1 Title: Comparison of Different Standards for Real Time PCR-based Absolute

2 Quantification

4 Authors: S. Dhanasekaran a, b, T. Mark Doherty c, d, John Kenneth a,* TB Trials Study

5 Group

6 Authors’ Addresses: a Infectious Diseases Unit, St.John’s Research Institute,

7 Koramangala, Bangalore – 560 034, India.

b
8 The Gade Institute, Section for Microbiology and Immunology, University of Bergen and

9 Haukeland University hospital, N-5021, Norway.

c
10 Infectious Disease Immunology, Statens Serum Institute, Artillerivej 5, Copenhagen

11 2300s, Denmark.
d
12 Center for International Health, University of Bergen, N-5021, Norway.

13

14

15

16 *Corresponding Author

17 Dr. John Kenneth

18 Associate Professor

19 Infectious Diseases Unit,

20 St.John’s Research Institute,

21 Koramangala, Bangalore - 560 034, India.

22 Phone: 91-80-22065059; Fax: 91-80-25501088

23 Email- johnkennt@gmail.com
1
 1 Abstract

2 Quantitative real-time PCR (qPCR) is a powerful tool used for both research

3 and diagnostic, which has the advantage, compared to relative quantification, of

4 providing an absolute copy number for a particular target. However, reliable standards

5 are essential for qPCR. In this study, we have compared four types of commonly-used

6 standards - PCR products (with and without purification) and cloned target sequences

7 (circular and linear plasmid) for their stability during storage (using percentage of variance

8 in copy numbers, PCR efficiency and regression curve correlation co-efficient (R2)) using

9 hydrolysis probe (TaqMan) chemistry. Results, expressed as copy numbers / µL, are

10 presented from a sample human system in which absolute levels of HuPO (reference gene)

11 and the cytokine gene IFN-γ were measured. To ensure the suitability and stability of the

12 four standards, the experiments were performed at 0, 7 and 14 day intervals and repeated 6

13 times. We have found that the copy numbers vary (due to degradation of standards) over

14 the period of time during storage at 4oC and -20oC, which affected PCR efficiency

15 significantly. The cloned target sequences were noticeably more stable than the PCR

16 product, which could lead to substantial variance in results using standards

17 constructed by different routes. Standard quality and stability should be routinely

18 tested for assays using qPCR.

19

20 Key Words: Quantitative real-time PCR; Absolute quantification; Percentage of Variance;

21 PCR efficiency; PCR standards; Cytokine.

22 Abbreviations: qPCR – Quantitative real-time PCR; Ct - Cycle of threshold; R2 –

23 Correlation coefficient; HuPO – Human acidic ribosomal phosphoprotein; IFN γ –

24 Interferon gamma.

2
 1 1. Introduction

2 Quantitative PCR (qPCR) is used to simultaneously amplify, detect and quantify

3 nucleic acids. Quantification can be achieved either by absolute or relative quantification.

4 In turn, relative quantification is obtained either by the comparative Ct method or the

5 standard curve method. Both these methods measure the target and reference gene

6 expression only in relative terms by using a calibrator, and may hence miss relevant

7 biological information (Leong et al., 2006). The absolute quantification method relies on a

8 standard plot constructed from known concentrations of standards to measure the actual

9 copy numbers of a particular target, and is therefore considered to be more informative

10 and reliable for comparisons. But the quantity of target in samples can be evaluated with

11 reasonable accuracy only by using a properly characterized standard (Jie et al., 2009).

12 A variety of materials have been used as standards, including PCR-amplified target

13 sequences (Leong et al., 2006; Rose’Meyer et al., 2003), plasmids containing the target

14 sequence (Lee et al., 2006; Whelan et al., 2003) or commercially prepared DNA (Dworkin

15 et al., 2002). Generating a reliable standard requires extensive effort, as illustrated by

16 Collins et al., (1995) who reported the use of an absolute RNA standard for viral

17 quantification.

18 The reliability of standards for quantitative PCR is a crucial issue and a great deal

19 of effort has been expended on identifying appropriate target genes for analysis (Dheda et

20 al., 2004). Once made, standards are typically used for extended periods of time and

21 stability of standards is therefore a critical, but overlooked, issue. This is particularly

22 important when choosing appropriate standards for different experimental designs such as

23 long term studies (where multiple samples must be compared over time with great

24 accuracy) as well as for short term studies. Therefore, we have evaluated 4 types of

3
1 commonly-used DNA standards constructed for HuPO and IFN γ: PCR sequences (with

2 and without purification) and cloned target sequences (circular and linear plasmid) using

3 quantitative real-time PCR. These four standard types were serially diluted and assayed for

4 suitability and stability with different storage conditions using the absolute quantification

5 method.

