Identification of New Rat Bone Marrow-Derived Population of Very Small Stem Cell with Oct-4A and Nanog Expression by Flow Cytometric Platforms

Abstract

Very small embryonic-like stem cells (VSELs) represent a unique rare population of adult stem cells (SCs) sharing several structural, genetic, biochemical, and functional properties with embryonic SCs and have been identified in several adult murine and human tissues. However, rat bone marrow- (BM-) derived SCs closely resembling murine or human VSELs have not been described. Thus, we employed multi-instrumental flow cytometric approach including classical and imaging cytometry and we established that newly identified population of nonhematopoietic cells expressing CD106 (VCAM-I) antigen contains SCs with very small size, expressing markers of pluripotency (Oct-4A and Nanog) on both mRNA and protein levels that indicate VSEL population. Based on our experience in both murine and human VSEL isolation procedures by fluorescence-activated cell sorting (FACS), we also optimized sorting protocol for separation of CD45(-)/Lin(-)/CD106(+) rat BM-derived VSELs from wild type and eGFP-expressing rats, which are often used as donor animals for cell transplantations in regenerative studies in vivo. Thus, this is a first study identifying multiantigenic phenotype and providing sorting protocols for isolation VSELs from rat BM tissue for further examining of their functional properties in vitro as well as regenerative capacity in distinct in vivo rat models of tissue injury.

Figure 1

Expression of CD54 (ICAM-I) and CD106 (VCAM-I) on rat BM cells. (a) Gating strategy for identification of CD54⁺ and CD106⁺ populations by classical flow cytometry. Total nucleated cells (TNCs) derived from WT rat BM were stained for CD45 (PE-Cy7), hematopoietic lineages markers (Lin: TCRαβ, CD3, CD11b, CD45RA; FITC), and CD54 (PE) or CD106 (PE) and further analyzed by LSR II (Becton Dickinson). BM cells are visualized on dot-plot showing FSC versus SSC signals, which are related to the size and granularity/complexity of the cell, respectively. 2% of total 1 × 10⁶ of analyzed TNCs is only displayed in this dot-plot to visualize the population distribution. Objects from region P1 (lymphgate including FSC^low/SSC^low objects) were further analyzed for Lin markers expression and only Lin⁻ events are included into region P2. Upper, middle dot-plot cells stained for CD54, derived from gate P2, and analyzed for CD54 versus CD45 expression. Two fractions of cells potentially enriched in stem/progenitor cells are gated: rHSCs, CD45⁺/Lin⁻/CD54⁺ (region P3), and rVSELs, CD45⁻/Lin⁻/CD54⁺ (region P4). Both populations are further “back-gated” on FSC versus SSC dot-plot to visualize cell size distribution. Lower, middle dot-plot cells stained for CD106, derived from gate P2, and analyzed for CD106 versus CD45 expression; rHSC, CD45⁺/Lin⁻/CD106⁺ (region P3), and rVSELs, CD45⁻/Lin⁻/CD106⁺ (region P4). Both populations are further “back-gated” on FSC versus SSC dot-plot to visualize cell size distribution. Percentages show content of each subpopulation among TNCs in representative sample. Total 1 × 10⁶ of TNCs was typically collected for each sample to identify the SC populations. (b) Representative images of nonhematopoietic cells expressing CD54 and CD106 by ImageStream X system (ISS-X). TNCs were stained for Lin markers (FITC, green), CD45 (Alexa Fluor 647, blue), CD54 (PE, orange), or CD106 (PE, orange). Cells were fixed and nuclei were stained with Hoechst 33342 prior to analysis by ISS. Extracellular expression of CD106 and CD54 is shown on combo as well as on magnified images (right panels). The scale bars indicate 10 μm.

Figure 2

Expression of mRNA for Oct-4, Nanog, and Rexo-1 by real-time RT-PCR in sorted fractions of BM cells. Indicated populations of nonhematopoietic (CD45⁻/Lin⁻/CD106⁺, CD45⁻/Lin⁻/CD106⁺/CD54⁺, and CD45⁻/Lin⁻/SSEA-1⁺) and hematopoietic (CD45⁺/Lin⁻/CD106⁺, CD45⁺/Lin⁻/CD106⁺/CD54⁺) cells were sorted with MoFlo XDP. The graphs show the fold difference in concentration of mRNA for Oct-4, Nanog, and Rexo-1 in sorted fractions when compared to unfractionated TNCs (shown as 1). Results are presented as mean ± SD. Statistically significant differences (P < 0.05) are shown when compared with TNCs. Analysis was performed three times with samples prepared from three independent sorts.

