Analysis of Relative Gene Expression Data Using Real Time Qpcr

8/3/2019 Analysis of Relative Gene Expression Data Using Real Time Qpcr

1/7

METHODS 25, 402408 (2001)

doi:10.1006/ meth.2001.1262, available online at http:/ / www.idealibrary.com on

Analysis of Relative Gene Expression Data Using Real-

Time Quantitative PCR and the 2

CT

MethodKenneth J . Livak* and Thomas D. Schmittgen,1

*Applied B iosystems, Foster City, California 94404; and Departm ent of Pharm aceutical S ciences, College of Pharm acy,Washington State University, Pullman, Washington 99164-6534

o f t h e t a rg e t g e n e re l a t i v e t o s o m e re fe re n c e g ro u pThe two most commonly used methods to analyze data from

such as an u nt reat ed cont rol or a sa mple at t ime zeroreal-time, quantitative PCR experiments are absolute quantifica-in a time-course study.tion and relative quantification. Absolute quantification deter-

Absolute qu an tificat ion sh ould be performed in sit u-mines the input copy number, usually by relating the PCR signalto a standard curve. Relative quantification relates the PCR signal ations where i t is necessary t o determ ine the a bsoluteof the target transcript in a treatment group to that of another transcript copy number. Absolute quantification hassample such as an untreated control. The 2CT method is a been combined with real-time PCR an d nu merous re-convenient way to analyze the relative changes in gene expression

port s ha ve appeared in th e l itera tu re (69) includingfrom real-time quantitat ive PCR experiments. The purpose of t his

two articles in this issue (10, 11). In some situations,report is to present the derivation, assumptions, and applicationsi t may be unnecessary to determine the absolut e tran -of the 2CT method. In addition, we present the derivation andscript copy nu mber and reporting th e relative chan geapplications of two variations of the 2CT method that may be

useful in the analysis of real-time, quantitative PCR data. 2001 in gene expression will suffice. For example, statingElsevier Science (USA) t h a t a g i v e n t re a t m e n t i n c re a s e d t h e e x p re s s i o n o f

Key Words: reverse transcription polymerase chain reaction; gene x by 2.5-fold may be m ore r elevant th an stat ingquantitative polymerase chain reaction; relative quantification;

tha t t he tr eatmen t increased the expression of gene xreal-time polymerase chain reaction; Taq Man.from 1000 copies to 2500 copies per cell.

Qua nt ifying the r elative cha nges in gene expression

u s i n g r e a l-t i m e P C R re q u ire s ce rt a i n e qu a t i on s , a s -

s u m p t i o n s , a n d t h e t e s t i n g o f t h e s e a s s u m p t i o n s t oReserve transcription combined with the polymer- properly ana lyze the dat a. The 2CT method may be

ase chain reaction (RT-PCR) has proven to be a power- used to calculate relative changes in gene expressionful method to quantify gene expression (13). Real- determined from real-time quantitative PCR experi-t i m e P C R t e c h n o l o g y h a s b e e n a d a p t e d t o p e rfo rm m e n t s . De ri v a t i o n o f t h e 2CT equation, includingqua nt itat ive RT-PCR (4, 5). Two differen t met hods of a s s u m p t i on s , e xp e rim e n t a l d e si gn , a n d v a li da t i onanalyzing data from real-time, quantitative PCR ex- test s, ha ve been d escribed in Applied Biosystem s User

periments exist: absolute quantification and relative Bullet in N o. 2 (P/N 4303859). Analyses of gene expr es-qua nt ificat ion. Absolut e quan tificat ion deter mines th es i o n d a t a u s i n g t h e 2CT m e t h o d h a v e a p p e a re d i n

input copy number oft he tra nscript of interest, usuallyt h e l it e ra t u r e (5 , 6). Th e p u rp o se o f t h i s r e p ort i s t o

by relating t he PCR signal to a sta ndar d curve. Rela-present t he derivation of the 2CT method, assump-tive quan tification d escribes th e cha nge in expressiontions involved in using t he meth od, an d a pplications

of this met hod for the general l itera tur e. In a ddition,

we present t he derivat ion an d applicat ion of two varia-

tions of the 2CT m e t h o d t h a t m a y b e u s e fu l i n t h e1 To whom r equests for reprints should be a ddressed. Fax: (509)335-5902. E-mail: Schm itt g@ma il.wsu.edu. ana lysis of real-time qua ntita tive PCR data.

