Impact of Serum Storage Conditions on Microrna Stability

LETTERS TO THE EDITOR Impact of serum storage conditions on microRNA stability

Leukemia (2012) 26, 2414–2416; doi:10.1038/leu.2012.106 intermediate-term storage (20 months) at À 20 versus À 80 1C on the abundance of individual miRNAs. Serum from patients with multiple myeloma (MM) was simultaneously frozen at À 20 1Cas As microRNAs (miRNAs) were found to be present in human serum well as À 80 1C and thawed after 20 months followed by RNA in a remarkably stable and cell-independent form, their potential extraction and quantification using the Bioanalyzer small RNA kit as novel noninvasive biomarkers for physiological and pathophy- (n ¼ 5) and qRT-PCR (n ¼ 3). No significant difference of total RNA siological conditions, including cancer, is of growing interest.1,2 concentration was detected (P ¼ 0.97) (Figure 1c). Out of the four Serum miRNAs can be found in exosomes3 and bound to selected miRNAs only miR-93 showed a significant decrease proteins, such as AGO2,4 increasing their stability and circulation (Po0.04) at À 20 1C (Figure 1d). These findings suggest differ- time. However, owing to limited space and availability of À 80 1C ences in the stability of individual serum miRNAs frozen at freezers, many hospitals and laboratories store patient serum different time points and temperatures. To further explore this samples at À 20 1C, for many years. We therefore investigated the matter, serum miRNAs derived from three MM patient samples, impact of short-term, intermediate-term and long-term storage at simultaneously frozen at À 20 and À 80 1C for 20 months, were À 80 and À 20 1C on the stability and abundance of individual quantified using Serum/Plasma Focus miRNA PCR panels (Exiqon, and total miRNAs in human serum. Copenhagen, Denmark). Of a total of 182 analyzed miRNA probes, Serum was obtained from coagulated whole blood after centri- only miR-20a* was undetectable in all patient samples. After inter- fugation at 2000 r.p.m. for 10 min. All patients gave informed consent plate calibration, 48 miRNA assays were removed from the analysis for the research use of their specimens, in accordance with the because of low and inconsistent expression across the samples. Declaration of Helsinki. The miRNA fraction was quantified using Of all analyzed miRNAs, the calculated DCt (average Ct À 201C À the Bioanalyzer small RNA kit (Agilent Technologies, Santa Clara, average Ct À 801C) values ranged between 4.19 and À 3.88 with CA, USA) according to the manufacturer’s protocol. This method has average DCt values of À 0.73 for miRNAs more abundant at a detection range of 50–2000 pg/ml (www.genomics.agilent.com). À 20 1C and 0.79 for miRNAs more abundant at À 80 1C All samples were analyzed in triplicates and on separate chips (Figure 1e). On average this reflects a less than twofold change to identify chip-to-chip variations. In addition, the expression assuming a 100% reaction efficacy. Assuming little changes or of selected miRNAs was tested by quantitative real time PCR even a decrease of individual miRNAs stored at À 20 1C, DCt (qRT-PCR) using Taqman amplicons (Life Technologies, Darmstadt, values X1 would be expected for all analyzed miRNAs. Germany). Four individual miRNAs (miR-223, miR-451, miR-93 However, only 27 (17.1%) miRNAs showed a DCt 41 and 14 and miR-24) were selected on the basis of their expression in miRNAs (8.2%) showed a DCt o À 1 (Figure 1e), potentially hematopoietic cells and amplification efficacy.5 miR-223 and miR- indicating that the storage temperature facilitated a release of 451 are highly abundant in mature myeloid6 and red blood cells,5 miRNAs from their protein or exosome-bound form in certain respectively, whereas miR-24 is ubiquitously expressed in all cases. The balanced increase and decrease of individual miRNAs hematopoietic lineages.5 miR-93 was included because of its might explain the lack of differences with regards to the overall previously published use as a housekeeping gene in serum amount of miRNAs between storage conditions at À 80 and samples.7 RNA was extracted from 200 ml of serum using Qiagen À 20 1C. These findings further highlight the stability of serum miRNeasy kit (Qiagen, Valencia, CA, USA) according to the miRNAs, suggesting that storage temperature is less relevant than manufacturer’s instructions, because of its superior efficacy and previously assumed for intermediate-term storage (p20 months). handling compared with other common RNA extraction protocols However, this data set also shows that several miRNAs are affected (results not shown). cDNA synthesis as well as qRT-PCR was by storage conditions. This implies that for the accurate performed according to the manufacturer’s instructions and all quantification of individual miRNAs, storage conditions should samples were analyzed in triplicate. Mean absolute Ct values and be kept constant within a cohort of samples. s.d. are displayed on the graphs. Statistics were performed using To explore the impact of long-term storage (up to 10 years) at Student’s t-test and Po0.05 was considered as significant. À 20 1C on serum miRNAs, matched samples from three MM To assess the effect of short-term storage on miRNA stability, patients, seen in our outpatient clinic in 2002, 2004, 2006, 2008 freshly isolated serum from three healthy donors was stored for 10 and 2010, were analyzed with the Bioanalyzer small RNA kit as well days at À 80 1C without thawing, at À 80 1C including a daily freeze/ as qRT-PCR to assess miRNA integrity. Only a small difference in thaw cycle, at À 20 1C without thawing and at room temperature miRNA levels could be observed after 2 and 4 years of storage (RT). No difference could be observed between storage at À 80 and time, whereas a significant decrease was detected after 6 years À 20 1C (Figure 1a), whereas storage at RT led to a significant (P ¼ 0.04) and a further decrease after 10 years (P ¼ 0.05; reduction of miRNA levels compared with storage at À 80 1C Figure 1f). Additionally, quantification of four individual miRNAs (P ¼ 0.0017) (Figure 1a). In contrast to previous reports,3 repeated by qRT-PCR confirmed the Bioanalyzer results (Figure 1g), showing freeze and thaw cycles also led to a significant decrease of miRNA the complementarity of both approaches. levels (P ¼ 0.0079) compared with continuous storage at À 80 1C These findings underline the stability of miRNAs in short- (10 (Figure 1a). Quantification of individual miRNAs by qRT-PCR days), intermediate- (p20 months) and long-term storages (p10 confirmed the Bioanalyzer results (Figure 1b). Remarkably, all four years) at À 80 and À 20 1C. To our surprise, storage at À 20 1C miRNAs were still detectable after 10 days at RT, highlighting the barely impacted the overall amount of miRNAs for at least 2–4 superior stability of serum miRNAs compared with serum mRNA.8,9 years with only minor changes in the levels of individual miRNAs. Owing to the fact that many institutions store serum samples Taken together, this study further highlights the potential of not uniformly at À 80 1C, we investigated the impact of miRNAs as biomarkers based on their stability in long-term stored