Figure 3

Expression of Oct-4A and Nanog in selected BM populations on protein level. Representative images of Oct-4A- and Nanog-expressing CD45⁻/Lin⁻/CD106⁺ cells by ImageStream X system. TNCs were stained for (1) Oct-4A (Alexa Fluor 488, green), CD45 (PE-Cy7, magenta), and CD106 (PE, orange) (a) and (2) Nanog (Alexa Fluor 647, blue), Lin markers (FITC, green), CD45 (PE-Cy7, magenta), and CD106 (PE, orange) (b). Cell were stained with Hoechst 33342 to visualize nuclei and analyzed by ISS to detect intranuclear expression of Oct-4A (a) and Nanog (b) as shown in magnified, combined images. The scale bars indicate 10 μm.

Figure 4

Quantitative analysis of morphological features of selected rat BM populations by ImageStream X system. (a) Quantitative analysis of cell size, nuclear to cytoplasmic (N/C) ratio, and cytoplasmic area of rVSELs when compared to other cell types. Analysis was performed based on the collected images of following rat BM populations: (i) rVSELs: CD45⁻/Lin⁻/CD106⁺ cells expressing Oct-4A or Nanog (upper and lower table, resp.); (ii) rHSCs: CD45⁺/Lin⁻/CD106⁺ cells with no expression of Oct-4A and Nanog, and (iii) TNCs (gated as singlets during the analysis). Analyses were performed with IDEAS software (Amnis Corp.). Data are shown as average values of each indicated feature within the cell population (mean ± SD). (b) Analytical approach for masking and feature calculation by IDEAS. Images show brightfield (BF), nuclear (Hoe), and combined (BF-Hoe) images of one example rat BM cell as well as exact masks (cyan) and features (formulas based on the defined masks) that were optimized and used for the analysis shown in (a). Cell size was expressed as minor cell axis computed within the mask covering BF image, while cytoplasmic area and N/C ratio were calculated based on the areas of BF and nuclear images (as indicated in red boxes). All masks and features including their nomenclature represent standard parameters provided by IDEAS software (Amnis Corp.)

Figure 5

Expression of Oct-4A and Nanog in sorted fractions of rVSELs (CD45⁻/Lin⁻/CD106⁺) and rHSCs (CD45⁺/Lin⁻/CD106⁺) by immunocytochemistry. Both fractions were sorted with MoFlo XDP cell sorter (Beckman Coulter) and further stained for Oct-4A (Alexa Fluor 488, green) (a) or Nanog (Alexa Fluor 546, red) (b) and analyzed with Leica DMI6000B (ver. AF7000) fluorescent microscope. Nuclei are stained with DAPI. Intranuclear staining for both transcription factors is visualized on combo images. The scale bars indicate 10 μm.

Figure 6

Optimized protocol for isolation of rVSELs from WT and eGFP-expressing rat BM tissue. Rat BM-derived VSELs were isolated from full population of BM cells stained for CD45 (PE-Cy7), Lin markers (TCRαβ, CD3, CD11b, and CD45RA; Alexa Fluor 647), and CD106 (PE) by MoFlo XDP cell sorter (Beckman Coulter). (a) Total nucleated cells (TNCs) are visualized on dot-plot showing FSC-H versus SSC-H. (b) Small agranular cells from gate R1 (with extension of lymphgate into low values of FSC; Figure 5(a)) are plotted on FSC-W versus FSC-H dot-plot to exclude doublets. (c) Single cells from gate R2 are subsequently analyzed for Lin markers expression. (d) Lin⁻ events included in region R3 are further plotted on dot-plot showing CD106 expression versus side scattered characteristics (SSC-H) of these cells. (e) CD106⁺ cells from gate R4 are eventually visualized on dot-plot based on their CD45 expression and FSC signal (FSC-H). Rat HSCs are identified as CD45⁺/Lin⁻/CD106⁺ (region R6), while rVSELs as FSC^low/SSC^low/CD45⁻/Lin⁻/CD106⁺ cells (region R5).