402 1046-2023/01 $35.00

2001 Elsevier Science (USA)All rights reserved.


2/7

ANALYSIS OF REAL-TIME PCR DATA 403

or1. THE 2CT METHOD

XN (1 E)CT K, [6]

1.1. Derivation of t he 2CT Methodwhere XN is equal to the normalized amount of targetThe equat ion th at describes the exponent ial amplifi-(X0/R 0) and CT is equal t o the difference in th resholdcation of PCR iscycles for target and reference (CT,X CT,R).

Rearr anging gives th e expressionXn X0 (1 EX)n, [1]

XN K (1 E)CT. [7]

where Xn is the number of target molecules at cycleThe final step is to divide the XN for any sample q byn of the reaction, X0 i s t h e i n i t i a l n u m b e r o f t a rg e tt h e XN for the calibrator (cb):molecules. EX is the efficiency of target amplification,

a n d n is th e nu mber of cycles. The th resh old cycle (CT)

i n di ca t e s t h e fra ct i on a l cy cl e n u m b e r a t wh ich t h e XN, q

XN,cb

K (1 E)CT,q

K (1 E)CT,cb (1 E)CT. [8]

amount of amplified target reaches a fixed threshold.

Thus,

Here CT (CT,q CT,cb).

For a mplicons designed t o be less th an 150 bp and forXT

X0

(1

EX)CT,X

KX [2] which the primer and Mg2+ concentra tions ha ve been

pr operly optim ized, th e efficiency is close to one. Th ere-where XT is the thr eshold n um ber of tar get molecules, fore, the amount of ta rget, norm alized to an endogenousCT,X is the threshold cycle for target amplification, and reference an d r elat ive t o a calibrator, is given byKX is a consta nt . A similar equ at ion for t he end ogenous

reference (internal control gene) reaction is amount of target 2CT. [9]

R T R 0 (1 ER)CT,R KR, [3]

1.2. Assumptions and Applications of the 2CT Method

For th e CT calculation to be valid, th e amplificat ionwhere R T is t he thr eshold n umber of r eference mole-efficiencies of th e ta rget an d r eferen ce must be appr oxi-cules, R 0 is t he initial nu mber of reference molecules,

mately equal. A sensitive method for assessing if twoER is the efficiency of reference amplification, CT,R isamplicons have the same efficiency is to look at howthe threshold cycle for reference amplification, and KRCT varies with template dilution. Figure 1 shows theis a constan t.

Dividing XT by R T gives the expression

XT

R T

X0 (1 EX)CT,X

R 0 (1 ER)CT,R

KX

KR K. [4]

For r eal-time a mplificat ion using TaqMan probes, th e

exact values ofXT a n d R T depend on a n um ber of factors

including the reporter dye used in the probe, the se-quen ce cont ext effects on t he fluorescence propert ies of

the probe, the efficiency of probe cleavage, purity of

the probe, and setting of the fluorescence threshold.

Therefore, t he const ant K does not have to be equal toFIG. 1. Validation of th e 2CT meth od: Amplification of cDNAone. Assuming efficiencies of the tar get a nd th e r efer-synt hesized from different am ount s of RNA. The efficiency of am plifi-

ence are the sa me,cation of the target gene (c-m yc) and internal control (GAPDH) wasexamined using r eal-time PCR and TaqMan detection. Using reversetranscriptase, cDNA was synthesized from 1 g total RNA isolatedEX ER E,from human Raji cells. Serial dilutions of cDNA were amplified byreal-time PCR using gene-specific primers. The most concentratedsample contained cDNA derived from 1 n g of total RNA. The CT(CT,cm yc CT,GAPDH ) was calculated for each cDNA dilution. The data

X0

R 0

(1 E

)

CT,XCT,R K, [5] were fit using least-squares linear regression analysis (N 3).


