Measuring the TREC ratio in dried blood spot samples: Intra- and inter-filter paper cards reproducibility

Journal of Immunological Methods 389 (2013) 1–8

Contents lists available at SciVerse ScienceDirect

Journal of Immunological Methods journal homepage: www.elsevier.com/locate/jim

Research paper

Measuring the TREC ratio in dried blood spot samples: Intra- and inter-filter paper cards reproducibility☆ P.O. Lang a, b,⁎, S. Govind a, M. Dramé c, R. Aspinall a a b c

Translational Medicine Research group, Cranfield Health, Cranfield University, Cranfield, UK Nescens Centre of preventive medicine, Clinic of Genolier, Genolier, Switzerland Methodological assistance unit, Department of Research and Innovation, Robert Debré Hospital, Reims teaching Hospitals, Reims, France

a r t i c l e

i n f o

Article history: Received 19 November 2012 Received in revised form 6 December 2012 Accepted 10 December 2012 Available online 9 January 2013 Keywords: TREC ratio Thymic TREC Dried blood spots Reproducibility Real time PCR Quantitative PCR

a b s t r a c t The level of T-cell receptor excision circles (TREC), which decline with advancing age in normal individuals, has recently gained interest as a reference marker for studies on premature or early immunosenescence under particular health conditions. In order to facilitate translational studies at population and clinical levels, essential for the understanding of how changes in TREC levels are associated with responsiveness of the immune system, we have developed and optimized a real-time polymerase chain reaction (qPCR) assay which quantifies the TREC ratio from dried blood spot (DBS) samples. The present study considers a fully automated procedure to purify DNA and amplify sequences of interests by means of qPCR from DBS samples collected in healthy adults. Both TREC:PBMC (peripheral blood mononuclear cell) and TREC:T-cell ratios were compared for intra- and inter individual reproducibility. Furthermore, the impact of the length of storage on the quality of the DNA generated was also analyzed. In conclusion we describe a fully automated procedure for extracting DNA and qPCR set up, which offers a high-precision, robust qPCR assay for the quantification of both TREC:T-cell ratio and TREC:PBMC from DBS samples. © 2012 Elsevier B.V. All rights reserved.

1. Introduction To ensure longevity and healthy life the maintenance of appropriate immunity is fundamental. Age- or diseaserelated changes of the immune system are of particular importance, contributing to the higher incidence and severity of infectious diseases, reduced efficacy of vaccination and possibly autoimmunity and cancer (Schneider, 2010). While these changes can affect many components of both the innate as well as adaptive immunity (Arnold et al., 2011) one of the most prominent features of the immune senescence process is changes in the composition of the T-cell compartment and particularly the decrease in antigen-inexperienced naïve T-cells (Lang et al., 2012a). This is also obvious in individuals suffering ☆ Conflict of interest: None of the authors declare conflict of interests for this manuscript. ⁎ Corresponding author at: Nescens, Centre of preventive medicine, c/o Clinique of Genolier, Route du Muids 3, CH-1272 Genolier, Switzerland. Tel.: + 41 22 366 93 09; fax: + 41 22 366 93 49. E-mail address: [email protected] (P.O. Lang). 0022-1759/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jim.2012.12.003

from combined immunodeficiency (Arnold et al., 2011; van Zelm et al., 2011; Lang et al., 2012a), in HIV-infected and lymphopenic cancer patients and at the end-stage of chronic renal diseases (Douek et al., 2000; Hazenberg et al., 2000, 2002; Betjes et al., 2011). Predicting individual immune responsiveness to any antigen using biological markers that distinguish between healthy and “immunologically vulnerable” states is desirable. To establish early identification of such individuals including an estimation of future complication risk, standardized and reliable assays are required. One of the newly explored methods involves quantification of circular DNA products that are generated during T-lymphocyte development with the creation of T-cell receptors (TCR) (van Zelm et al., 2011). This receptor is employed by naïve T-lymphocytes to recognize foreign antigens. In order to create a boarder repertoire of TCR molecules, each immature T-lymphocyte during intra-thymic development undergoes unique somatic rearrangements in its TCR loci (Lang et al., 2012a). During this rearrangement process, the intervening DNA sequences are deleted and circularized into

