Quantification of mRNA in Whole Blood by
Assessing Recovery of RNA and Efficiency of
Masato Mitsuhashi,1,2*Shigeru Tomozawa,2,3Katsuya Endo,4and Atsushi Shinagawa5
Background: Current gene expression analysis relies on
the assumption that the isolated RNA represents all
species of mRNA in proportions equal to those in the
original materials. No system is available for absolute
quantification of mRNA.
Methods: We applied whole blood to 96-well filter-
plates to trap leukocytes. Lysis buffer containing cock-
tails of specific reverse primers and known concentra-
tions of synthetic external control RNA (RNA34) was
added to filterplates, and cell lysates were transferred to
oligo(dT)-immobilized microplates for hybridization.
We then synthesized the cDNA in the oligo(dT)-immo-
bilized microplates from these primer sites and used the
cDNA for real-time PCR. RNA34 acted as a universal
control, and gene amplification results were converted
to quantities of mRNA per microliter of whole blood
after the recovery of RNA34 in each sample was deter-
Results: Under fully optimized conditions, both added
RNA34 and native mRNA species exhibited ?10% re-
covery from whole blood to real-time PCR. When whole
blood was stimulated ex vivo, changes in gene expres-
sion as low as 30%–40% were detected with statistical
significance, and the experimental CVs were low (10%–
Conclusion: This new system to estimate mRNA copies
per microliter of whole blood may allow standardiza-
tion of gene-expression–based molecular diagnostics.
© 2006 American Association for Clinical Chemistry
Methods to quantify mRNA expression include Northern
blot analyses (1), RNase protection assay (2), PCR (3),
nucleic acid sequence–based amplification (4), and
branched-DNA amplification (5). With current real-time
PCR methods, the amounts of various gene-specific
cDNA sequences can be determined in each reaction
vessel (6). Because the efficiencies of RNA extraction and
cDNA synthesis in each sample are unknown, however,
the quantities of various mRNA molecules per milliliter
blood (7, 8) or per cell (9) “determined” by these methods
are presumed. Moreover, no information is available
regarding whether the purified RNA or synthesized
cDNA represents all species of mRNA as they exist in the
starting materials. Although mRNA data are expressed as
means and SD and statistical analysis is used, values are
derived from multiple determinations of the final stage of
gene amplification alone and therefore do not represent
the whole procedure. Microarray chip technologies (10)
can analyze many mRNA molecules simultaneously, but
the results are dependent on the quality of purified RNA
and amplified cDNA. The quantity of extracted RNA is
measured by absorbance at 260 nm and the quality is
assessed by gel electrophoresis to confirm the presence of
rRNA bands (1); however, whether the purified RNA
represents all species of mRNA in the same proportions
present in the original material is unknown.
The reporting of mRNA quantification results as quan-
tities of target mRNA per microgram of total RNA may be
inaccurate because mRNA represents only 1%–5% of total
RNA. Furthermore, mRNA concentration can vary even
when the total RNA concentration is constant. Yields of
total RNA or mRNA also vary widely depending on the
isolation method used (7). In HIV viral load tests (11),
total RNA is extracted from patient plasma samples to
which control RNA chosen to be similar in length and
sequence to the target RNA is added. No control RNA
that can act as a universal internal standard has been
Relative quantification by comparison of target gene
data to data for housekeeping genes or rRNA is used
1Hitachi Chemical Research Center, Inc., Tokyo, Japan.
2Department of Pathology, College of Medicine, University of California
at Irvine, Irvine, CA.
3Department of Surgical Oncology, Faculty of Medicine, University of
Tokyo, Tokyo, Japan.
4Hitachi Chemical Co., Ltd., Tokyo, Japan.
5Hitachi, Ltd., Hitachi General Hospital, Tokyo, Japan.
* Address correspondence to this author at: 1003 Health Sciences Road,
Irvine, CA 92617. Fax 949-725-2727; e-mail mmitsuhashi@HCRcenter.com.
Received January 31, 2005; accepted January 20, 2006.
Previously published online at DOI: 10.1373/clinchem.2005.048983
Clinical Chemistry 52:4
Papers in Press. First published February 23, 2006 as doi:10.1373/clinchem.2005.048983
Copyright © 2006 by The American Association for Clinical Chemistry
widely, although the quantities of control genes may
change during experiments (12). This method relies on
the assumption that both target RNA and internal stan-
dard RNA are extracted with the same efficiency and have
the same rate of cDNA synthesis. Because of a lack of
standardization, however, it is difficult to compare results
from one experiment with results obtained in another
experiment or with published results (13). As the number
of gene expression studies continues to grow, the need for
standardization becomes greater. We investigated a novel
method to accurately quantify the amount of human
leukocyte-specific poly(A)?mRNA per microliter of
whole blood as a model for standard gene expression
Materials and Methods
Primers and TaqMan probes were designed by Primer
(RNAture). DNA oligonucleotides were purchased from
Applied Biosystems, IDT, Proligo, TriLink, or GeneScript.