3/7

LIVAK AND SCHMITTGEN404

results of a n experiment where a cDNA prepara tion chan ge in gene expression relative to an untreat ed con-

tr ol, for example, if one wan ted to determine t he expres-was diluted over a 100-fold range. For each dilution

sample, amplificat ions were perform ed using pr imers sion of a part icular mRNA in an organ. In these cases,

th e calibra tor ma y be th e expr ession of th e same mRNAan d fluorogenic pr obes for c-m yc an d GAPDH. The aver-

a ge CT was calculated for both c-m yc a n d GAP DH a n d in a n ot h e r org a n . Ta b le 1 p re s e n t s m e a n CT values

determined for c-m yc and GAPDH t ran scripts in totalt h e CT (CT,myc CT,GAPDH ) was determined. A plot of

the log cDNA dilution versus CT was made (Fig. 1). RNA samples from bra in and kidney. The brain was

ar bitr ar ily chosen as th e calibra tor in this example. TheIf the absolute value of the slope is close to zero, theefficiencies of th e tar get and r eference genes are simi- amount of c-m yc, normalized to GAPDH an d relative to

brain, is reported. Although the relative quan titativelar, an d the CT calcula tion for t he r elative quan tifica-

tion of ta rget m ay be used. As shown in Fig. 1, th e slope met hod can be used to ma ke this type of tissu e compar i-

son, biological interpretation of the results is complex.of the line is 0.0471; therefore, the assumption holds

a n d t h e CT method may be used to analyze the data. The single r elative quantity r eported actua lly reflects

varia tion in both t ar get and reference tr an scripts acrossIf the efficiencies of the two amplicons are not equal,

t h e n t h e a n a l ys is m a y n e ed t o b e p erform e d v ia t h e a v a rie t y of ce ll t y p es t h a t m i gh t b e p res en t i n a n y

particular tissue.absolut e quan tification meth od using st an dard cur ves.

Alter na tively, new primer s can be designed a nd/or opti-

mized to achieve a similar efficiency for the target and1.4. Data Analysis Using the 2CT Methodreference amplicons.

Th e CT values provided from real-time PCR instru-

mentation are easily imported into a spreadsheet pro-1.3. Selection of Internal Control and Calibrator for the gram such a s Microsoft E xcel. To demonst ra te th e an al-2CT Method ys is , d a t a a r e r e por t e d fr om a q u a n t it a t ive gen e

expression experiment a nd a sam ple spreadsheet is de-The pu rpose of th e int ern al contr ol gene is to norm al-

ize t he P CRs for th e am oun t of RNA added t o the re- scribed (Fig. 2). The chan ge in expression of th e fosglo

m yc target gene normalized to -actin was monitoredv ers e t ra n s cri pt i on re a ct i on s . We h a v e fou n d t h a t

stan dar d housekeeping genes usua lly suffice as inter- over 8 h . Triplicat e sam ples of cells were collected a t

each time point . Real-time P CR was perform ed on t henal control genes. Suitable internal controls for real-

time quantitative PCR include GAPDH, -actin, 2- corresponding cDNA synth esized from each sample.

The dat a were ana lyzed using Eq. [9], where CT microglobulin, and rRNA. Other housekeeping genes

will u n d ou b t ed ly work a s wel l. I t i s h i gh l y re com - (CT,Target CT,Actin)Time x (CT,Target CT,Actin)Time 0. Timex is any time point an d Time 0 repr esent s the 1 expres-mended t ha t t he inter na l cont rol gene be properly vali-

dat ed for each experiment to determine t ha t gene ex- sion of th e ta rget gene normalized to -actin. The mea n

CT values for both t he ta rget a nd int erna l cont rol genespression is unaffected by the experimental treatment.