2

P.O. Lang et al. / Journal of Immunological Methods 389 (2013) 1–8

episomal DNA molecules, also called signal joint TCR excision circles or sj-TRECs (thereinafter referred as TREC) (van Zelm et al., 2011). Thus in T-cell expressing TCR-αβ (i.e. 95% of T-cells), rearrangements of both α and β chains produce TREC. While TRECs are identical in 70% of developing αβ T-cells (Verschuren et al., 1997), they have no specific function. They are also stable products that are not replicated during subsequent mitosis. Consequently, TREC are diluted during cell division and conversion of naïve lymphocyte to memory/effector T-cells (Lang et al., 2012a). The TREC level, which declines with advancing age in healthy individuals (Hazenberg et al., 2003; Mitchell et al., 2010; Lang et al., 2012a), could be thus used as a reference for studies on premature or early immunosenescence under particular health conditions (Douek et al., 2000; Hazenberg et al., 2000, 2002; Mitchell et al., 2010; Betjes et al., 2011; van Zelm et al., 2011). In female rhesus macaques vaccinated with inactivated influenza vaccine, higher levels of influenzaspecific CD8 + T cell were among those with the highest TREC levels (Aspinall et al., 2007). Finally, to facilitate complementary clinical and translational studies at population and clinical levels needed to address how changes in TREC levels are associated with responsiveness of the immune system (World Health Organization (WHO), 2011) and to elucidate the exact contribution of the resting naïve T-lymphocyte pool to the immune senescence process (Arnold et al., 2011; Lang et al., 2012a), we have developed and optimized a real-time polymerase chain reaction (qPCR) assay quantifying the TREC: T-cell ratio (Lang et al., 2011, 2012b). The measurements obtained with this qPCR assay in whole blood and in DBS samples were comparable to those obtained by Mitchell et al., where total leucocyte and/or T-cell count measured by flow cytometer after anti-CD3 staining and TREC measurement with fluorescent dye qPCR assay (Mitchell et al., 2010). We have also provided some arguments of preferentially choosing an automated procedure for DNA extraction from DBS samples (Lang et al., 2012b). Thus, by using a fully automated procedure to purify DNA and amplify albumin gene (ALB), the TCR-β germline area downstream of the VD gene (thereinafter referred as VD gene) and TREC sequences by means of qPCR, the present study has contributed to further validate this qPCR assay. In our present data we provide both TREC:PBMC (peripheral blood mononuclear cells) and TREC:T-cell ratios, which were compared in terms of intra- and inter-filter paper card reproducibility in healthy adults. Furthermore, the impact of the length of storage on the variability of the ratio measurements was assessed by using DBS samples spotted from the same individuals at two different periods of time. Our present findings supplement our preliminary studies (Lang et al., 2011, 2012b), conducted in healthy and middle age adults to optimize and further validate this qPCR assay before applying the technique to the DBS collection of the SAGE project (Study on Global AGEing and Adult Health) (World Health Organization, 2011). 2. Material and methods 2.1. Blood samples Peripheral blood samples were collected from 5 volunteers either in June 2010 (#1: 38-years old — y.o.-man; #2: 38-y.o.

man; and #3: 26-y.o woman) or in November 2010 (#4: 28-y.o. man) and in March 2011 (#5: 40-y.o. man). In addition, peripheral blood samples were also collected from the same individual #1 (40-y.o. at this time) in July 2012 and #2 (39-y.o. at this time) in April 2012. All blood collections were performed at the University of Oregon (Eugene, USA) except for sample #2 that were collected at Cranfield University (Cranfield, UK). For each blood test, ten milliliters (mL) of venous blood samples was collected into EDTA purple top vacutainer tubes (BD Bioscience). Volunteers were all chosen on the criteria that they were healthy with no known disease conditions or blood pathogens at the time of sample extraction. They willingly contributed blood samples and were not remunerated for these samples. The Health Research Authority (NHS), National Research Ethics Service (NRES South central authority, Bristol, Berkshire, UK) approved the study (No. 09/H0603/34-AM02). 2.2. Dried blood spot (DBS) samples From all venous blood samples, DBS samples were spotted by applying 50 μL of blood onto Schleicher and Schuell (S&S) no 903 filter papers (Schleicher & Schuell BioScience, Inc., Keene, New Hampshire, UK). A total of 10 DBS were spotted onto each card and this for a total of 8 cards (thereinafter referred as Cards #1A, #1B, #1C, #2A #2B, #3, #4 and #5). All S&S 903 cards were then air-dried at room temperature for a minimum of 12 h and stored at a minimum of − 20 °C until further analysis. These conditions were similar to those by which the DBS collection has been obtained during the SAGE project (World Health Organization (WHO), 2011). 2.3. Extraction of DNA solutions DNA was extracted from each entire blood spot by punching out 6 × 3 mm circles using a paper punch (Harris Micro-Punch™, Tip diameter 3.00 mm — cat. no. 15094, Ted Pella Inc., Redding, CA). Fully automated DNA purification, was performed using the QIAamp® DNA Investigator kit, a column-based extraction method (QIAamp® MinELute column), on the QIAcube® (Qiagen group, Crawley, UK), as previously described (Lang et al., 2012b). 2.4. Assessment of the quality and quantity of DNA generated The ratio of the absorbance at 260 and 280 nm (A260:280 ratio) was determined by optical density readings using the Picodrop Microliter UV/Vis Spectrophotometer (Picodrop Ltd., Saffron Walden, UK) to assess the purity of nucleic acids. All measurements were performed in triplicate and an average used for statistical analyses. Because extracted DNA amounts were expected to be low (b50 ng/μL), DNA concentrations were concurrently determined by qPCR using the count for albumin gene (ALB) as further detailed below. ALB was considered as a housekeeping gene; the count for ALB corresponded to the total number of peripheral blood mononuclear cells (PBMC) into the blood samples (Ct value of 24.0 cycles≈1× 105 PBMC) (Lang et al., 2012b). The DNA content in each of these diploïd cells was estimated as being 6.6 pg (i.e. 3× 10 9 (size of the human genome in base pair — bp)× 2 (diploïd cell) ×660 (molecular weight of 1 bp) ×1.67× 10−12 pg (weight of 1 Da)= 6.6 pg) (Ranek, 1976).

