 30-min. on-column DNase treatment is included using RNase-free DNase (Qiagen). 2 Students are trained in commonly-used procedures for handling RNA and minimizing RNase contamination and use nuclease-free water (Fisher Scientific) throughout the project; students also wash down surfaces and instruments with RNase Zap (Ambion). Students determine the concentration of their RNA using the A260 value and a modified Beer-Lambert Equation, with an RNA absorption coefficient of 40 ng·cm/µL 3; students typically obtain RNA concentrations ranging from 250 to 700 ng/µL. Finally, students freeze their RNA samples at −20 °C until laboratory Period 2. If desired, students could also assess the quality of their RNA using formaldehyde agarose gel electrophoresis, as sufficient time is available in laboratory Period 2 for students to run RNA gels.
Laboratory Period 2
Students synthesize cDNA from their extracted RNA using Invitrogen SuperScript III Reverse Transcriptase (RT) and an oligo (dT)20 primer. 4 Each RT reaction contains 1000 ng of total RNA. Control samples containing no RT are prepared in parallel to assess contamination due to genomic DNA. 5 All “+RT” and “−RT” samples are stored at −20 °C until laboratory Period 3. Reverse transcription does not typically require three full laboratory hours. Therefore, students might use the remaining time in laboratory Period 2 to assess the quality of their RNA by formaldehyde agarose gel electrophoresis. Alternatively, students could set up their qPCR plates immediately following reverse transcription, condensing the project timeframe from four to three laboratory periods.
Laboratory Period 3
qPCR Set Up
Students set up qPCR reactions for relative gene expression analysis. Each student pair collectively analyzes their two cDNA samples (vern. and nonvern.) and corresponding “-RT” controls prepared during laboratory Period 2 for both the FLC target gene and the POLYUBIQUITIN-10 (UBQ10) endogenous control gene. 6 Primer sequences used for amplifications are shown in Table 1 [20, 21].
Table 1. Oligonucleotide primers used for qPCR
Primer sequence (5′ → 3′)
Product size (bp)
Product Tm (°C)
Primer sequences for the FLC target gene and the POLYUBIQUITIN-10 (UBQ10) endogenous control gene were obtained from Halliday et al.  and Czechowski et al. , respectively. Approximate melting temperature (Tm) values for PCR products were determined empirically in our laboratory.
Student pairs thaw their samples and dilute them 1:10 with nuclease-free H2O (Fisher Scientific). Each “+RT” qPCR reaction is set-up in triplicate, and each “−RT” control qPCR reaction is set-up in duplicate. For each primer pair, additional template-free H2O control reactions are also run in duplicate. For each 10-µL qPCR reaction, students combine 5 µL SYBR Green PCR Master Mix (Applied Biosystems), 1.5 µL RNase-free H2O, 0.5 µL primer mixture (10 µM forward primer and 10 µM reverse primer in RNase-free H2O), and 3 µL of template (diluted “+RT” cDNA, diluted “−RT” control, or RNase-free H2O). To minimize pipetting, students prepare primer-specific reaction mixtures containing SYBR Green PCR Master Mix, RNase-free H2O, and primer mixture for all of their reactions. Students pipet 3 µL of either diluted template cDNA, diluted “−RT” control, or RNase-free H2O into the appropriate well of a 96-well qPCR plate, followed by 7 µL of the appropriate reaction mixture. Two student pairs share one 96-well plate, although one 48-well plate would also provide sufficient wells for two student pairs. Students seal their plates with film and store them at 4 °C until the qPCR analysis, which can be carried out after class by the instructor or teaching assistant.
Each plate is analyzed by qPCR on an Applied Biosystems StepOnePlus™ instrument. Amplifications are performed using a two-step program: 10 min at 95 °C followed by 40 cycles of 15 s at 95 °C and 1 min at 60 °C. Figure 3A shows student-generated triplicate amplification curves for the amplification of FLC and UBQ10 in Arabidopsis cDNA samples. To determine the melting temperature (Tm) of each qPCR product, a melt curve analysis over the temperature range of 60–95 °C is completed at the end of 40 amplification cycles by the StepOnePlus instrument. Figure 3 shows representative melting curves for the FLC (Fig. 3C) and UBQ10 (Fig. 3D) qPCR products.