Accepted article preview online 16 April 2012; advance online publication, 22 May 2012 Letters to the Editor 2415 n.s. n.s.

300 p<0.01 20 p<0.01 p=0.017

200 25 p=0.05 pg/ul

100 Ct value 30 miR-24 miR-223 miR-93 miR-451 0 35

RT RT -80°C -20°C -80°C -20°C 10xF/T 10xFT storage condition storage condition

n.s. 300 20 -80°C -20°C 200 25 p=0.04 pg/ul

100 Ct value 30

0 35

-80°C -20°C miR-24miR-24miR-93miR-93 storage condition miR-223miR-223 miR-451 miR-451

n.s.

p=0.04 250 4 miR-20b 3 miR-22* 200 2 miR-297 n.s. 17.1% 150 1 35.0% p=0.05 0 pg/ul 100 ((Ct-20)-(Ct-80)) 36.9% -1 8.2% -2 50 miR-92a deltaCt -3 miR-125 -4 miR-629 0

2002 2004 2006 2008 2010 year of storage

n.s. 20 p<0.01 miR-24 p<0.01 miR-93 25 miR-223 miR-451 30 Ct value 35

2002 2004 2006 2008 2010 year of storage Figure 1. (a) Bioanalyzer quantification of small RNAs in samples from three healthy donors stored at different conditions for 10 days. (b) Four miRNAs quantified by qRT-PCR in the same samples (average Ct values). (c) Bioanalyzer quantification of small RNAs in samples from three MM patients stored at À 20 or À 80 1C for 20 months. (d) Comparison of four miRNAs in three MM samples stored at À 20 or À 80 1C for 20 months (average Ct values). (e) DCt (Ct À 201C–Ct À 801C) values for individual miRNA assays. (f) Bioanalyzer quantification of small RNAs in samples from three MM patients that were frozen at different time points. (g) Four miRNAs quantified by qRT-PCR in the same samples (average Ct values). 10 Â F/T ¼ 10 daily freeze ( À 80 1C)/thaw cycles.

& 2012 Macmillan Publishers Limited Leukemia (2012) 2414 – 2444 Letters to the Editor 2416 samples, even at conditions that are usually considered as 7These authors contributed equally to this work. incompatible with subsequent analyses of messenger RNA. 8These authors contributed equally to this work.