P.O. Lang et al. / Journal of Immunological Methods 389 (2013) 1–8

2.5. Quantitative real-time PCR analyses 2.5.1. ALB, VD and TREC primer sets and qPCR conditions qPCR amplification for ALB and VD genes and human sj-TREC (TREC) levels were performed on the Rotor-Gene Q qPCR cycler (Qiagen, Crawley, UK). A reaction mixture containing 2 μL of DNA solution, 0.5 μM of forward and reverse primers and 2 × SYBR Green mix Quantitect (Qiagen, Crawley, UK) in a final reaction volume of 10 μL was made using nuclease-free water. For rapid and high-precision PCR setup, the qPCR mixtures were automatically prepared on the QIAgility® (Qiagen, Crawley, UK). The primer sequences used were: ALB forward: 5′-CTA TCC GTG GTC CTG AAC CAG and reverse: 5′-CTC TCC TTC TCA GAA AGT GTG (Lang et al., 2012b); VD forward: 5′-TGG CCA CAG GAG GTC GGT TT and reverse: 5′-TTA CTC CTG CGC CTC TGT GTC (Lang et al., 2011); and TREC forward: 5′-GGC AGA AAG AGG GCA GCC CTC and reverse: 5′-GCC AGC TGC AGG GTT TAG G (Lang et al., 2012b); which produced amplicons of 206 bp, 256 bp, and 195 bp respectively. The thermocycling conditions for ALB were 95 °C for 15 min, followed by 45 cycles at 95 °C for 20 s, 60 °C for 20 s and 72 °C for 30 s (fluorescent acquisition). The VD-J and TREC reactions were performed as described above, except that annealing temperatures were changed to 57 °C and 61 °C respectively. The 206 bp, 256 bp and 195 bp qPCR products were identified by melting point analysis. All samples were run in triplicates and an average of the results used for statistical analyses. Where Ct values of the triplicates were greater than 1.5 cycles the samples were excluded from subsequent analyses. 2.5.2. qPCR for ALB, VD and TREC absolute quantification ALB, VD and TREC copy number was obtained by extrapolating the respective sample quantities from the standard curve obtained by serial dilutions (10 8, 10 7, 10 6, 10 5, 10 4, 10 3, 10 2 and 10) of linearized plasmids containing inserts corresponding to fragments of interest. The copy number of ALB, VD and TREC (X) was calculated using the following equations: YALB = −3.374X+ 40.593, YVD = −3.340X + 39.810 and YTREC =−3.468X + 42.09, where the cycle threshold (Ct) value is substituted as Y. A standard concentration of 1 × 103 ALB, VD or TREC molecules was included to determine variance between each run and comparability of the samples. ALB count was used in order to determine the total number of PBMC (Lang et al., 2012b); the unrearranged TCR-β count corresponded to the total germline cell number (Chain et al., 2005). Hence, the total T-cell count was calculated by subtracting the unrearranged TCR-β cell count (VD level) from the total cell count (ALB level) (Lang et al., 2011). Finally based on ALB, VD and TREC copy numbers, two ratios were calculated by using the following formulas: TREC : PBMC ¼

mean of TREC cn 6  10 mean of ALB cn=2

TREC : T cells ¼

mean of TREC cn 6  10 ðmean of ALB cn−mean of VD cnÞ=2

* where cn= copy number. The mean quantity of ALB and ALB-VD had to be divided by 2 because these control genes are biallelically present in

3

the genome, while excision circles are not replicated (Motta et al., 2010; van Zelm et al., 2011). 2.6. Statistical analysis The results of the descriptive analysis are presented in the form of means ± standard deviation (SD). In order to analyze how measurements made from blood spots punched-out of the same filter paper card resemble each other in terms of quantity of DNA generated and downstream PCR measurements, different statistic methods have been considered. To assess the agreement between the two techniques used to measure the concentration of the DNA samples (i.e. spectrophotometer vs. ALB levels quantified by means of qPCR), both Bland–Altman plots (difference plot) and intra-class correlation coefficient (ICC) and its 95% confidential interval (CI) were computed (Altman and Bland, 1983; Rodgers and Nicewander, 1988). The Bland–Altman plot is a method of data plotting used in analyzing agreement between two different assays, neither being considered as a gold standard, by calculating the mean difference between the two methods of measurement (the bias) and 95% limits of agreement as the mean difference (1.96 SD) (Altman and Bland, 1983). It is expected that the 95% limits include 95% of differences between the OD and qPCR measurements. Bland–Altman plots were drawn using SPSS software (IBM® SPSS® release 20.0; IBM Corp., Armonk, NY, USA). The ICC ranges from 0 to 1 with values closer to 1 indicating greater concordance between the two techniques (Rodgers and Nicewander, 1988) were computed with the PROC MIXED for SAS software. We determined whether there was a significant difference between the mean values of DNA concentration (OD vs. qPCR) and the TREC-based ratios measured from blood spots within the same filter paper card or from the same individual's blood but spotted onto different filter cards (Card #1A, #1B and 1C; and Card #2A and #2B). The statistical test was chosen according to the sample size and the number of comparable groups (i.e. Kruskal–Wallis or Mann–Whitney test). Finally, to test the influence of the length of storage on the concentration of DNA (ng/μL) extracted, the linear Pearson correlation coefficient (r) was calculated. Statistical analyses were computed using SAS software for Window (version 9.0; SAS Institute®, Inc., Cary, NC, USA) and the level of significance was set at p= 0.05. 3. Results 3.1. Quantity and quality of generated DNA solutions The concentration and purity of DNA solutions generated from DBS are shown in Table 1. There is an overall agreement between the two techniques of quantification (spectrophotometry vs. qPCR), with ICC values close to 1 as shown in Table 2. Of the 240 measurements 17 (7.7%) were outside the 95% limits in the Bland–Altman plot (Fig. 1). On average, the difference observed between the two estimates of the same DNA solution was ranged from 1.28 to 1.91 ng (mean ± SD: 1.63 ±0.25) with no systematic error for either spectrophotometry or qPCR technique. The purity of the DNA solutions generated was confirmed by the 260:280 ratio. The OD measurements ranged from 1.71 to 2.87 (mean ± SD: 1.99± 0.01). When the possible effects of the length of storage