Laboratory Period 4
Product Verification and Data Analysis
Students view the melting curves and obtain Tm data for all of their qPCR reactions from the StepOnePlus instrument. Each amplification product should produce a single peak in the derivative reporter melting curve, and Tm values should be comparable to those in Table 1. Students export their CT values from the StepOnePlus instrument into Microsoft Excel and verify that any observed amplification in the H2O only samples (no-template controls) and “−RT” controls is negligible. CT values for these controls are typically >35 cycles, indicating that amplification is negligible when compared with CT values in reactions with cDNA. If time permits, students also verify the size of each amplification product by gel electrophoresis on a 1.5% agarose gel.
Students estimate the relative change in FLC transcript levels between the vern. and nonvern. plants using the method . In this method, the CT value for a target gene is first normalized for the amount of RNA added to each reverse transcription reaction. Since CT is an exponential term, this is done by subtracting the CT value for the amplification of the endogenous control gene in a sample from that of the target gene in the same sample to obtain a ΔCT value for each sample. Then, to determine the relative quantities of transcripts between two samples, the ΔCT value for a “calibrator sample” is subtracted from the ΔCT value for each sample to obtain ΔΔCT values. In the analysis, the calibrator sample is assigned a relative transcript value of unity, and the amount of transcripts in each additional sample is calculated relative to the calibrator sample . While the StepOnePlus software is capable of automatically determining relative quantities of transcripts, carrying out the calculation manually gives students a much better understanding of their overall results.
Sample calculations using student-generated data are shown in Table 2. First, students calculate the mean CT value and standard deviation for the triplicate technical replicates of FLC (target gene) and UBQ10 (endogenous control) amplifications in both vern. and and nonvern. plant samples and use these to calculate ΔCT values [Eq. (1); . Students calculate the standard deviation in ΔCT using standard error propagation methods [Eq. (2)], where sFLC and sUBQ10 are the standard deviations of the triplicate CT(FLC) and CT(UBQ10) values, respectively.
Table 2. Sample calculations using student data
Ave. FLC CT
Ave. UBQ10 CT
Details of the calculations are described in the text.
25.38 ± 0.13
16.80 ± 0.13
8.58 ± 0.18
−3.82 ± 0.18
29.18 ± 0.17
16.8 ± 0.4
12.4 ± 0.4
0.0 ± 0.4
Students calculate ΔΔCT values according to Eq. (3), using their vern. sample as the calibrator. In this way, expression of FLC in the nonvern. sample is estimated relative to FLC expression in the vern. sample. The standard deviation in ΔΔCT is the same as the standard deviation in ΔCT for each sample .
Students calculate the relative quantity (RQ) of FLC transcripts in the nonvern. sample relative to the vern. sample according to Eq. (4). Because CT is exponentially related to the number of transcript copies in a sample, the error associated with RQ is estimated by determining the range of possible RQ values. This is accomplished by calculating the minimum and maximum RQ values (RQmin and RQmax) according to Eqs. (5) and (6), respectively . Figure 3B shows a graphical representation of the student results from Table 2.
A Note About Statistical Significance of Results
There is often considerable confusion about how to apply statistics to qPCR results. While it might be tempting for students to perform a Student's t-test or confidence interval test on their data to determine the statistical significance of their observed fold change in FLC expression, it must be stressed that qPCR replicates only measure technical or method variation, and not biological or sample variation [3, 22]. Of course, independent biological replicates within the class could be used for a statistical analysis; the number of replicates depends on the statistical test being used, the reproducibility of the qPCR measurements, and the confidence range [3, 22]. In this project, the difference in FLC expression between vern. and nonvern. plants is generally so striking that a statistical test is not necessary.
In Arabidopsis (ecotype Columbia), vernalization produces a marked reduction in FLC transcripts  (Table 2; Fig. 3B) and provides a convenient system for the study of relative gene expression by undergraduate students. As a summer annual ecotype of Arabidopsis, Columbia plants have relatively low levels of FLC transcripts and display a small vernalization response relative to winter-annual ecotypes . However, the use of the Columbia ecotype in this experiment highlights the power and sensitivity of RT-qPCR for differential gene expression analysis and introduces students to the Arabidopsis ecotype most commonly used in plant research. Instructors may choose to use a winter-annual Arabidopsis ecotype, such as C24, 7 with higher FLC transcript levels and a larger vernalization-induced reduction in FLC expression .