The sequences of TaqMan probes, primers, and templates
are summarized in Table 1 of the Data Supplement that
accompanies the online version of this article at http://
preparation of rna
Template oligonucleotides and cDNA were amplified by
PCR with T7 forward primers and oligo(dT) reverse
primers (see Table 1 in the online Data Supplement) with
30–35 cycles of 95 °C for 30 s (denaturing), 55–65 °C for
30 s (annealing), with final extension at 72 °C for 1 min.
The PCR products were analyzed by agarose gel electro-
phoresis to confirm a clear single band of the expected
size. PCR products were then applied to spin columns
(SigmaSpin; Sigma) to remove free nucleotides and prim-
ers, extracted twice with organic solvents (phenol, chlo-
roform, and isoamyl alcohol; Invitrogen), and then sub-
jected to ethanol precipitation. RNA was then synthesized
with an in vitro transcription system (T7 RiboMax; Pro-
mega) at 37 °C for 30 min, followed by three 10-min
DNase treatments. Each synthetic RNA was extracted
with organic solvents and precipitated with ethanol as
described above. The final RNA product was suspended
in nuclease-free water, and the absorbance was measured
at 260 nm [Ultroapec 3100 pro (Amersham) and DU7400
(Beckman-Coulter)]. The lengths of the RNAs [109 for
RNA34, 775 for CD4, 473 for p21, and 982 for Fas ligand
(FasL)6] were confirmed by microchip electrophoresis
(iChip; Hitachi Chemical), and the peak areas were ?70%.
RNA34 was also synthesized by Dharmacon with 86%
purity by HPLC analysis. The 40mer poly(A) tail was used
because preliminary experiments showed similar recov-
eries for the 20-, 40-, and 60-mer (data not shown). The
cDNA was synthesized from each RNA under different
conditions [e.g., varying the concentration of Moloney
murine leukemia virus (MMLV), the incubation time, and
the primer/template ratio] to find the maximum yield,
followed by TaqMan real-time PCR with known concen-
trations of HPLC-purified DNA oligonucleotides as cali-
brator. The molar quantities of cDNA were then deter-
mined as the amounts of RNA based on the assumption
that cDNA synthesis efficiency was 100% under optimum
The blood protocol used was approved by the Institu-
tional Review Board. Blood samples were collected at the
University of California, Irvine, and Hitachi General Hos-
pital from adult volunteers after written informed consent
was obtained. After collection samples were stored at 4 °C
The assay procedure consists of 3 major steps, as
shown in Fig. 1: (a) leukocyte isolation and lysis on
filterplates; (b) mRNA isolation, reverse primer hybridiza-
tion, and cDNA synthesis in oligo(dT)-immobilized mi-
croplates; and (c) real-time quantitative PCR. The custom
96-well filterplates were manufactured by Whatman or
Pall by assembly with leukocyte reduction membranes
(Leukosorb; Pall). These filterplates were placed over
collection plates, and 150 ?L of 5 mmol/L Tris (pH 7.4)
was applied to wet the filter membranes. After centrifu-
gation at 120g for 1 min at 4 °C to remove the Tris solution
from the membranes, 50 ?L of well-mixed whole blood
sample was applied to each well and immediately centri-
fuged at 120g for 2 min at 4 °C. The wells were then
washed once with 300 ?L of phosphate-buffered saline.
After centrifugation at 2000g for 5 min at 4 °C to remove
the saline solution, 60 ?L of stock lysis buffer [5 g/L
mmol/L Tris-HCl (pH 7.4), 1 mmol/L EDTA, 1 mL/L
IGEPAL CA-630 (substitute of NP-40), 1.79 mol/L guani-
dine thiocyanate (all from Sigma)], supplemented with 1
mL/L 2-mercaptoethanol (Bio-Rad), 0.5 g/L proteinase K
(Pierce), 0.1 g/L salmon sperm DNA (5 Prime Eppen-
dorf/Brinkman), 0.1 g/L Escherichia coli tRNA (Sigma), 5
nmol/L each of the specific reverse primers, and 109
molecules/L of synthetic RNA34 (as external control),
was added to each well of the filterplates. The plates were
then incubated at 37 °C for 10 min, placed over oligo(dT)-
immobilized microplates (GenePlate; RNAture), and cen-
trifuged at 2000g for 5 min at 4 °C. After overnight storage
at 4 °C, the microplates were washed 3 times with 100 ?L
of plain lysis buffer and then 3 times with 150 ?L of wash
buffer [0.5 mol/L NaCl, 10 mmol/L Tris (pH 7.4) 1
mmol/L EDTA] at 4 °C.
cDNA was synthesized directly in each well by addi-
tion of 30 ?L of buffer containing 1? reverse transcription
buffer [50 mM KCl, 10 mM Tris-HCl (pH 8.3), 5.5 mM
6Nonstandard abbreviations: FasL, Fas ligand; MMLV, Moloney murine
leukemia virus; PBMC, peripheral blood mononuclear cell; Ct, threshold cycle;
ACD, acid-citrate-dextrose; PMA, phorbol 12-myristate 13-acetate; CaI, cal-
cium ionophore; PHA, phytohemagglutinin-P; and IL-2, interleukin-2.