A method to validate the effect of experimental tr eat- were determined at t ime zero (Fig. 2, column 8) and

were used in E q. [9]. The fold cha nge in th e tar get gene,ment on the expression of the internal control gene is

descr ibed in Sect ion 2.2. n or m a lized t o -actin an d r elat ive to th e expression at

time zero, was calculated for each sample using Eq. [9]The choice of calibrator for the 2CT method de-

pends on the type of gene expression experiment th at (Fig. 2, column 9). The mean, SD, a nd CV are then

determined from the triplicate samples at each timeone has planned. The simplest design is to use the un-

t re a t e d con t ro l a s t h e ca l ib ra t or. Us i n g t h e 2CT point. Using this analysis, the value of the mean fold

change at time zero should be very close to one (i.e.,method, the data are presented as the fold change ingene expression norma lized to an endogenous reference since 20 1). We have found the verification of the

mea n fold cha nge at time zero to be a convenient met hodgene an d r elative to the un treated contr ol. For the u n-

tr eated cont rol sample, CT equals zero and 20 equals to check for errors and variation am ong the triplicate

samples. A value that is very different from one sug-one, so that the fold change in gene expression relative

to the unt reated contr ol equals one, by definition. For gests a calculation error in the spreadsheet or a very

high degree of experiment al variation.th e tr eated sa mples, evaluation of 2CT indicates t he

fold chan ge in gene expression relative to th e un tr eated In the preceding example, thr ee separa te RNA prepa-

ra t i on s were m a d e for e a ch t i m e p oi n t a n d ca rri edcont rol. Similar an alysis could be applied to st udy t he

time course of gene expression where the calibrator thr ough the analysis. Therefore, i t made sense to treat

each sample separately and average the results aftersample represents th e amount of transcript t hat is ex-

pr essed a t t im e zer o. t h e 2CT calculation. When replicate PCRs are run

on the same sample, it is more appropriate to averageS it u a t i on s e xi st wh ere on e m a y n ot com p a re t h e


4/7


CT data before performing the 2CT calculation. Ex- (GAPDH) were am plified in separate wells. There is

no reason to pair any particular c-m yc well with anyactly how th e averaging is perform ed depends on if th e

tar get and reference a re amplified in separate wells particular GAPDH well. Therefore, it ma kes sense to

average the c-m yc and GAPDH CT values separatelyo r i n t h e s a m e we l l . Ta b l e 1 p re s e n t s d a t a fro m a n

e xp eri m en t wh ere t h e t a rg et (c-m yc) a n d re fe ren ce b efore p erform i n g t h e CT calculation. The variance

TABLE 1

Treatment of Replicate Data Where Target and Reference Are Amplified in Separate Wells a

CT (Avg. c-m y c C T CT (Avg. CT Normalized c-m yc amountTissu e c-m y c C T GAPDH CT Avg. GAPDH CT Avg. CT,Brain) r ela t ive t o br a in 2

CT

Br a in 30.72 23.7030.34 23.5630.58 23.4730.34 23.6530.50 23.6930.43 23.68

Aver age 30.49 0.15 23.63 0.09 6.86 0.17 0.00 0.17 1.0 (0.9 1.1)

Kidn ey 27.06 22.7627.03 22.6127.03 22.6227.10 22.6026.99 22.6126.94 22.76

Aver age 27.03 0.06 22.66 0.08 4.37 0.10 2.50 0.10 5.6 (5.3 6.0)

a Total RNA from hum an brain and kidney were purchased from Clontech. Using reverse tra nscriptase, cDNA was synthesized from 1g total RNA. Aliquots of cDNA were used a s tem plate for real-time PCR r eactions conta ining either primer s an d pr obe for c-m yc or primersand probe for GAPDH. Each reaction contained cDNA derived from 10 ng total RNA. Six replicates of each reaction were performed.

FIG. 2. Sample spreadsheet of data analysis using the 2CT method. The fold change in expression of the target gene (fosglomyc)relative to t he internal contr ol gene (-actin) at various time points was st udied. The samples were ana lyzed using real-time qua ntitativeP C R a n d t h e Ct data were imported into Microsoft Excel. The mean fold change in expression of the ta rget gene a t each time point wascalculat ed us ing Eq. [9], where CT (CT,Target C,Actin)Time x (CT,Target C,Actin)Time 0. The mean CT at time zero ar e shown (colored boxes)

as is a sample calculation for the fold change using 2CT (black box).