4

P.O. Lang et al. / Journal of Immunological Methods 389 (2013) 1–8

Table 1 Concentration and purity of DNA solutions generated from filter paper cards. Concentrations (ng/μL) and ratios of absorbance at 260 and 280 nm are presented as mean ± standard deviation (SD), median and inter-quartile range (IQR). For each filter paper card analyzed, the number of measurements is presented as well as the date of spotting when applicable. One card corresponded to 10 dried blood spot samples collected from one healthy adult. Volunteers were ranged in age from 26 to 40 years old at the time of collection. Each DNA solution purified from 1 DBS sample was analyzed in triplicate. Letters A, B and C indicate that blood samples were collected from the same individual either on a different card at the same time (Card #1A and #1B) or at two different dates (Card #1C vs. Card #1A and #1B or Card #2B vs. #2A). Filter paper card

Total (N = 240)

Card #1A (n = 30) Spotted on June 2010 Card #1B (n = 30) Spotted on June 2010 Card #1C (n = 30) Spotted in July 2012 Card #2A (n = 30) Spotted in June 2010 Card #2B (n = 30) Spotted in April 2012 Card #3 (n = 30) Spotted in June 2010 Card #4 (n = 30) Spotted in November 2010 Card #5 (n = 30) Spotted in March 2011

Concentration of DNA

Mean ± SD Median IQR Mean ± SD Median IQR Mean ± SD Median IQR Mean ± SD Median IQR Mean ± SD Median IQR Mean ± SD Median IQR Mean ± SD Median IQR Mean ± SD Median IQR Mean ± SD Median IQR

Purity of DNA

OD

qPCR

p-Value

260:280 ratio

42.85 ± 3.81 43.60 39.22–45.30 41.31 ± 4.37 42.05 38.50–43.60 41.55 ± 4.31 42.30 37.90–44.00 45.41 ± 1.88 45.15 44.00–45.60 38.53 ± 4.64 37.73 36.66–42.67 45.02 ± 0.56 45.05 44.90–45.30 39.35 ± 1.69 39.00 38.10–40.70 40.42 ± 2.17 39.75 38.50–41.90 46.89 ± 1.93 46.60 45.00–49.30

43.07 ± 3.79 43.82 39.79–45.59 41.54 ± 4.30 42.21 38.28–43.97 41.54 ± 4.30 42.20 38.28–43.97 45.44 ± 1.79 44.91 44.27–45.65 39.35 ± 4.27 37.60 36.23–43.52 45.14 ± 0.56 45.01 44.90–45.24 39.31 ± 1.90 38.95 38.05–40.44 41.34 ± 2.36 40.79 39.79–41.47 47.19 ± 1.85 46.42 45.57–49.03

b0.0001

1.99 ± 0.12 2.02 1.90–2.05 1.99 ± 0.09 1.94 1.85–2.03 1.97 ± 0.19 1.95 1.85–2.03 1.94 ± 0.09 1.95 1.85–2.03 1.99 ± 0.12 1.89 1.70–2.00 2.05 ± 0.03 2.04 2.02–2.07 1.94 ± 0.12 1.99 1.82–2.04 2.09 ± 0.07 2.09 2.04–2.13 2.01 ± 0.10 2.05 1.95–2.08

0.07

0.9

0.6

0.09

0.3

0.6

b0.0001

0.06

OD: optical density; qPCR: quantitative real-time PCR; p-value b 0.05 indicates that there is a significant difference when compared OD to qPCR. Card #1 was spotted with blood collected from 38 years-old (y.o.) man in June 2010 (Card #1A and #1B) and 40 in July 2012 (Card #1C). Card #2 was spotted with blood collected from a 38 y.o. man in June 2010 (Card #2A) and 39 in April 2012 (Card #2B). Card #3 was spotted with blood collected from a 26 y.o. woman in June 2010. Card #4 was spotted with blood collected from a 28 y.o. man in November 2010. Card #5 was spotted with blood collected from a 40 y.o. man in March 2011.

(in weeks) at −20 °C on both concentration and purity of DNA were analyzed, where a significant decrease on the quantity was globally observed, regardless of the technique of measurement. The Spearman's r coefficients were respectively −0.50 and −0.49 (p b 0.0001) for OD and qPCR measurements. More specifically, the interference was also observed when DNA solutions were extracted from DBS spotted with the blood of the same individual (Card #2A vs. Card #2B; Card #1A and #1B vs. Card #1C) were considered (p b 0.0001). While globally no interference was observed on the purity of the DNAs (r −0.08, p = 1.00), an intra-individual inter-card significant difference was however observed (p b 0.0001).