CONFLICT OF INTEREST REFERENCES The authors declare no conflict of interest. 1 Scholer N, Langer C, Dohner H, Buske C, Kuchenbauer F. Serum microRNAs as a novel class of biomarkers: a comprehensive review of the literature. Exp Hematol 2010; 38: 1126–1130. ACKNOWLEDGEMENTS 2 Scholer N, Langer C, Kuchenbauer F. Circulating microRNAs as biomarkers—true This study was supported by the Deutsche Krebshilfe grant (109420), the European blood? Genome Med 2011; 3:72. Hematology Association fellowship (2010/04) and by the Deutsche Forschungsge- 3 Mitchell PS, Parkin RK, Kroh EM, Fritz BR, Wyman SK, Pogosova-Agadjanyan EL et al. meinschaft grant (KU2288/3-1). Circulating microRNAs as stable blood-based markers for cancer detection Proc Natl Acad Sci USA 2008; 105: 10513–10518. S Grasedieck1,7, N Scho¨ler2,7, M Bommer1, JH Niess3, H Tumani4, 4 Arroyo JD, Chevillet JR, Kroh EM, Ruf IK, Pritchard CC, Gibson DF et al. Argonaute2 A Rouhi1, J Bloehdorn1, P Liebisch5, D Mertens1,HDo¨hner1, complexes carry a population of circulating microRNAs independent of vesicles in human plasma. Proc Natl Acad Sci USA 2011; 108: 5003–5008. C Buske6, C Langer1,8 and Florian Kuchenbauer1,6,8 1 5 Petriv OI, Kuchenbauer F, Delaney AD, Lecault V, White A, Kent D et al. Department of Internal Medicine III, Comprehensive microRNA expression profiling of the hematopoietic University Hospital of Ulm, Ulm, Germany; hierarchy. Proc Natl Acad Sci USA 2010; 107: 15443–15448. 2 Klinische Kooperationseinheit Dermato-Onkologie, 6 Kuchenbauer F, Mah SM, Heuser M, McPherson A, Ruschmann J, Rouhi A et al. German Cancer Research Center, Heidelberg, Germany; Comprehensive analysis of mammalian miRNA* species and their role in myeloid 3Department of Internal Medicine I, cells. Blood 2011; 118: 3350–3358. University Hospital of Ulm, Ulm, Germany; 7 Song J, Bai Z, Han W, Zhang J, Meng H, Bi J et al. Identification of suitable 4Department of Neurology, reference genes for qPCR analysis of serum microRNA in gastric cancer University Hospital of Ulm, Ulm, Germany; patients. Dig Dis Sci 2012; 57: 897–904. 8 Kopreski MS, Benko FA, Kwak LW, Gocke CD. Detection of tumor messenger 5Oncological Clinic Moers, Moers, Germany and 6 RNA in the serum of patients with malignant melanoma. Clin Cancer Res 1999; 5: Institute of Experimental Cancer Research, Comprehensive Cancer 1961–1965. Centre, University Hospital of Ulm, Ulm, Germany 9 Tsui NB, Ng EK, Lo YM. Stability of endogenous and added RNA in blood speci- E-mail: fl[email protected] mens, serum, and plasma. Clin Chem 2002; 48: 1647–1653.

SPARC is dispensable for murine hematopoiesis, despite its suspected pathophysiological role in 5q-myelodysplastic syndrome

Leukemia (2012) 26, 2416–2419; doi:10.1038/leu.2012.97 indications that SPARC may have a role in the hematopoietic system, we aimed to elucidate its role in the regulation of HSCs, as HSCs have not been characterized in the SPARC null mice. SPARC is a glycoprotein that binds calcium and interacts with All mature blood cell types arise from a limited number of members of the extracellular matrix including collagens and hematopoietic stem cells (HSCs). HSCs sustain hematopoiesis thrombospodin. SPARC is expressed in a variety of tissues under- throughout the lifetime of an individual and any insult to them going rapid turnover and is a major component of the bone. can lead to severe pathologies. Elucidating the molecular SPARC expression is associated with cell proliferation, migration, mechanisms that govern the maintenance and differentiation of changes in morphology, matrix-remodeling and cell–cell interac- HSCs is critical to understanding their role in hematological tions, although the precise function of the gene still remains pathologies. In a search for novel regulators of HSCs, we selected unknown.5 SPARC knockout mice (SPARC À / À ) are viable and SPARC (secreted protein acidic and rich in cysteine, also known as fertile but display a number of abnormalities, including increased Osteonectin and BM40), because it was identified as one of the deposition of fat, highly contractile skin, alterations in wound candidate regulators of HSCs by molecular profiling of differ- healing and an altered response to tumor growth, most severe of entiating hematopoietic cells in an in vitro model system.1 Several which are osteopenia and early onset cataract.6,7 SPARC À / À lines of evidence have also indicated an important role for mice generated previously by the targeted mutation of exon 6 SPARC in hematopoiesis. In zebrafish, a knockdown of SPARC that completely lacked SPARC mRNA and protein6 were used in resulted in lower numbers of circulating blood cells in developing this study to carry out functional assays for HSCs. embryos.2 In a murine model, SPARC deficiency resulted in Steady state hematopoiesis was unperturbed in SPARC À / À thrombocytopenia and reduced erythroid colony formation.3 mice as seen by normal peripheral blood counts (Supplementary In humans, SPARC is transcriptionally silenced in acute myeloid Table 1). A similar observation was made by Barlow et al.8 leukemia patients and in cell lines with mixed lineage leukemia using a mouse model carrying a different mutation in SPARC. Also, rearrangements.4 SPARC maps to the deleted region in the bone marrow (BM) of SPARC À / À mice showed normal 5q-myelodysplastic syndrome (MDS), an HSC disorder. With morphology and cellularity (Supplementary Figure 1). A thorough

Accepted article preview online 5 April 2012; advance online publication, 1 May 2012

Leukemia (2012) 2414 – 2444 & 2012 Macmillan Publishers Limited