4
 1 2. Material and Methods

2 2.1. RNA purification and cDNA conversion

3 Blood (2.5 ml) was collected from healthy community controls in PAXgene Blood

4 RNA tubes (PreAnalytiX, Hilden, Germany) according to the manufacturer’s instructions.

5 Total RNA was purified using PAXgene Blood RNA kit (Qiagen, Hilden, Germany) with

6 an on-column DNase digestion step to remove genomic DNA. DNase digestion was done

7 at room temperature for 15 minutes. The purity (A260/280nm ratio) and the concentration of

8 RNA were measured, using spectrophotometry (Gene Quant 1300, Uppsala, Sweden).

9 RNA quality was confirmed by gel electrophoresis and the starting material was

10 consistently found to be of good quality. Of the total RNA, 0.5 - 2 µg was used for cDNA

11 conversion using High capacity cDNA Reverse transcription kit (Applied Biosystems,

12 California, USA). The reaction was carried out according to the manufacturer’s instruction,

13 using thermocycler (Corbett Gradient Palm-Cycler, Sydney, Australia). The converted

14 cDNA was stored at -20oC.

15

16 2.2. Construction of standards

17 PCR was carried out with real-time PCR primers (200nM), Hotstar Taq® Master

18 Mix (Qiagen, Hilden, Germany) and cDNA (5 µl) for each gene (HuPO and IFN γ) as a

19 template in a final volume of 100 µl. Cycling conditions used for all genes were: initial

20 denaturation 10 minutes at 95oC followed by 45 cycles of 15 seconds at 95oC and 1 minute

21 at 60oC. 10 µl of PCR product was visualized in a 2% agarose gel. The product was

22 analyzed and quantified using DNA markers of known concentration (Fermentas, European

23 Union) as reference. This amplified PCR product (without purification) was used as the

24 first type of standard in quantitative real-time PCR.

5
 1 The second standard was made by purifying the above PCR product using a PCR

2 purification kit (Qiagen, Hilden, Germany). For both the first and second type of standards,

3 DNA was quantified using a Volume Regression curve analysis on Quantity one software

4 v.4.6 in a Gel Doc XR system (Bio-Rad, California, USA) according to the

5 manufacturer’s protocols.

6 Each gene was cloned separately by using the TOPO TA pCR 2.1 cloning vector kit

7 (Invitrogen, California, USA.) according to the manufacturer’s instructions. The

8 recombinant vector was transformed into competent E.coli and 25 µl of transformed culture

9 was spread onto LB plates containing ampicillin (75µg/ml) and X-gal/IPTG. Transformed

10 (white) colonies were picked and processed for plasmid isolation. Plasmid purification was

11 done using a Plasmid Mini kit (Qiagen, Hilden, Germany). The presence of the insert in

12 the recombinant clones was confirmed by restriction digestion. The cloned circular

13 plasmid was quantified using a spectrophotometer and used as the third type of standard.

14 The fourth standard was prepared by linearizing the plasmid with KpnI (Fermentas,

15 European Union). Digestion products were detected by electrophoresis and gel purified

16 using the MinElute Gel Extraction kit (Qiagen, Hilden, Germany). Linearised plasmid was

17 quantified using a spectrophotometer and copy numbers were calculated for all standards

18 by the following formula (Godornes et al., 2007):

19

20 6.022x1023 (molecules/mole) × DNA concentrations (g/µl)

21 Number of Copies/µl =
22 Number of bases pairs × 660 Daltons
23

24 6.022x1023 (Molecules/mole) - Avogadro’s number

25 660 Daltons - Average weight of a single base pair.