(r = − 0.1; p = 0.08) globally, no significant correlation was calculated with the measurements obtained from the Card #1 (A, B and C: r = − 0.08; p = 0.4) and Card #2 (A and B: r = −0.2; p = 0.08). The mean values of the ratio were not significantly different between Card #1A and Card #1B, Card #1A+ B compared to Card #1C and the Card #2A vs. #2B. The inter-spot consistency for ratio values obtained with the analyses performed from blood samples spotted on the same card was good as depicted by Fig. 2. Similar results were observed by analyzing samples spotted with the blood collected from the same individual but on different cards (Card #1A vs. #1B) and even with blood collected at different periods of time (Card #1A+ B vs. #1C; Card #2A vs. #2B).

3.2. Downstream qPCR application The descriptive results of the downstream qPCR applications are presented in Table 3. The results of the assessment of consistency or reproducibility of the two ratio measurements within each card are presented in (Fig. 2 and Table 3). 3.2.1. TREC:PBMC ratio In addition to the lack of significant impact of the length of storage at − 20 °C on the TREC:PBMC ratio quantification

3.2.2. TREC:T-cell ratio Similar to the TREC:PBMC ratio, no impact of the length of storage was measured globally (r = − 0.1–p = 0.4) and when only the samples collected from the same individuals were considered (Card #1A, Card #1B and Card #1C: r = − 0.1; p = 0.3; Cards#2A and Card #2B: r = 0.1; p = 0.2). The mean values were not significantly different between Card #1A and #1B, the Card #1A + B vs. #1C and the Card #2A vs. #2B.

P.O. Lang et al. / Journal of Immunological Methods 389 (2013) 1–8 Table 2 Overall and intra-card intra-class correlation coefficients (ICC) measuring the variability of the concentration of the DNA solutions generated from filter paper cards. Each ICC is presented with its 95% confidential interval (CI) and the corresponding p-value. Closer to 1 is the ICC greater that is the intra-sample homogeneity; below 0.05 the p value confirms that the intra-sample homogeneity is statistically significant. One card corresponded to 10 dried blood spot samples collected from one healthy adult. Volunteers were arranged in age from 26 to 40 years old at the time of collection. Each DNA solution purified from 1 DBS sample was analyzed in triplicate. Letters A, B and C indicate that blood samples were collected from the same individual either on a different card at the same time (Card #1A and #1B) or at two different dates (Card #1C vs. Card #1A and #1B or Card #2B vs. #2A). Filter paper card

Total (N = 240) Card #1 (n = 90) Card #1A (n = 30) Spotted on June 2010 Card #1B (n = 30) Spotted on June 2010 Card #1C (n = 30) Spotted in July 2012 Card #2 (n = 60) Card #2A (n = 30) Spotted in June 2010 Card #2B (n = 30) Spotted in April 2012 Card #3 (n = 30) Spotted in June 2010 Card #4 (n = 30) Spotted in November 2010 Card #5 (n = 30) Spotted in March 2011

Concentration of DNAs ICC

95% CI

p-value

0.991 0.991 0.989

0.988–0.993 0.987–0.994 0.864–0.968

b0.0001 b0.0001 b0.0001

0.992

0.982–0.996

b0.0001

0.974

0.946–0.988

b0.0001

0.943 0.847

0.941–9.956 0.710–0.992

b0.0001 0.001

0.971

0.950–0.997

b0.0001

0.974

0.946–0.988

b0.0001

0.933

0.864–0.968

b0.0001

0.898

0.797–0.950

b0.0001

Card #1 was spotted with blood collected from 38 years-old (y.o.) man in June 2010 (Card #1A and #1B) and 40 in July 2012 (Card #1C). Card #2 was spotted with blood collected from a 38 y.o. man in June 2010 (Card #2A) and 39 in April 2012 (Card #2B). Card #3 was spotted with blood collected from a 26 y.o. woman in June 2010. Card #4 was spotted with blood collected from a 28 y.o. man in November 2010. Card #5 was spotted with blood collected from a 40 y.o. man in March 2011.

The inter-spots consistency for the two ratio values obtained from analyses performed from DBS spotted with the blood collected from the same individuals (Card #1A, #1B and #1C) is presented by Fig. 2. With respect to the control genes used we did not observe considerable differences for preferentially choosing one approach compared to the other. 4. Discussion We have analyzed the concentration and integrity of the DNA extracted from DBS samples of healthy adults stored up to 21 months at − 20 °C and performed downstream qPCR with the DNAs generated. Using a fully automated procedure for both DNA extraction and PCR set up, the intra- and inter filter paper card comparisons revealed a good homogeneity and reproducibility of all measurements. Two approaches have been tested to quantify the TREC contents: TREC:PBMC and TREC:T-cell ratio and both approaches seemed to be of similar internal consistency. We have analyzed DNA extracted from DBS samples that were stored up to 21 months and report the ability to amplify gene sequences. Our study supports a preference for choosing an automated procedure of extraction when downstream qPCR applications are intended. A previous study, comparing