5/7


e st i m a t ed from t h e re p li ca t e CT values is carr ied of an arbitrary constan t. This gives results equivalent

to those reported in Fig. 2 wher e CT values for n onrepli-through to the final calculation of relative quantities

using stan dar d propagation of error meth ods. One diffi- cated samples were carr ied thr ough th e ent ire 2CT

calculation before averaging. Alternatively, it is possi-culty is tha t CT is exponen tially relat ed to copy nu mber

(see Section 4 below). Thu s, in the fina l calcula tion, th e ble to report resu lts with the calibra tor qua nt ity defined

a s 1 without any error. In this case, the error esti-error is estimated by evaluating the 2CT term using

CT plus the stan dard deviation a nd CT m i n u s t h e m a t e d for t h e a v era g e CT,cb value must be propagated

into each of the CT values for the test samples. Instan dar d deviation. This leads to a r an ge of values th atis asymmetrically distributed relative to the a verage Table 1, the CT value for the kidney sample would

become 2.50 0.20 an d th e norma lized c-m yc amountvalue. The asymmetric distribution is a consequence of

converting the results of an exponent ial process int o a would be 5.6 with a range of 4.9 to 6.5. Results for

brain would be reported as 1 without any error.linear compar ison of amount s.

By usin g probes labeled with dist inguish able reporter

dyes, it is possible to ru n th e tar get and reference am pli-

fications in th e sam e well. Table 2 pr esents data from 2. THE 2CT METHODan experiment where the tar get (c-m yc) and reference

(GAPDH) were am plified in th e sam e well. In an y par-2.1. Derivation of t he 2CT Methodticular well, we know that the c-m yc reaction and the

GAPDH r eaction h ad exactly the same cDNA input. Norma lizing to an endogenous reference provides a

meth od for corr ecting results for differing a mount s ofTherefore, it makes sense to calculate CT separately

for each well. These CT values can then be averaged input RNA. One hallmark of the 2CT method is th at

i t u s e s d a t a g e n e ra t e d a s p a rt o f t h e re a l -t i m e P C Rbefore proceeding with the 2CT calculation. Again,

the estimated error is given as an asymmetric ran ge of experiment to perform this normalization function.

This is particularly attractive when it is not practicalvalues, r eflectin g conversion of an exponen tia l varia ble

t o a lin ea r com pa r ison . t o m ea su r e t h e a m ou n t of in pu t RNA by ot h er m et h ods.

Such situations include when only limited amounts ofIn Tables 1 a nd 2, th e estimated error has not been

increased in proceeding from the CT column to th e RNA ar e available or when high-th roughput processing

of many samples is desired. It is possible, though, toCT column . This is because we h ave decided t o dis-

play the data with err or shown both in th e calibrator normalize to some measurement external to the PCR

experiment . The most comm on meth od for norma liza-an d in the test sam ple. Subtr action oft he average CT,cb

to determine th e CT value is treated as subtra ction tion is to use UV absorbance to determine the amount

TABLE 2

Treatment of Replicate Data Where Target and Reference are Amplified in the Same Wella

c-m yc CT (Avg. c-m y c C T CT (Avg. CT Normalized c-m yc amountTissue CT GAPDH CT Avg. GAPDH CT) Avg. CT,Brain) r ela t ive t o br a in 2

CT

Br a in 32.38 25.07 7.3132.08 25.29 6.7932.35 25.32 7.03

32.08 25.24 6.8432.34 25.17 7.1732.13 25.29 6.84

Aver a ge 6.93 0.16 0.00 0.16 1.0 (0.9 1.1)Kidn ey 28.73 24.30 4.43

28.84 24.32 4.5228.51 24.31 4.2028.86 24.25 4.6128.86 24.34 4.5228.70 24.18 4.52

Aver a ge 4.47 0.14 2.47 0.14 5.5 (5.0 6.1)

a An experiment like that described in Table 1 was performed except the reactions contained primers and probes for both c- m yc a n dGAPDH. The probe for c-m yc was labeled with the reporter dye FAM and the probe for GAPDH was labeled with the reporter dye JOE.Becau se of th e different r eporter dyes, th e real-time PCR signals for c-m yc an d GAPDH can be distinguished even th ough both am plifications

are occurring in the same well.