6
 1

2 2.3. qPCR using hydrolysis probe

3 qPCR was performed three times in duplicate after 0, 7 and 14 days of storage using

4 a Rotor-Gene 6000 (Corbett, Sydney, Australia). Six independent experiments were

5 conducted. 100 nM of dual labeled hydrolysis probe, 350nM each of forward and reverse

6 primers were used. The standard DNA (template) was serially diluted in nuclease-free

7 water (from 1010 to 102 specific copies/5 µl). Serial dilutions from 106 to102 (copies/5 µl)

8 were used for all four types of standard in a final volume of 20 µl alongside a negative

9 control (RNA) and a non-template control (NTC). The Quantitect probe PCR kit (Qiagen,

10 Hilden, Germany) was used for amplification. The reaction was set manually in a 72-Well

11 rotor (Rotor-Gene 6000, Corbett, Sydney, Australia). Cycling conditions for all four types

12 of standards were: 2 min at 50oC then 10 min at 95oC followed by 45 cycles of 15 seconds

13 at 95oC and 1 minute at 60oC, with the data acquired at 60oC in green channel (510 nm). A

14 standard curve was obtained by plotting the threshold cycle (Ct) on the Y-axis and the

15 natural log of concentration (copies/µl) on the X-axis. Ct, slope, PCR efficiency,

16 correlation co-efficient (R2) and percentage of variance in copy numbers were calculated

17 by using the default settings of Rotor-Gene 6000 series software 1.7. The same

18 fluorescence threshold used to construct the standard curve was also used to

19 determine Ct values (Rutledge and Cote, 2003). Diluted standards were stored at 4oC and

20 -20oC for the experiment.

21

22 2.4. qPCR using SYBR Green I

23 The Quantitect SYBR Green PCR kit (Qiagen, Hilden, Germany) was used to

24 amplify the target sequence. The specificity of the reaction was confirmed by melt curve

7
 1 analysis, using the reaction setup detailed above. Nuclease free water was used in place of

2 probe. A melt curve was acquired from 65oC to 99oC on the green channel (510 nm).

3 2.5. Absolute quantification

4 Absolute quantification determines the actual copy numbers of target genes by

5 relating the Ct value to a standard curve (Yu et al., 2005). PCR efficiency plays an

6 important role in absolute quantification. Ct values in each dilution were measured in

7 duplicate and were plotted against the logarithm of their initial template copy numbers.

8 Each standard curve was generated by a correlation co-efficient (R2) of the plotted points.

9 From the slope of each standard curve, PCR amplification efficiency (E) was calculated

10 according to the equation (Lee et al., 2006):

11 E = 10 (–1/slope) – 1

8
 1 3. Results

2 Melt curve analysis of HuPO and IFN γ showed the primers (Table 1) to be

3 specific for the cDNA prepared (data not shown) and the reaction efficiency and linearity

4 for the serially diluted standards prepared was of consistently good quality (Fig 1) for both

5 HuPO and IFN-γ . We then processed the DNA to prepare standard curve material in the

6 four ways most commonly used in research laboratories. Simple, unpurified PCR-amplified

7 cDNA was designated as Type 1, while column-purified material was designated type 2.

8 The PCR products were also cloned into the TOPO TA pCR 2.1 plasmid and purified

9 plasmid, either unmodified, or linearised with the restriction enzyme KpnI was used as type

10 3 and 4 standards, respectively. These preparations were then diluted by a series of 10-

11 fold steps to generate the standard curve material, which was tested for stability on storage.

12 We have considered the following criteria in this study: PCR efficiency [1.00(±0.05)],

13 regression curve (R2 = 0.99) and percentage of variance (≤ 10%) between given copy

14 numbers and calculated copy numbers per microlitre. As shown in Fig 1, PCR efficiency

15 was high, and this was true across all of the standard curve material (Data not shown). The

16 percentage of variance of the diluted standards for different storage duration and conditions

17 are shown in Fig 2.

18 As is clear from the data, that all of the diluted standards showed a variance of ≤

19 10% on day 0, which is generally considered acceptable. But the type I diluted standards

20 stored at 4oC showed a variance of > 10 % after only seven days of storage at 4oC and by

21 day 14, this had become extreme, reaching over 30% at the lower dilutions (Fig 2A).

22 Storage at -20oC improved stability, but after only 14 days there were some signs that a

23 similar tendency was emerging, even in the frozen samples. Type 2 diluted standards

24 appeared to be slightly more stable, but increasing variance when stored at 4oC was seen by

9
 1 day 14 (Fig 2B). There was no appreciable increase seen in variance in the type 2 standard

2 when frozen for up to 14 days. Type 3 diluted standards appeared to be the most stable and

3 when stored at -20oC showed ≤ 10% variance for up to 14 days, or even when stored at 4oC

4 for up to a week (Fig 2C). Finally, the type 4 standard, while apparently an improvement

5 on PCR product, appeared to be marginally less stable than the Type 3 (uncut) plasmid

6 with a tendency towards increased variance in standards stored to 2 weeks at 4oC (Fig 2D).