Mitsuhashi et al.: Method for Quantifying mRNA in Whole Blood
12. Glare EM, Divjak M, Bailey MJ, Walters EH. ?-Actin and GAPDH
housekeeping gene expression in asthmatic airways is variable
and not suitable for normalizing mRNA levels. Thorax 2002;57:
13. Vlems FA, Ladanyi A, Gertler R, Rosenberg R, Diepstra JH, Roder
C, et al. Reliability of quantitative reverse-transcriptase-PCR-
based detection of tumour cells in the blood between different
laboratories using a standardized protocol. Eur J Cancer 2003;
14. Slifman NR, Loegering DA, McKean DJ, Gleich GJ. Ribonuclease
activity associated with human eosinophil-derived neurotoxin and
eosinophil cationic protein. J Immunol 1986;137:2913–7.
15. Rainen L, Oelmueller U, Jurgensen S, Wyrich R, Ballas C, Schram
J, et al. Stabilization of mRNA expression in whole blood samples.
Clin Chem 2002;48:1883–90.
16. de Vries TJ, Fourkour A, Punt CJ, Ruiter DJ, van Muijen GN.
Analysis of melanoma cells in peripheral blood by reverse tran-
scription-polymerase chain reaction for tyrosinase and MART-1
after mononuclear cell collection with cell preparation tubes: a
comparison with the whole blood guanidinium isothiocyanate RNA
isolation method. Melanoma Res 2000;10:119–26.
17. Wingo ST, Ringel MD, Anderson JS, Patel AD, Lukes YD, Djuh YY,
et al. Quantitative reverse transcription-PCR measurement of
thyroglobulin mRNA in peripheral blood of healthy subjects. Clin
18. Ishikawa T, Ichikawa Y, Miura Y, Momiyama M, Keller, C, Koo K,
et al. Construction of cDNA bank from biopsy specimens for
multiple gene analysis of cancer. Clin Chem 1997;43:764–70.
19. Chomczynski P, Sacchi N. Single-step method of RNA isolation by
acid guanidinium thiocyanate-phenol-chloroform extraction. Anal
20. Boom R, Sol CJ, Salimans MM, Jansen CL, Wertheim-van Dillen
PM, van der Noordaa J. Rapid and simple method for purification
of nucleic acids. J Clin Microbiol 1990;28:495–503.
21. Eggesbo JB, Hjermann I, Hostmark AT, Kierulf P. LPS induced
release of IL-1 ?, IL-6, IL-8 and TNF-? in EDTA or heparin
anticoagulated whole blood from persons with high or low levels of
serum HDL. Cytokine 1996;8:152–60.
22. Sattentau QJ, Dalgleish AG, Weiss RA, Beverley PC. Epitopes of
the CD4 antigen and HIV infection. Science 1986;234:1120–3.
23. Liu S, Bishop WR, Liu M. Differential effects of cell cycle regulatory
protein p21(WAF1/Cip1) on apoptosis and sensitivity to cancer
chemotherapy. Drug Resist Updat 2003;4:183–95.
24. Sabelko-Downes KA, Russell JH, Cross AH. Role of Fas–FasL
interactions in the pathogenesis and regulation of autoimmune
demyelinating disease. J Neuroimmunol 1999;100:42–52.
25. Mitsuhashi M, Cooper A, Ogura M, Shinagawa T, Yano K,
Hosokawa T. Oligonucleotide probe design–a new approach.
26. Artuso M, Esteve A, Bresil H, Vuillaume M, Hall J. The role of the
ataxia telangiectasia gene in the p53, WAF1/CIP1(p21)- and
GADD45-mediated response to DNA damage produced by ionising
radiation. Oncogene 1995;11:1427–35.
27. Gillis S, Smith KA, Watson J. Biochemical characterization of
lymphocyte regulatory molecules. II. Purification of a class of rat
and human lymphokines. J Immunol 1980;124:1954–62.
28. Hartel C, Schumacher N, Fricke L, Ebel B, Kirchner H, Muller-
Steinhardt M. Sensitivity of whole-blood T lymphocytes in individ-
ual patients to tacrolimus (FK 506): impact of interleukin-2 mRNA
expression as surrogate measure of immunosuppressive effect.
Clin Chem 2004;50:141–51.
29. Giese T, Zeier M, Schemmer P, Uhl W, Schoels M, Dengler T, et al.
Monitoring of NFAT-regulated gene expression in the peripheral
blood of allograft recipients: a novel perspective toward individu-
ally optimized drug doses of cyclosporine A. Transplantation
30. Hamaguchi Y, Aso Y, Shimada H, Mitsuhashi M. Direct reverse
transcription-PCR on oligo(dT)-immobilized polypropylene micro-
plates after capturing total mRNA from crude cell lysates. Clin
31. Mitsuhashi M, Keller C, Akitaya T. Gene manipulation on plastic
plates. Nature 1992;357:519–20.
Clinical Chemistry 52, No. 4, 2006