6/7


of RNA added to a cDNA reaction. PCRs are then set equation where CT CT,Time x CT,Time 0 (Fig. 3). A sta-

tistically significan t relationsh ip exists between th eup u sing cDNA derived from t he sam e amount of inpu t

RNA. One example of using this external normalization treatm ent and expression of GAPDH but not for 2-

microglobulin (Fig. 3). Therefore, 2-microglobulinis to study th e effect of experiment al tr eatm ent on t he

expression of an endogenous reference to determ ine if mak es a suitable intern al cont rol in quan titat ive serum

stim ulat ion st udies while GAPDH does not. This exam -th e intern al contr ol is affected by trea tmen t. Thu s, the

t a rg et g en e a n d t h e e n d og en ou s r e fe ren ce a re on e i n p le d em on s t ra t e s h ow t h e 2CT method may be used

to analyze relative gene expression data when only oneth e sam e. In t his case, Eq. [2] is n ot divided by Eq. [3]a n d E q. [5] becom es gen e is bein g st u died.

X0 (1 EX)CT,X KX. [10]

3. STATISTICAL ANALYSIS OF REAL-TIMEPCR DATARearra nging gives t he expression

The endpoint of real-time PCR an alysis is the t hr esh-X0 KX (1 EX)

CT,X. [11]old cycle or CT. The CT is determ ined from a loglinear

plot of the PCR signal versus th e cycle num ber. Thus,Now, dividing X0 for any sample q b y t h e X0 for the CT is a n e xp on e n t ia l a n d n ot a l in e a r t e rm . F or t h i s

calibrator (cb) gives reason, any statistical presentation using the raw CT

X0,q

X0,cb

KX (1 EX)CT,q

KX (1 EX)CT,cb

(1 EX)CT, [12]

where CT is equal to CT,q CT,cb. The prime is used

to distinguish this expression from the previous CTcalcula tion (see Eq. [6]) th at involved su btr action ofCTvalues for target and reference.

As stated in Section 1.1, if properly optimized, the

efficiency is close to one. The amount of endogenous

reference relative to a calibrator th en becomes

2CT. [13]

2.2. Application of t he 2CT Method

An appr opriate applicat ion of the 2CT method is to

determine the effect of the experimental treatment on

th e expression of a candidat e int erna l cont rol gene. To

demonstra te this ana lysis, serum st ar vation an d induc-

tion experim ent s were perform ed (7). Seru m sta rva tion/induction is a commonly used model to study the decayFIG. 3. Application of the 2C

T meth od. The following experimentof certain mRNAs (8). However, serum may alter the

was conducted to validate the effect of treatment on the expressione xp re ss ion of n u m e rou s ge n es i n cl u di n g s t a n d a rd of candidate int erna l control genes. NIH 3T3 fibroblasts were seru mhousekeeping genes (9). star ved for 24 h and th en indu ced with 15% seru m over an 8-h period.

Samples were collected at various times following serum stimulation;Gene expression was induced in NIH 3T3 cells bymRNA was extracted a nd converted to cDNA. The cDNA was sub-adding 15% serum following a 24-h period of serum

jected t o real-time qua ntita tive PCR u sing gene-specific primers forstarvation. Poly(A)+ RNA was extr acted from th e cells 2-microglobulin and GAPDH. The fold change in gene expressionan d equivalent amount s were converted to cDNA. The was calculat ed using Eq. [13], where CT (CT,Time x CT,Time 0) a nd

is presented for both 2-microglobulin (A) an d GAPDH (B). Reprin tedamounts of2-microglobulin and GAPDH cDNA werefrom T. D. Schmit tgen a nd B. A. Zak ra jsek (2000)E ffect of experim en-determined by real-time quantitative PCR with SYBRtal treatment on housekeeping gene expression: Validation by real-

Green detection (7). The relative am oun ts of2-micro- time quantitative RT-PCR, J. Biochem. Biophys. Methods 46 , 6981,with permission of Elsevier Science.g l o b u l i n a n d GAP DH a re p re s e n t e d u s i n g t h e 2CT