5

automated with non-automated protocol of extraction had already suggested the benefit of using automated procedure (Lang et al., 2012b). The present study confirms that the purity of automated extracted DNA is close to the optimum value as illustrated by the 260:280 ratios (see Table 1 and Fig. 1). As demonstrated by the Bland–Altman spot (see Fig. 2) the amount of albumin molecule significantly correlated the concentration of DNA extracted. This finding confirms not only the quality of the DNA solutions purified from the DBS samples but also the accuracy of the qPCR assay. Use of DBS is a convenient and a far from onerous method for biobanking in low-resource countries as well as a back-up strategy for conventional biobanks (Ivarsson and Carlson, 2011). It is widely described that this type of sample can be stored for days and even years as long as they are not exposed to elevated humidity (Li et al., 2004; Sherman et al., 2005; McDade et al., 2007; Masciotra et al., 2012). Recently, by using DBS samples stored up to 11 weeks at − 20 °C from the same individual we have analyzed the impact of the length of storage not only on the quality of the DNA generated but also on the variability of the TREC:PBMC ratio (Lang et al., 2012b). No impairment was observed when DNA samples were extracted with an automated procedure compared to a manual protocol. As demonstrated by previous work, TREC values can be reliably detected and measured with either fluorescence dyes such as SYBR Green (Mitchell et al., 2010) or probe based TaqMan-based qPCR (van Zelm et al., 2011). The exact number of TREC can be obtained for a given DNA sample as the number of TREC/μg DNA for example. However for more accurate quantification, additional steps usually need to be undertaken. As performed in the present study, a more accurate TREC evaluation can be obtained when quantification is performed relative to a control gene, such as ALB (Zubakov et al., 2010; Lang et al., 2012b), or the T-cell-count (Mitchell et al., 2010; Lang et al., 2011) or the TCRα constant region as proposed by other authors (Hazenberg et al., 2000; van Zelm et al., 2011). The selection of methodology utilized for quantifying excision circles is very important as it will impact the interpretation of the data obtained. Thus, while in few settings straightforward quantification of TREC values with or without respect to a control gene could be sufficient to address some specific question, this could not be sufficient for others (van Zelm et al., 2011). Since TRECs are specifically formed in developing lymphocytes they could be used to determine the presence or lack of naïve T-cells. This has been exploited in recent years to develop newborn screening tests for severe combined immunodeficiency (Chan and Puck, 2005). Recent studies have also demonstrated the influence of human age on the frequency of TRECs in blood (Mitchell et al., 2010; Zubakov et al., 2010; Lang et al., 2011) as this may be reflected by the differences observed between cards collected from adults of different ages (Table 1). Finally, while our work lends further support to the consideration of an automated procedure for DNA extraction (Lang et al., 2012b) that enables rapid and high-precision setup of the PCR reaction setup, we cannot provide a convincing enough argument to bias selection of either the TREC:T-cell or the TREC:PBMC ratio as the better readout. Indeed, both intracard and intra-individual heterogeneity, even if not statistically significant, were observed with both (see Section 3.2.1 and

6

P.O. Lang et al. / Journal of Immunological Methods 389 (2013) 1–8

Fig. 1. Agreement between the two techniques used to measure the concentration of the DNAs extracted from 6×3 mm punch-out circles. Results are presented for DNA extractions with a final volume of 50 μL. With the Bland–Altman plot the differences between the two techniques of measurement (i.e. spectrophotometry and qPCR) are plotted against the averages of the two techniques. The Bland–Altman plot method graph displays a scatter diagram of the differences plotted against the averages of the two measurements. Horizontal lines are drawn at the mean difference, and at the limits of agreement, which are defined as the mean difference+ and −1.96 times the standard deviation of the differences.

Section 3.2.2). However, the TCR-β germline area downstream of the VD gene may not the perfect selection criteria for excluding proper T-cells by means of qPCR. Indeed, as some

T-cells have only one rearranged allele, the qPCR level of germline TCR-β amplicons partly was derived from T-cells as well. Consequently, TREC:T-cell readout may be overestimated

Table 3 TREC:PBMC (peripheral blood mononuclear cells) and TREC:T-cell ratios measured by qPCR from each filter paper card. Measurements indicate the number of TREC molecules per 106 cells (PBMC or T-cells) and are presented as mean ± standard deviation (SD), median and inter-quartile range (IQR). One card corresponded to 10 dried blood spot samples collected from one healthy adult. Volunteers were arranged in age from 26 to 40 years old at the time of collection. Each DNA solution purified from 1 DBS sample was analyzed in triplicate. Letters A, B and C indicate that blood samples were collected from the same individual either on a different card at the same time (Card #1A and #1B) or at two different dates (Card #1C vs. Card #1A and #1B or Card #2B vs. #2A). Filter paper card

TREC:PBMC ratio

Card #1 (n = 90) Card #1A (n = 30) Spotted on June 2010 Card #1B (n = 30) Spotted on June 2010 Card #1A-B (n = 60) Spotted on June 2010 Card #1C (n = 30) Spotted in July 2012 Card #2 (n = 60) Card #2A (n = 30) Spotted in June 2010 Card #2B (n = 30) Spotted in April 2012 Card #3 (n = 30) Spotted in June 2010 Card #4 (n = 30) Spotted in November 2010 Card #5 (n = 30) Spotted in March 2011 Card Card Card Card Card