7 These data indicate that preparation and storage of standard curve materials can

8 potentially have dramatic effects on the accuracy of the assay. Copy number variations

9 affect the PCR efficiency of standards significantly, over the period of storage. We have

10 found that a deviation of as little as 0.06 – 0.10 efficiency (from 1.00) of standards can

11 result in misleading copy numbers of the genes analyzed (data not shown).

10
 1 4. Discussion

2 The development of qPCR to directly quantify absolute copy number of cDNA

3 (equivalent to mRNA copy numbers) represented a major step forward in PCR technology

4 and it is now routinely applied in a wide variety of studies. Cytokines are important

5 mediators of the immune system and are being studied extensively using this approach.

6 Many of the most informative studies in terms of immune response are longitudinal studies

7 that follow the development of responses over time, to try and establish causal effects in

8 disease and protection. This raises the issue of comparing multiple samples over extended

9 periods of time – all the more important, since samples in such studies are irreplaceable.

10 Our interest in the work presented in this manuscript was therefore to study potential

11 problems in measuring cytokine gene expression using qPCR technique, which requires a

12 robust standard for absolute quantification. Different types of detection chemistries are

13 available such as hybridization and hydrolysis probes or SYBR Green I and Eva Green

14 dyes, which are widely used for qPCR based quantification. Among these, hydrolysis

15 probes have the advantages of better specificity, sensitivity and reproducibility. Further,

16 this chemistry is especially suitable for accurate quantification (Overbergh et al., 1999;

17 Stordeur et al., 2002).

18 Currently, relative and absolute quantification methods are available. Relative

19 quantification is generally calculated with the 2-∆∆Ct formulae by comparative Ct method

20 (Schmittgen and Livak, 2008). The assumption for valid use of these formulae is that for

21 the gene of interest and the reference gene, the PCR efficiency must be similar (1.0 ± 0.05).

22 But, common reference genes may not be consistently expressed in all the samples and

23 therefore the relative quantification can be biased if the technique uses just one reference

24 gene.

11
 1 Alternatively, in absolute quantification, the only normalization done between

2 samples is for the starting amount of total RNA, thus decreasing potential sources of error

3 (Bustin, 2002). Consequently, absolute quantification requires standardization. By using

4 the copy number per microlitre of cDNA, it should be possible to compare results between

5 laboratories (Whelan et al., 2003). Absolute quantification is a powerful tool used in

6 various tasks, from quantifying viral load to gene expression studies. However, all such

7 uses rely on suitable standards. In the present study, we have demonstrated the importance

8 of stable standards for absolute quantification method using the qPCR technique. We

9 found that the reported copy numbers vary substantially over even a short period of

10 time, depending on storage conditions. Significant copy number variance between four

11 groups were found when stored, showing that the type 1 and 2 standards (PCR product) are

12 comparatively unstable during storage. Conversely, the type 3 and 4 serially diluted

13 standards (Cloned target) were more stable.

14 These copy number variations appear due to the degradation of target sequences,

15 which inturn directly alters the PCR efficiency of standards significantly, but not the

16 correlation co-efficient R2. Hence it is suggested that the correlation coefficient alone may

17 not be a reliable measure. All of the above mentioned criteria (PCR efficiency, copy

18 number variance and correlation co-efficient R2) are equally important for absolute

19 quantification.

20 Type 1 and 2 standards (PCR product) are short sequences which have good

21 stability in concentrated form but may lose integrity when serially diluted and aliquoted

22 for standard construction. The chance of degradation is likely to be increased by repeated

23 freeze-thaw and general handling of diluted standards, especially Type 1 standards. But

24 Type 3 and 4 standards (circular plasmid) appear to survive such handling better. Our data

12
 1 support the findings of Giulietti et al. (2001) that plasmid standards can be stored

2 comparatively longer at -20oC, without significant degradation.

3 Our data suggest that serial dilutions of cloned plasmid as appear to be the most

4 robust and suitable standard for long-term study purposes, even if plasmid construction and

5 cloning requires extensive work and is expensive. Though PCR product standards are

6 easier to make and less expensive, our data suggest that care should be taken to use only

7 freshly prepared serial dilutions. Certainly, one should avoid storage of diluted PCR

8 standards at 4oC, even if these standards can be stored at -20oC for up to one week.