7/7


values should be avoided. As described within th e previ- should be used. Other wise, presen ta tion of th e relative

gene express ion should suffice. Relative qu an tificationous sections of th is ar ticle, present at ion of relat ive PCRm a y b e e a s i e r t o p e rfo rm t h a n t h e a b s o l u t e m e t h o ddata is most often calculated along with an internalbecause the use of standard curves is not required.cont rol an d/or calibra tor sam ple an d is rar ely presen ted

The equations provided herein should be sufficient fora s t h e CT . An exception is when one is interested inan investigator to analyze quantitative gene expressionexamin ing the sam ple-to-sam ple var iation among repli-data using r elative quan tification. To sum mar ize thecate reactions.

i m p o rt a n t s t e p s i n t h e d e s i g n a n d e v a l u a t i o n o f t h eTo demonstrate this, 96 replicate reactions of theexperiment : (i) select an internal control gene, (ii ) vali-identical cDNA were performed using real-time PCRd a t e t h e i n t e rn a l c o n t ro l t o d e t e rm i n e t h a t i t i s n o ta n d S YB R Gree n d et e ct i on . A m a s t e r m i xt u re con -affected by experimental treatm ent, a nd (ii i) PCR onta ining all of the ingredients was pipetted into individ-perform dilut ions of RNA or cDNA for both th e t ar getual tubes of a 96-well reaction plate. The samples werean d int ern al contr ol genes to ensu re th at t he efficienciessubjected to real-time PCR an d th e individual CT valuesar e similar. Fin ally, stat istical data s hould be conver tedwere determined. To examine the intrasample varia-to the l inear form by the 2CT calculation and shouldtion, the mea n SD was determined from t he 96 sam-not be presented by the raw CT values.ples. If calculated from the raw CT the mean SD was

20.0 0.194 with a CV of 0.971%. However, when th e

individual CT values were converted to th e linear form

REFERENCESu s in g t h e t e rm 2C

T, the mean SD was 9.08 10

7

1.33 107 with a CV of 13.5%. As demonstrated1. Mur ph y, L. D., Herzog, C. E., Rudick, J . B., Fojo, A. T., and Ba tes,

by this simple example, reporting the data obtainedS. E. (1990) Biochemistry 29 , 1035110356.

from the ra w CT values falsely represents t he var iation 2. Noona n, K. E ., Beck, C., Holzmayer, T. A., Chin, J . E., Wunder,an d sh ould be a voided. Converting t he individual dat a J . S., Andru lis, I. L., Gazdar, A. F., Willman, C. L., Griffith, B.,

Von-Hoff, D. D., a nd Robinson, I. B. (1990) Proc. Natl. Acad. Sci.to a linear form using 2CT more accur ately depicts th eU S A 87 , 71607164.

individual var iation among r eplicate reactions.3. Horikoshi, T., et al. (1992) Cancer Res. 52 , 108116.

4. Heid, C. A., Stevens, J ., Livak, K. J ., and Williams, P. M. (1996)Genome Res. 6, 986994.

5. Winer, J . , J ung, C. K., Shackel, I. , a nd Williams, P. M. (1999)4. CONCLUDING REMARKS Anal. Biochem. 270, 419.

6. Schmit tgen , T. D., Zak ra jsek, B. A., Mills, A. G., Gorn, V., Singer,

Experimental design and data analysis from real- M. J ., and Reed, M. W. (2000) Anal. Biochem. 285, 194204.time, quantitative PCR experiments may be achieved 7. Schmittgen, T. D., and Zakr ajsek, B. A. (2000) J . Biochem. Bio-

phys. Methods 46 , 6981.using either relative or absolut e qua nt ificat ion. When8. Chen , C. Y., and Sh yu, A. B. (1994)Mol. Cell. Biol. 14 , 84718482.designing quantitative gene expression studies using9. Iyer, V. R., et al. (1999) S cience 283, 8387.

real-time PCR, the first question that an investigator10. Giuliett i, A., Overbergh , L., Valckx, D., Decallone, B., Bouillon,

should ask is h ow should th e data be present ed. If abso- R., and Mathieu, C. (2001) Methods 25 , 386401.11. Niesters, H. G. M. (2001) Methods 25 , 419429.lute copy num ber is required, then t he absolut e meth od

Documents

Analysis of Relative Gene Expression Data Using Real Time Qpcr