#1 was spotted #2 was spotted #3 was spotted #4 was spotted #5 was spotted

with with with with with

blood blood blood blood blood

TREC:T-cell ratio

Mean ± SD

Median

IQR

Mean ± SD

Median

IQR

5050 ± 52 5032 ± 61

5052 5044

5018–5093 4986–5059

24,134 ± 314 24,122 ± 389

24,181 24,216

24,000–24,337 23,767–24,487

5062 ± 51

5065

5026–5101

24,094 ± 340

24,054

23,946–24,413

5047 ± 58

5051

5005–5097

24,108 ± 362

24,192

23,857–24,429

5057 ± 58

5056

5019–5092

24,184 ± 180

24,181

24,140–24,242

5184 ± 119 5212 ± 140

5168 5191

5123–5237 5132–5238

20,946 ± 272 20,899 ± 294

21,014 20,860

20,710–21,153 20,668–21,285

5155 ± 88

5168

5101–5207

20,993 ± 243

21,034

20,777–21,090

4928 ± 261

4908

4773–5016

18,547 ± 271

18,494

18,282–18,824

4972 ± 131

5038

4829–5081

20,993 ± 330

21,034

20,967–21,207

4517 ± 119

4551

4423–4623

19,606 ± 492

19,453

19,209–19,981

collected collected collected collected collected

from from from from from

38 years-old (y.o.) man in June 2010 (Card #1A and #1B) and 40 in July 2012 (Card #1C). a 38 y.o. man in June 2010 (Card #2A) and 39 in April 2012 (Card #2B). a 26 y.o. woman in June 2010. a 28 y.o. man in November 2010. a 40 y.o. man in March 2011.

P.O. Lang et al. / Journal of Immunological Methods 389 (2013) 1–8

7

populations suffering from health disorders involving and/or affecting the cell-mediated immune system. All these issues will be addressed in the framework of the SAGE project (World Health Organization (WHO), 2011). 5. Conclusion By analyzing DNA samples extracted from DBS samples collected from healthy adults that have been stored up to 21 months and the ability to amplify albumin, TCR-β germline area downstream of the VD gene and TREC sequences, this study provides complementary findings, in terms of intra- and inter-card reproducibility, to further validate the qPCR assay measuring the TREC:PBMC and/or TREC:T-cell ratio in DBS samples. While this preliminary study did not provide a robust argument to consider one readout over another, this question will be addressed as well as the relationship of these readouts with health and immunological outcomes by analyzing the SAGE project samples. Acknowledgments We are grateful to the World Health Organization (WHO) for supporting and funding the research project referenced as HQIVB1207048 concerning the “Optimization of the QPCR mono assay measuring the TREC ratio in DBS samples” and more particularly Dr Maria Teresa Aguado de Ros and Dr Judith Thomas-Crusells (Initiative for Vaccine Research (IVR), Immunization, Vaccines and Biologicals (IVB), WHO, Geneva, Switzerland) and Dr Somnath Chatterji (Health Statistics and Informatics, WHO, Geneva, Switzerland). The WHO has not had any role in the design, subject recruitment, methods, data collections, analysis, and preparation of the present manuscript. We would like to extend our gratitude to Prof J. Josh Snodgrass for providing us the filter paper cards from the University of Oregon (Department of Anthropology, Eugene, OR, USA) and all the volunteers for agreeing to participate. References Fig. 2. “Spider-web” schemas depict the internal consistency or homogeneity and inter-individual variability in measuring the TREC:PBMC ratio (Panel A) and TREC:T-cell ratio (Panel B) from blood spots collected from the same individuals. The Card #1A and Card #1B were collected in November 2012 and the Card #1C in July 2012. Onto each filter paper card, 10 samples were spotted. From each blood spot, a DNA solution was generated and TREC, ALB and VD-J levels were quantified in triplicate using qPCR. CT values obtained were averaged and used to calculate the two ratios that are presented. Each stratum line of the spider-web corresponds to the same TREC ratio value.

because there is a risk of underestimation of the T-cell count in the calculation. Thus, theoretically, the TREC:PBMC ratio should be preferred as better qPCR readout of the TREC ratio. However, some authors have reported the rearranged TCR-β gene count correlated well with the total lymphocyte counts in HIVseropositive adults (Redd et al., 2010) and in healthy young adults (Lang et al., 2011). Thus, complementary studies are still needed to elucidate which is the better readout and above all what will be that one able to accurately follow immunological changes in individual groups with wider age ranges and/or in

Altman, D.G., Bland, J.M., 1983. Measurement in medicine: the analysis of method comparison studies. Statistician 32, 307. Arnold, C.R., Wolf, J., Brunner, S., Herndler-Brandstetter, D., Grubeck-Loebenstein, B., 2011. Gain and loss of T-cell subsets in Old age — age related reshapping of the T-cell repertoire. J. Clin. Immunol. 31, 137. Aspinall, R., Pido-Lopez, J., Imami, N., Henson, S.M., Ngom, P.T., Morre, M., Niphuis, H., Remarque, E., Rosenwirth, B., Heeney, J.L., 2007. Old rhesus macaques treated with interleukin-7 show increased TREC levels and respond well to influenza vaccination. Rejuvenation Res. 10, 5. Betjes, M.G., Langerak, A.W., van der Spek, A., de Wit, E.A., Litjens, N.H., 2011. Premature aging of circulating T cells in patients with end-stage renal disease. Kidney Int. 80, 208. Chain, J.L., Joachims, M.L., Hooker, S.W., Laurent, A.B., Knott-Craig, C.K., Thompson, L.F., 2005. Real-time PCR method for the quantitative analysis of human T-cell receptor gamma and beta gene rearrangements. J. Immunol. Methods 300, 12. Chan, K., Puck, J.M., 2005. Development of population-based newborn screening for severe combined immunodeficiency. J. Allergy Clin. Immunol. 2, 391. Douek, D.C., Vescio, R.A., Betts, M.R., Brenchley, J.M., Hill, B.J., Zhang, L., Collins, R.H., Koup, R.A., 2000. Assessment of thymic output in adults after haematopoietic stem-cell transplantation and prediction of T-cell reconstruction. Lancet 355, 1875. Hazenberg, M.D., Stuart, J.W., Otto, S.A., Borleffs, J.C., Boucher, C.A., de Boer, R.J., Miedema, F., Hamann, D., 2000. T-cell division in human immunodeficiency virus (HIV)-1 infection is mainly due to immune activation: a