9 The data here suggest that cloned target standards are relatively more stable

10 than PCR products, that storage at -20oC is preferable to 4oC even for relatively short

11 periods and that storage at higher concentrations improved stability. It is possible

12 that stability could be further improved by storage at -80oC, though here we have

13 focused on 4oC or -20oC since these conditions are used in most laboratories for

14 storage of PCR products and cloned plasmids. For the same reason, we have not

15 investigated the effects of longer periods of storage, since typically it is unusual to

16 store diluted aliquots for longer than 2 weeks at 4 or -20oC. Nonetheless, given that we

17 found over 30% variance in standard curves prepared and stored under typical

18 laboratory conditions, in just 1-2 weeks, it can be assumed that longer storage in

19 similar conditions would be unwise. The impact of such variation may vary depending

20 on the type and abundance of the target, but added to the variations derived from

21 efficiency of the sampling, extraction and cDNA conversion methods, it may prove

22 very significant, especially for targets present in low copy number or which are

23 especially labile.

13
1 In conclusion, in order to obtain the most reliable PCR efficiency, it is generally

2 accepted that one should design primers to span exon-exon junctions (so as to avoid

3 genomic DNA amplification) and avoid primer dimer formation. For gene expression

4 studies, primers giving 60 base pair (bp) to 150 bp amplicon sizes are considered ideal

5 (amplicons more than 150 bp may give <1.0 PCR efficiency). To this should be added the

6 suggestion that PCR standard curve materials should be tested for robustness, and stored

7 concentrated at -20oC as cloned plasmid products.

14
 1 Acknowledgements

2 We thank our colleague Geojith George and Division of Infectious Diseases, St.

3 John’s Research Institute, Bangalore. This work was supported by a grant from the

4 Research Council of Norway (TBTrials-RCN179342).

6 TB Trials Study Group

7 India:

8 1. Mario Vaz, Principal Investigator, Professor, Epidemiology and Biostatistics, St.

9 John’s Research Institute, Bangalore 560034, India.

10 2. Anura V Kurpad, Professor and Dean, St. John’s Research Institute, Bangalore

11 560034, India.

12 3. John Kenneth, Associate Professor, Infectious Diseases, St. John’s Research

13 Institute, Bangalore 560034, India.

14 4. Rajini Macaden, Professor, Infectious Diseases, St. John’s Research Institute,

15 Bangalore 560034, India.

16 5. Auburn Jacob, Director, Emmaus Swiss Leprosy Project and Referral Hospital, PO

17 Palamaner 517408, Chittoor District, A.P, India.

18 Norway:

19 1. Harleen M.S. Grewal, Principal Investigator, July 2009-present Professor and

20 Senior Consultant, The Gade Institute, Section for Microbiology and Immunology,

21 University of Bergen and Haukeland University Hospital, Bergen, Norway.

22 2. Bernt Lindtjorn, Prinicipal Investigator: July 2007-July 2009, Professor, Centre for

23 International Health, Faculty of Medicine P.O. Box 7804, N-5020 Bergen, Norway.

15
 1 3. Frode Jahnsen, Professor, Pathology Clinic, Rikshospitalet-Radiumhospitalet

2 Medical Center, N-0027, Oslo, Norway.

3 USA:

4 1. Sean Bennett, Clinical Project Physician, Aeras Global TB Vaccine Foundation,

5 1405 Research Blvd., Rockville, MD 20850, USA.

6 2. Lawrence Geiter, Former Senior Director, Epidemiology, Aeras. Currently: Senior

7 Director, TB Products Unit Otsuka Pharmaceutical Development &

8 Commercialization, Inc., 2440 Research Boulevard Rockville, MD

9 Denmark:

10 1. Mark Doherty, Professor, Statens Serum Institute, Artillerivej 5, Copenhagen

11 2300s, Denmark.

12 South Africa:

13 1. Anneke C. Hesseling, Professor, Desmond Tutu TB Centre, Department of

14 Paediatrics and Child Health, Stellenbosch University, Cape Town, South Africa.

16
 1 Reference

2 Bustin, S.A., 2002. Quantification of mRNA using real-time reverse transcription PCR

3 (RT-PCR): trends and problems J. Mol. Endocrinol. 29, 23–39.

5 Collins, M.L., Zayati, C., Detmer, J.J., Daly, B., Kolberg, J.A., Cha, T.A., Irvine, B.D.,

6 Tucker, J., Urdea, M.S., 1995. Preparation and characterization of RNA standards for use

7 in quantitative branched DNA hybridization assays. Anal Biochem. 226, 120–129.

