β2-Microglobulin is an appropriate reference gene for RT-PCR-based gene expression analysis of hematopoietic stem cells

Abstract

Real-time reverse transcription polymerase chain reaction (RT-PCR) is regarded as one of the most useful and powerful tools for characterizing hematopoietic stem cells (HSCs), because samples of extremely small cell numbers can be analyzed. The expression levels determined by RT-PCR are based on relative quantification; therefore, the selection of an appropriate reference gene with a relatively stable expression level under most conditions is crucial. Here, we determined that beta2-microglobulin (B2m) is an appropriate reference gene for analyzing mouse HSCs by a novel method using single-cell RT-PCR. Clonally sorted HSCs were subjected to RT reactions with exogenous RNA fragments and then to real-time PCR. Next, the relative gene expression levels of 4 well-known housekeeping genes were quantified in each single cell sample based on the threshold cycle of exogenous RNA. The analysis revealed that B2m expression was reproducibly detected in almost all HSCs and that B2m had the most stable expression level among the compared genes, even after the cells had been cultured under various conditions. Thus, our results indicate that B2m can reliably be used as a reference gene for the relative quantification of expression levels in HSCs across various conditions. Furthermore, our work proposes a novel method for the selection of appropriate reference genes.

Keywords: Actb, beta-actin; B2m, beta2-microglobulin; Beta2-microglobulin; Ct, threshold cycles; ERCC, External RNA Controls Consortium; Gapdh, glyceraldehyde-3-phosphate dehydrogenase; HKGs, housekeeping genes; HSCs, hematopoietic stem cells; Hematopoietic stem cells; Hprt, hypoxanthine phosphoribosyl transferase; MHC, major histocompatibility complex; MPPs, multi-potential progenitors; RT-PCR, reverse-transcription polymerase chain reaction; Reference gene; SCF, stem cell factor; Single-cell RT-PCR; TPO, Thrombopoietin.

Fig. 1

Single-cell gene expression analysis strategy. (A) Experimental design of the single-cell gene expression assay. HSCs were clonally sorted into 96-well PCR plates containing initial reaction mix, including Ambion^® ERCC RNA Spike-In Mixes for quality control. Using the indicated primers, an RT reaction followed by an SMART reaction was used to synthesize first-strand cDNAs with adapter sequences at both the 5′ and 3′ ends. Subsequently, cDNA was amplified by 20 cycles of PCR using a primer containing the adapter sequences, and these amplicons were used as the templates for single-cell real-time RT-PCR. (B) For quality control of these templates, the threshold cycle (Ct) values of control ERCC RNA were examined by real-time PCR simultaneously with the target genes. As the first step of quality control, only single-cell samples that exhibited Ct values within ±2 cycles of the median Ct of each of 3 control ERCC RNAs in each independent experiment were used. In addition, samples that had no detectable signal for any gene were deemed to be empty samples (lacking a cell), attributed to a cell sorting error, and were removed from the analysis. These samples repeatedly showed negative signals for all target genes even when templates were further amplified through 10 additional cycles of PCR (data not shown).

Fig. 2

B2m expression could be stably detected in uncultured, fresh HSCs. The expression of the 4 indicated housekeeping genes was examined using single-cell RT-PCR in fresh HSCs (CD150⁺CD34⁻KSL cells). (A) Frequency of HSCs in which the expression of the indicated genes could be detected. The numbers in parentheses represent cells with positive signal/examined cells in an experiment; 3 independent experiments were performed. To investigate the influence of scaling down to the single-cell level, the expression levels of Actb (B), Gapdh (C) and Hprt (D) were compared between templates prepared from either 10,000 cells or single cells. The graphs represent the mRNA expression levels of indicated genes that were normalized to the Ct values of B2m. Data are shown as mean ± S.D. (single cell: n = 159; 10,000 cells: n = 3).

Fig. 3

B2m is stably expressed in HSCs, even after ex vivo culture. (A) CD150⁺CD34⁻KSL cells were cultured for 5 days under 3 different conditions in the presence of SCF and/or TPO. After the culture, CD48⁻KSL cells, referred to as the cultured HSC population , were clonally sorted and subsequently subjected to single-cell gene expression analysis. (B) Frequency of cells in which the expression of the indicated genes could be detected before and after culture. The numbers in parentheses represent cells that had positive signals/examined cells under each condition. (C) Heat maps showing the expression values of the indicated genes in individual HSCs before and after the culture under the 3 different conditions. The expression levels of these genes were normalized to the Ct values of ERCC-00130, an exogenously added control RNA. (Fresh: n = 159, SCF + TPO: n = 148, SCF: n = 163, TPO: n = 127).

Fig. 4

The expression level of B2m in HSCs is relatively independent of the influence of ex vivo culture. Using the results of single-cell real-time RT-PCR, changes in the expression levels of the indicated genes in HSCs were examined after ex vivo culture. The expression levels of these genes in individual single-cell samples were normalized to the Ct values of ERCC-00130, an external control RNA. The graphs show relative expression values of Actb (A), B2m (B), Gapdh (C) and Hprt (D). Data are shown as mean ± S.D. (**p < 0.05, Fresh: n = 159, SCF + TPO: n = 148, SCF: n = 163, TPO: n = 127).