New insights into cell cycle and DNA damage response machineries through high‐resolution AMICO quantitative imaging cytometry

Abstract

Objective

Progress in biology and medicine research is being driven by development of new instrumentation and associated methodologies which open analytical capabilities that expand understanding of complexity of biological systems. Application of cytometry, which is now widely used in so many disciplines of biology, is the best example of such a progress.

Methodology

Recent publications push the envelope in expanding capabilities of cytometry by introducing a high resolution imaging cytometry defined as Automated Microscopy for Image CytOmetry (AMICO). This instrumentation is utilized to further elucidate mechanisms of the cell cycle progression and also the DNA damage response. This approach is going beyond the presently possible analytical technologies regarding throughput and depth of information.

Conclusions

The possibility of multiparametric analysis combined with the high resolution mapping of individual constituents of cell cycle and DNA damage response machineries provides new tools to probe molecular mechanism of these processes. The capability of analysis of proximity of these constituents to each other offered by AMICO is a novel and potentially important approach that can be used to elucidate mechanisms of other biological processes.

Flow cytometry revolutionized progress in cell biology and medicine by offering the possibility of rapid and accurate measurements of numerous constituents, concurrently, in individual cells. A particular virtue of flow cytometry stems from the capability of multiparametric analysis that makes it possible to correlate quantities of such constituents with respect to each other. Applications of flow cytometry in the areas of the cell cycle, cell necrobiology and DNA damage response has yielded results that have helped comprehension of many details of intricate molecular mechanisms that drive the systems.

For deeper insight into the underlying processes, combination of multiple markers measured simultaneously in the same cell, and up to thousands or millions of cells, is essential. Using fluorescence‐based flow‐cytometry, cell cycle progression and onset of cell death can be characterized by ten or more markers simultaneously. This remarkable number of markers can even be exceeded in the new technology of mass cytometry, where combination of 30 or more characteristic phenotypic and cell cycle indicators can be analysed in a single run 1, 2. Although flow cytometry is highly sensitive in detecting even low intensity signals (thus low expression levels) and is unsurpassed in simultaneously measuring multiple markers in identical individual cells, it lacks detailed structural and locational information.

Further progress in research in these areas has thus been advanced by use of imaging cytometry, utilizing either laser scanning cytometry (LSC) image cytometry 3, 4, 5, 6, 7, 8, 9, 10, flow imaging cytometry 11, 12, 13 or quantitative confocal microscopy 4, 14. To expand capabilities of quantitative confocal microscopy in terms of analysis of proximal sites of DNA damage induced by anticancer drugs vis‐á‐vis DNA replication sites (‘replication factories’), further approaches have recently been developed utilizing computerized algorithms that allow measurement of spatial immediacy of these sites in 3D 15, 16. By applying quantitative nuclear imaging, even clinical diagnosis is within reach 17. All the above technologies have fair to excellent spatial resolution, but are relatively limited in their ability to combine quantification of multiple markers on individual cells in a single run. Technologies such as chip‐based explorative cytometry 18 can overcome this limitation by using multiple sequential staining and de‐staining steps, which can add virtually endless numbers of markers per individual cell. Unfortunately, such methods are relatively time consuming and only a limited number of cells, so far, can be analysed per sample.

Continuation of such progress has recently been reported by Furia et al. 19, 20, who describe a novel approach in high‐resolution imaging cytometry (schematic workflow, Fig. 1). Here, the authors defined it as Automated Microscopy for Image CytOmetry (AMICO) and utilized it to elucidate signalling pathways and molecular mechanisms associated with controls of cell cycle progression. They also investigated interaction of these pathways with DNA damage signalling pathways activated during constitutive (spontaneous) DNA injury in untreated mammary cells in relation to cell cycle phase. The latter was monitored by DNA content measurement combined with DNA replication (as revealed by incorporation of nucleotide precursor EdU), and with expression analysis of proliferation marker Ki‐67. Checkpoint activation markers p53 and p21, which coordinate cell cycle regulatory machinery with DNA damage response, were studied in parallel. Concurrently analysed also were primary markers of DNA damage response, by measurement of histone H2AX‐Ser139 phosphorylation and recruitment of p53BP1. Proximity Ligation Analysis 21 of images made it possible to measure proximity of detected DNA replication sites to DNA damage and repair foci, revealing the probable cause/effect association of molecular processes ongoing in these sites 19, 20.

Figure 1.