9 Dheda, K., Huggett, J.F., Bustin, S.A., Johnson, M., Rook, G., Zumla, A., 2004. Validation

10 of housekeeping genes for normalizing RNA expression in real-time PCR. Biotechniques.

11 37, 112-119.

12

13 Dworkin, L.L., Gibler, T.M., Van Gelder, R.N., 2002. Real-time quantitative polymerase

14 chain reaction diagnosis of infectious posterior uveitis. Arch. Ophthalmol. 120, 1534-1539.

15

16 Fu, J., Li, D., Xia, S., Song, H., Dong, Z., Chen, F., Sun, X., Tang, Z., 2009. Absolute

17 quantification of plasmid DNA by real-time PCR with genomic DNA as external standard

18 and its application to a biodistribution study of an HIV DNA vaccine. Anal sci. 25, 675-

19 680.

20

21 Giulietti, A., Overbergh, L., Valckx, D., Decallonne, B., Bouillon, R., Mathieu, C., 2001.

22 An overview of real-time quantitative PCR: Applications to quantify cytokine gene

23 expression. Methods. 25, 386–401.

17
 1 Godornes, C., Leader, B.T., Molini, B.J., Centurion-Lara, A., Lukehart, S.A., 2007.

2 Quantitation of rabbit cytokine mRNA by real-time RT-PCR. Cytokine. 38, 1–7.

4 Lee, C., Kim, J., Shin, S.G., Hwang, S., 2006. Absolute and relative qPCR quantification of

5 plasmid copy number in Escherichia coli. J. Biotech. 123, 273–280.

7 Leong, D.T., Gupta, A., Bai, H.F., Wan, G., Yoong, L.F., Too, H.P., Chew, F.T.,

8 Hutmacher, D.W., 2007. Absolute quantification of gene expression in biomaterials

9 research using real-time PCR. Biomaterials. 28, 203–210.

10

11 Overbergh, L., Valckx, D., Waer, M., Mathieu, C., 1999. Quantification of murine cytokine

12 mRNAs using real time quantitative reverse transcriptase PCR. Cytokine. 11, 305–312.

13

14 Rose’Meyer, R.B., Mellick, A.S., Garnham, B.G., Harrison, G.J., Massa, H.M., Griffiths,

15 L.R., 2003. The measurement of adenosine and estrogen receptor expression in rat brains

16 following ovariectomy using quantitative PCR analysis Brain Res. Brain Res. Protoc. 11,

17 9-18.

18

19 Rutledge, R. G., Cote, C., 2003. Mathematics of quantitative kinetic PCR and the

20 application of standard curves. Nucleic Acids Res., 31, e93.

21

22 Schmittgen, T.D., Livak, K.J., 2008. Analyzing real-time PCR data by the comparative CT

23 method. Nat Protoc. 3, 1101– 1108.

24

18
 1 Stordeur, P., Poulin, L.F., Craciun, L., Zhou, L., Schandene, L., Lavareille, A., Goriely, S.,

2 Goldman, M., 2002. Cytokine mRNA quantification by real-time PCR. J. Immunol.

3 Methods. 259, 55–64.

5 Whelan, J.A., Russell N.B., Whelan, M.A., 2003. A method for the absolute quantification

6 of cDNA using real-time PCR. J. Immunol.Methods. 278, 261-269.

8 Yu, Y., Lee, C., Kim, J., Hwang, S., 2005. Group-specific primer and probe sets to detect

9 methanogenic communities using quantitative real-time polymerase chain reaction.

10 Biotechnol. Bioeng. 89, 670–679.

19
 1 Legends - Figures

2 Fig. 1. Amplification graph and standard curve constructed using 106 to 102 copies/µL of

3 HuPO (A) or IFN-γ (B). The data represent a representative result from six

4 independent experiments with samples analyzed in duplicate, at three time points and

5 show the high level of reproducibility and PCR efficiency obtained

7 Fig. 2. Mean percentage of variance for analysis of HuPO and IFN-γγ stored for 0, 7

8 and 14 days at 4oC or -20oC, using (A) Type 1 standard (PCR product without

9 purification), (B) Type 2 standard (PCR product with purification) (C) Type 3 standard

10 (PCR product cloned into plasmid) and (D) Type 4 standard (PCR product cloned into

11 linearized plasmid). Data is the mean of duplicate analyses performed over 6

12 independent experiments (intra-experimental variability was consistently less than

13 10%), indicating that increased variance appears to be associated with lower standard

14 concentration, longer storage times, higher temperatures and the nature of the

15 standard.

20