8

P.O. Lang et al. / Journal of Immunological Methods 389 (2013) 1–8

longitudinal analysis in patients before and during highly active antiretroviral therapy (HAART). Blood 95, 249. Hazenberg, M.D., Otto, S.A., de Pauw, E.S., Roelofs, H., Fibbe, W.E., Hamann, D., Miedema, F., 2002. T-cell receptor excision circle and T-cell dynamics after allogenic stem cell transplantation are related to clinical events. Blood 99, 3449. Hazenberg, M.D., Borghans, J.A.M., Boer, R.J., Miedema, F., 2003. Thymic output: a bad TREC record. Nat. Immunol. 4, 97. Ivarsson, M., Carlson, J., 2011. Extraction, quantification, and evaluation of function DNA from various sample types. Methods Mol. Biol. 675, 261. Lang, P.O., Mitchell, W.A., Govind, S., Aspinall, R., 2011. Real time-PCR assay estimating the naive T-cell pool in whole blood and dried blood spot samples: pilot study in young adults. J. Immunol. Methods 369, 133. Lang, P.O., Govind, S., Aspinall, R., 2012a. Reversing T cell immunosenescence: why, who, and how. Age (Dordr). http://dx.doi.org/10.1007/s11357-0129393-y (Feb 26. [Epub ahead of print]). Lang, P.O., Govind, S., Dramé, M., Aspinall, R., 2012b. Comparison of manual and automated DNA purification for measuring TREC in dried blood spot (DBS) samples with qPCR. J. Immunol. Methods 384, 118. Li, C.C., Beck, I.A., Seidel, K.D., Frenkel, L.M., 2004. Persistence of human immunodeficiency virus type 1 subtype B DNA in dried-blood samples on FTA filter paper. J. Clin. Microbiol. 42, 3847. Masciotra, S., Khamadi, S., Bilé, E., Puren, A., Fonjungo, P., Nguyen, S., Girma, M., Downing, R., Ramos, A., Subbarao, S., Ellenberger, D., 2012. Evaluation of blood collection filter papers for HIV-1 DNA PCR. J. Clin. Virol. 55, 101. McDade, T.W., Williams, S., Snodgrass, J.J., 2007. What a drop can do: dried blood spots as a minimally invasive method for integrating biomarkers into population-based research. Demography 44, 899. Mitchell, W.A., Lang, P.O., Aspinall, R., 2010. Tracing thymic output in older individuals. Clin. Exp. Immunol. 161, 497.

Motta, M., Chiarini, M., Ghidini, C., Zanotti, C., Lamorgese, C., Caimi, L., Rossi, G., Imberti, L., 2010. Quantification of newly produced B and T lymphocytes in untreated chronic lymphotic leukemia patients. J. Transl. Med. 8, 111. Ranek, L., 1976. Cytophotometric studies of the DNA, nucleic acid and protein content of human liver cell nuclei. Acta Cytol. 20, 151. Redd, A.D., Ciccone, E.J., Nakigozi, G., Keruly, J.C., Ndyanabo, A., Iga, B., Gray, R.H., Serwadda, D., Quinn, T.C., 2010. T-cell enumeration from dried blood spots by quantifying rearranged T-cell receptor-β genes. J. Immunol. Methods 354, 40. Rodgers, J.L., Nicewander, W.A., 1988. Thirteen ways to look at the correlation coefficient. Am. Stat. 42, 59. Schneider, R., 2010. T cells with commitment issues. Sci. Signal. 3, jc3. Sherman, G.G., Stevens, G., Jones, S.A., Horsfield, P., Stevens, W.S., 2005. Dried blood spots improve access to HIV diagnosis and care for infants in low-resource settings. J. Acquir. Immune Defic. Syndr. 38, 615. van Zelm, M.C., van der Burg, M., Langerak, A.W., van Dongen, J.J., 2011. PID comes full circle: applications of V(D)J recombination excision circles in research, diagnostics and newborn screening of primary immunodeficiency disorders. Front. Immunol. 2, 12. Verschuren, M.C., Wolvers-Tettero, I.L., Breit, T.M., Noordzij, J., van Wering, E.R., van Dongen, J.J., 1997. Preferential rearrangements of the T cell receptordelta-deleting elements in human T cells. J. Immunol. 159, 4341. World Health Organization (WHO), Initiative of vaccine research (IVR) of the Immunization, Vaccines and Biologicals department and the Ageing and Life Course (ALC) department. (2011) Report on the ad-hoc Consultation on Ageing and Immunization. 21–23 March 2011. In, Geneva, Switzerland, p. 1–33. Zubakov, D., Liu, F., van Zelm, M.C., Vermeulen, J., Oostra, B.A., van Duijn, C.M., Driessen, G.J., van Dongen, J.J., Kayser, M., Langerak, A.W., 2010. Estimating human age from T-cell DNA rearrangements. Curr. Biol. 20, R970.