Schematic strategic workflow of the ‘Automated Microscopy for Image CytOmetry’ approach. Left to right: Following appropriate cell labelling, the software is fed images acquired from the automated microscope system. It then receives user‐guided definition of basic parameters to determine the type of analysis; this step is followed by corrections for imaging artefacts. Then, image analysis is performed and a single‐cell data repository is generated, which in turn yields flow‐ or image‐cytometry such as data analysis, to target cell populations and perform statistical analyses. Every individual cell can then be physically retrieved to be re‐analysed at high resolution, adopting appropriate image‐acquisition parameters. (Scheme adapted from reference 20).

The AMICO approach has advanced several interesting findings 19, 20. Constitutive (spontaneous) DNA damage sites, identified by presence of γH2AX and 53 binding protein 1 (53BP1) – which has previously been reported as caused primarily by endogenous oxidants – the by‐products of oxidative phosphorylation 22, 23 were found to be strongly spatially correlated with the DNA replication sites. Concurrent presence of γH2AX and 53BP1 in the same foci provides more definite evidence that these foci are markers of DNA double‐strand breaks than does the presence of γH2AX alone. However, similar degrees of correlation have been observed for frequency of γH2AX foci and colocalization of γH2AX with EdU foci, as well as for integrated value of nuclear expression of γH2AX in DNA‐replicating cells, when DNA damage occurred constitutively induced by endogenous‐, or exogenous‐ oxidants 4. Collectively, these data strongly support the notion of double‐strand breaks, potentially the most dangerous DNA lesions, constitutively being induced in untreated either normal 22, 23 or tumour (A549) cells 4 being generated within DNA replication sites, probably by endogenous oxidants.

In addition to data linking constitutive DNA damage with the cell DNA replication machinery, results showing association between checkpoint activation markers p53 and p21 with respect to Ki‐67 – reporter of the transition point between quiescence (G₀) and proliferation status (G₁) prior to cell entrance to S‐phase 19, 20 – are interesting and of importance. This has been achieved by elegant and informative ‘paint‐a‐gate’ multiparametric analysis of the data. Gating strategy allowed these authors to subdivide the total heterogeneous cell population progressing through the cell cycle into several discrete sections of the cycle, including early‐ mid‐ and late‐ S phase, and to correlate expression of these checkpoint‐ and proliferation‐ markers with respect to each other, and to particular sub‐sections of the cell cycle as well as with DNA replication. This is the first such approach in the literature and certainly has substantial merit, in terms of both methodology and in of yielding important biological information.

An attractive finding was the observation that while cells replicating DNA essentially did not express p21, they had a high degree of p53 expression. As upregulation of p21 is mediated by p53, these findings are consistent with the position that not abundance of p53, but status of its phosphorylation, controls intracellular level of p21 24. This is also consistent with the observation by LSC, using phospho‐specific Ab, that after DNA damage by UV light, p53 becomes phosphorylated on Ser15, subsequent to activation of ATM and phosphorylation of histone H2AX, at times of cell arrest 25.

From a technological point of view, microscopic components of mosaic image acquisition and analysis combine image cytometry of the whole cell population by wide field microscopy. Subsequently, detailed target‐driven sub‐cellular analysis of individual cells is possible using a confocal setup. Method optimization is described by the authors and includes different filter settings, labelling, microscope objectives, and image pre‐processing, among others. Results obtained by AMICO are comparable to those obtained by flow and LSC with regard to data display, population distribution and signal resolution. The methodology can even be extended to analysis of live cells or detection of hot‐spots and repair, or proliferation foci, in tissue sections 26.

They applied the AMICO approach 19, 20 to normal unperturbed cells in which constitutive DNA damage was being primarily caused by endogenous oxidants 27, 28. However, it is ideally suited to test mechanisms of DNA damage response in cells treated with anticancer drugs, particularly those targeting DNA, perturbing cell cycle progression and inducing apoptosis 14, 29, 30, 31. Application of this approach may lead to better understanding of involvement of different signalling pathways engaged in DNA damage repair and activation of checkpoint mechanisms in response to drug treatment. Comprehension of these mechanisms is of clear importance in designing strategies to treat cancer and design new drugs. This is of particular importance currently, with development of personalized approaches to treating cancer 32. In designing the most effective treatment strategies, knowledge of signalling pathways characteristic of particular cancers, related to cell proliferation, and that have proclivity to undergo apoptosis, is as valuable as that obtained from genome‐DNA sequencing analyses.