A Note About Controls
Students use three types of controls in this experiment. Nonvern. plants are used as a “treatment” control, “−RT” samples are used to confirm that DNA amplified during the PCR process is indeed cDNA derived from RNA and not contaminating genomic DNA, and separate qPCR reactions with the constitutive gene UBQ10 are carried out to control for the amount of RNA used in each individual RT reaction.
All chemical wastes should be disposed of properly and should not be poured down the drain. RNase Zap is an eye and skin irritant. β-mercaptoethanol, used in RNA extraction, is mutagenic and a potential neurotoxin. Dithiothreitol (DTT), used in reverse transcription, is an eye, skin, and respiratory irritant. SYBR Green is an eye and skin irritant; inhalation of SYBR Green can cause respiratory, kidney, and liver damage. As a DNA intercalating agent, SYBR Green is a possible mutagen and should be used with caution.
Occasionally, students are unable to successfully extract RNA. However, most students obtain abundant RNA, and students are able to share RNA with others in the class. Some students may inadvertently switch their samples; make sure students label their tubes well! As is common with biological materials, on occasion the Arabisopsis seedlings do not grow well following vernalization of seeds. We recommend having frozen seedlings on hand for students to use. Detailed instructions for vernalizing seeds and growing Arabidopsis seedlings are provided in the Supporting Information Set-Up Sheets.
After the completion of the experiment, students prepare individual laboratory reports presenting their data and results, including representative amplification plots, melt curves, a table of CT values, sample calculations, and both tabular and graphical representations of their fold expression results. In addition, students determine whether their results are in agreement with other students in the class, as well as data presented in the primary literature . Finally, students answer specific postlaboratory questions concerning the various techniques covered in the laboratory. Sample postlaboratory questions are provided with the Supporting Information Student Handout. Students in Experimental Biochemistry II (CHEM 347) at Willamette University carried out this experiment in the fall of 2012 and a similar experiment in the fall of 2011. In general, the student responses to the postlaboratory questions matched our expectations, thereby demonstrating their understanding of both theoretical and practical aspects of gene expression analysis by RT-qPCR. Unfortunately, several of the students were unsure why it was not appropriate to use a Student's t-test to assess whether their calculated differences in gene expression were statistically significant, because they failed to recognize the difference between technical (or method) variation and biological (or sample) variation in their experiments. Furthermore, a few students failed to understand the importance of the “−RT” control sample in assessing the presence or absence of contamination from genomic DNA in their PCR reactions. To address these misconceptions, we will focus more of our attention on these important concepts in prelaboratory discussions.
Many of these students who completed this project did not have prior experience with molecular biology techniques, and they appreciated getting hands-on experience with RT-qPCR. Some of our students expressed an interest in designing their own PCR primers for the experiment. In the future, we may incorporate exercises in PCR primer design [24, 25] at the beginning of this laboratory. In addition, several students expressed a desire to design their own RT-qPCR experiment. A major strength of this project is that it allows students to develop the necessary skills required for future laboratory projects on the regulation of gene expression in Arabidopsis plants (see below). Students gain experience in several general biochemistry/molecular biology laboratory skills , including experimental design, data collection and analysis, statistical considerations of biochemical data, the proper use of controls, and the use of a commercial kit to isolate RNA from plants, as well as writing and reporting scientific results. They also learn the more specialized techniques of reverse transcription and quantitative analysis of gene expression by qPCR.
