Imaging flow cytometry for morphologic and phenotypic characterization of rare circulating endothelial cells

Abstract

Endothelial cells in the peripheral circulation are rare events that require technically rigorous approaches for detection by flow cytometry. Visualization of these cells has been even more demanding, as this has historically required extensive enrichment and processing prior to attempting imaging. As a result, few, if any, examples exist on images of peripheral blood circulating endothelial cells (CECs) that include verification of the cell lineage both phenotypically and genomically. In this study, we have devised a method whereby CECs can be directly visualized after lysis of red blood cells and staining, without pre-enrichment or additional processing. Peripheral blood is stained with CD45, CD146, CD3, Hoechst, and DAPI to permit identification of CD146 positive, nonleukocyte, nucleated, and live cells that fit the description of CECs. These cells are imaged using the Amnis ImageStream(X), an imaging flow cytometer. Genomic verification of the endothelial nature of these cells is accomplished by using an aliquot of the same stained samples for sorting CECs using similar gating strategies. This proof of principle of direct imaging of CECs by imaging flow cytometry will permit studies to be conducted heretofore not possible, as the ImageStream(X) has the capability of detecting additional fluorochromes other than those used to identify the CECs. Such potential investigations include antigen colocalization or capping, autophagy and apoptosis, morphologic changes in response to therapy, and others. Thus, this method will enable a broad range of novel studies to be conducted using CECs as surrogates of the endothelium.

Keywords: circulating endothelial cells; imaging cytometry; rare event cytometry.

Published 2013 Wiley-Periodicals, Inc.

Figures

Figure 1

Flow chart depicting concomitant CEC identification in sorting and IFC platforms. Peripheral blood from sickle cell patients was immunophenotypically stained for CEC identification. The sample was split into two aliquots whereby one was sorted on a MoFlo for Q-RTPCR and confirmation of endothelial gene expression in CECs (left column) and the other was imaged on the ImageStreamX for characterization of CECs (right column). Similarities and differences between the two platforms and their associated gating strategies are evident by comparing the two columns.

Figure 2

Gating strategy utilized to identify nucleated, live, single, CD3-, CD45-, CD 146+ CECs on ImageStreamX. For display purposes, 180,000 events are shown from merged files. Low side scatter mononuclear cells were selected using an Area_Brightfield versus Intensity_SSC plot. Nucleated cells were selected as Hoechst+ on an Intensity_Hoechst versus Intensity_SSC plot gated on low side scatter cells. Live cells were selected as 7AAD- on an Intensity_7AAD versus Intensity_SSC plot gated on nucleated cells. Single cells were selected on an Intensity_CD45 versus Aspect Ratio_Brightfield plot gated on live nucleated cells. Rounder objects and single cells have higher aspect ratios. Aggregates, two cells in one frame, or elongated cells have lower aspect ratios (see Figure 4, 3rd panel down). CD3 negative events were selected on an Intensity_CD3 versus Intensity_SSC plot gated on live nucleated single lymphocytes. CD146+CD45- CECs were selected on an Intensity_CD45 versus Intensity_CD146 plot gated on live nucleated single CD3- cells.

Figure 3

Gating strategy utilized to identify nucleated, live, single, CD3-, CD45- CD146+ CECs which fall within a tight Time gate on MoFlo. Four million total events are shown. Low side scatter mononuclear cells are identified on a FSC versus SSC plot. Nucleated events are selected as being Hoechst+ on a Hoechst versus SSC plot gated on low side scatter cells. Live cells were selected as being 7AAD- on a 7AAD versus SSC plot gated on nucleated cells. Events arising from a perturbation in fluidics are identified and gated out using a tight gate on a Raw Time versus SSC plot gated on live nucleated cells. Doublets and aggregates were identified and gated out using a CD45 versus Pulse Width plot gated on live nucleated single cells which fall within a tight Time gate. CD3- events are selected on a CD3 versus SSC plot gated on live nucleated single cells which fall within a tight Time gate. Finally, CD146+CD45-CECs are selected on a CD45 versus CD146 plot that is gated on live nucleated single CD3- cells which fall within a tight Time gate.

Figure 4

Representative imagery of cells gated out of the analysis, obtained at 60X magnification on ImageStreamX. To demonstrate the necessity of the sequential gating strategy for flow cytometric detection of CECs described in this manuscript, events not meeting the CEC gating criteria are shown originating from red boxes, and corresponding imagery is displayed. The top panel shows CD146+ Hoechst- anuclear cells. The second panel shows CD146+ 7AAD+ dead cells. The third panel shows CD146+ and CD146- cells arising from low aspect ratio together in one frame. The fourth panel shows CD146+ CD3+ activated T-cells. For display purposes, intensity of color was adjusted but raw data remained unchanged.

Figure 5

Representative imagery of CECs obtained at 60X magnification on the ImageStreamx Hoechst+ (nucleated), 7AAD- (live), single (high aspect ratio), CD3- (non-activated T-cells) CD146+CD45- CECs are shown originating from the red box, and corresponding CEC imagery in displayed. The bottom panel shows enlarged imagery of CD146 and nuclear merged images. For display purposes, intensity of color was adjusted but raw data remained unchanged.

Figure 6

Fold Changes of relative gene expression resulting from Q-RTPCR on sorted CECs, HUVECs, and T, and B cells. Sorted CECs and HUVECs showed mRNA expression levels consistent with endothelial genes while not showing high levels of leukocytic gene expression, while T and B cell controls showed high mRNA expression levels for corresponding leukocytic genes, but not for endothelial genes. Levels of gene expression are fold changes of A) CECs as compared to B cells B) CECs as compared to T cells C) HUVECs compared to B cells D) HUVECs compared to T cells E) B cells compared to CECs F) T cells compared to CECs G) B cells compared to HUVECs, and H) T cells compared to HUVECs.