As presented, the RT-qPCR project described here is a “structured inquiry” laboratory experience. In structured inquiry, the instructor provides the problem and the procedure, but students collect and analyze their own data and draw their own conclusions. Ideally, students are not informed of the desired outcome ahead of time . Moreover, this project provides the foundation for students to transition to “guided-inquiry” activities. In guided inquiry, the instructor provides a problem but students are tasked with devising their own experimental procedure . As a possible guided-inquiry project, each student pair could analyze FLC transcript levels in a different Arabidopsis ecotype. 8 Hundreds of different naturally-occurring Arabidopsis ecotypes have been collected from all over the world, and levels of FLC transcripts vary greatly among natural ecotypes . Students, guided by their instructor, could develop hypotheses about FLC transcript levels and vernalization requirements based on the climate of origin of their ecotype. Students in our Experimental Biochemistry II course during the Fall of 2011 used RT-qPCR to compare FLC transcript levels in two-week-old Columbia (small vernalization response) and C24 (large vernalization response) seedlings. These students observed much higher FLC transcript levels in the C24 ecotype (data not shown). Another worthy guided-inquiry project for undergraduate students would be to correlate vernalization time, FLC transcript levels, and flowering time in different Arabidopsis ecotypes. For example, Sheldon et al.  demonstrated a correlation between the duration of vernalization, the reduction in FLC transcripts, and the time to flowering in the C24 ecotype of Arabidopsis. Students could devise their own procedure, based on the project described here, to investigate whether there is a correlation in these three parameters in additional ecotypes, as well. Note that flowering time in Arabidopsis plants is commonly measured as the developmental age at flowering, determined by counting the number of leaves once the plant has bolted .
The project described here could also initiate a number of “open inquiry” research projects, where students develop both a problem to investigate and an appropriate procedure to investigate the problem . For example, some research suggests that short-term cold treatment during the day (cold stress) delays the flowering of Arabidopsis plants by increasing the expression of FLC . In addition, several hormone signaling pathways, such as the brassinosteroid signaling pathway  and the ethylene signaling pathway (A. Fisher, G. Eickelberg, and G. Laudenbach, unpublished), also reduce FLC transcript levels in Arabidopsis plants to help promote the floral transition. Finally, the expression of FLC in Arabidopsis is regulated, in part, by covalent chromatin modifications that are stable through mitotic divisions and confer onto plants an “epigenetic memory” of winter [30-32]. The regulation of FLC in Arabidopsis plants, therefore, is a fascinating biological system that lends itself extremely well to student-centered open inquiry projects at the undergraduate level.
The experiment described here provides upper-level biochemistry or molecular biology students with experience using RT-qPCR to analyze the relative expression of a target gene. Students complete the entire multistep RT-qPCR process, including data analysis with the method, in the context of a robust “real-world” biological system: the environmental influence on gene expression controlling flowering time in Arabidopsis thaliana plants. Appropriate controls are used throughout the experiment, providing students with experience in the proper design of gene expression studies by RT-qPCR. Students should come away from the project with a better understanding of gene expression analysis by RT-qPCR, the ability to critically review RT-qPCR experiments in the scientific literature, and the skills and confidence to carry out gene expression analysis by RT-qPCR in the future.
This research was supported by a Research Start-Up Grant for New Science Faculty from the M. J. Murdock Charitable Trust (Reference Number 2009188:JVZ:11/19/2009). This material is based on work supported by the National Science Foundation under Grant No. IOS-1061750 and Grant No. DUE-1044737. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. The authors thank Dr. Karen Hinkle for sharing a protocol for reverse transcription, Drs. Rebecca Laurie and Mauren Jaudal for help with qPCR, and Dr. Sarah Kirk and Geri Laudenbach for help running the laboratory in the Experimental Biochemistry II course at Willamette University. The authors are grateful for helpful comments on the manuscript from Drs. Melissa Marks, Jason Duncan, Todd Silverstein, and Sarah Kirk.
reverse transcription/reverse transcriptase
quantitative polymerase chain reaction
FLOWERING LOCUS C
To reduce reagent costs, each student pair might analyze only one plant sample (vern. or nonvern.) or extract total RNA using TRIzol reagent according to Chomczynski and Sacchi .
This step is optional.
User's manual for the NanoDrop ND-1000 spectrometer available at http://www.nanodrop.com/Library/nd-1000-v3.8-users-manual-8%205x11.pdf.
To reduce reagent costs, RNA samples from the entire class could be pooled prior to reverse transcription.
While the primer pair for amplification of FLC spans introns, the primer pair for amplification of UBQ10 does not. Therefore, it is essential that students analyze a “-RT” control in parallel to assess contamination from genomic DNA.
To minimize reagent costs and instrument time, cDNA from the entire class could be pooled for one qPCR analysis.
Arabidopsis C24 seeds are available in bulk from Lehle Seeds (www.arabidopsis.com).
Seeds of a large number of Arabidopsis ecotypes are available for a nominal fee from the Arabidopsis Biological Resources Center (ABRC) at Ohio State University.