Chemical and biological single cell analysis

Single cells represent the minimal functional unit of life. A major goal of biology is to understand the mechanisms operating in this minimal unit. Nowadays, analysis of the single cell can be performed at unprecedented resolution using new lab-on-a-chip devices and advanced analytical methods. While cell handling and cultivation devices can be classified into finite volume reactors and flow systems, the analytical approaches differ in respect to invasive (i.e. chemical) and noninvasive (i.e. biological/living cell) analysis. Using these new and exciting technologies cell-to-cell differences, originating from regulatory circuits and distinct microenvironments, can now be explored. For example, it could be shown that the rates of transcription and translation are stochastic. Chemical and biological single cell analyses provide an unprecedented access to the understanding of cell-to-cell differences and basic biological concepts.

Introduction

Most cellular research from fundamental cell biology and microbiology to applications in biotechnology is performed using cell populations with high cell numbers. This is convenient, as a huge variety of experimental techniques exist to analyze cell behavior. However, cell population data cover only information of an imaginary average cell (i.e. the average of the microbial culture), and not the mechanistic information of one cell. Properties of single cells within a population can be quantified by flow cytometry, deciphering cell-to-cell differences and thereby elucidating the heterogeneity of the population. Nevertheless, it is a snapshot analysis, reflecting the state of a cell at a certain moment in time. Neither the history nor the temporal development of a cell is traceable. Now, next to flow cytometry, developments are rapid and aim at the ultimate goal: spatiotemporal high-throughput analysis of single cells to mechanistically elucidate cellular functions. The enthusiasm of the field is reflected by the exponential increase of single cell studies and conferences in the last years. The development is pushed by the technical advances in microfluidic chip manufacturing and analytical methods, and pulled by the increasing demand from biologists.

We here define two alternative strategies for single cell analysis, which in the authors’ opinions should be ideally used in parallel as they deliver complementing information about cellular processes. First, live cell (biological) analysis allows the combination of cell cultivation and analysis, with the advantage of spatiotemporal analysis and possibility of subsequent cell retrieval. Single live cell analysis hence increases the toolbox of microbiologists and cell biologists. Second, in chemical analysis cells are lysed and their inventory is (partly) quantified [1^••].

After a detailed motivation why to study single cells, we briefly summarize latest technologies for single cell cultivation and analytics. It is important to note that most devices described are prototypes. In addition, we highlight some recent achievements of this emerging field of research.

Cells within populations are not identical and do not necessarily behave alike. For example, cell populations can be subject to rapid environmental changes (often referred to as stress or perturbation). During such changes it was observed that some subpopulations cope better with the new conditions, resulting in higher survival or even growth rates, with the outcome that the subpopulation forms the basis of the subsequent main population. Hence, it was suggested that such cell population heterogeneities secure the existence of species [2]. It is assumed that the cell-to-cell differences persist at all time and become relevant only during environmental perturbations [3]. Antibiotic resistance of bacteria, as an example, can originate from population heterogeneities [4, 5•]. A potential mechanism for cell-to-cell differences is feedback circuits [6], implying that population heterogeneity is actively generated by epigenetic mechanisms to secure higher fitness of the species under ever fluctuating conditions [4, 7]. Another mechanism generating population heterogeneities are stochastic events (also referred to as noise) [8••, 9]. Especially regulatory reaction cascades, in which only a small number of molecules are involved are a potential source of cell-to-cell differences, small changes in absolute numbers can result in high relative changes [10].¹ One can speculate that population heterogeneity is indicative of spatiotemporal fluctuations in protein concentration in single cells. Contrary to benign effects of population heterogeneity, very high noise levels can be harmful, since signaling pathways may become nonfunctional [10].

In conclusion, population heterogeneity is an inherent condition of cell populations, possibly as a result of stochasticity in gene expression and protein synthesis, existence of feedback loops, and differences in external stimuli (e.g. differences in the microenvironment). A better understanding of this heterogeneity will give mechanistic insight in cellular processes, ultimately resulting in new strategies for engineering biological systems, by minimizing cell population heterogeneities [11].

Arguably, for measuring cell-to-cell differences, one has to analyze cell functions on the basis of the single cell, as an average readout might blur these differences. In addition, precisely controlled microenvironments have to be used to decipher the individual contributions (e.g. microenvironment, stochasticity) to cell-to-cell differences. Such microenvironments can be realized on lab-on-a-chip (LOC) devices as outlined below. Ideally, analytical methods are integrated into the LOC devices, although few examples exist. The nature of analytes in single cell analysis does principally not differ from population analysis. What differs significantly, are the amounts of target molecules available and their variations in time and space. An insight into the small numbers of analytes present in a single cell is given in Table 1. Moreover, the analyte amount is dramatically lower in small bacteria as compared to relatively large mammalian cells. Despite higher total amounts of analytes in large cells, analyte gradients due to diffusion might have to be considered [12].

The targets of interest in single cell analysis are not only specific mRNAs or large numbers of mRNA species, specific proteins or many proteins, metabolites, but also biomass and metabolite production rates as they are evidence for fundamental cellular functions of transcription, translation, biochemical performance, and cellular fitness, respectively. Exemplarily for the latter, we estimated from population data the single cell production rates of lysine (0.2 fmol/cell/s, 1.3 × 10⁸ molecules/cell/s [13]), ethanol (6 fmol/cell/s, 3.8 × 10⁹ molecules/cell/s [14]), and IgG antibody (6 zmol/cell/s, 4000 molecules/cell/s [15]), which are synthesized by a bacterium, a yeast, and a high producing Chinese hamster ovary (CHO) cell, respectively (Table 1). For the latter, the production rate of 15,000 antibodies/cell/s, reported for mammalian lymphocytes [16], is not yet achieved by recombinant antibody production. As indicated, most cellular processes have multiple steps involved (e.g. information transport through a regulation cascade), hence requiring multitarget analyses to elucidate underlying molecular mechanisms. For some cell-to-cell difference studies, analysis of a single target might be sufficient.

The prospects of single cell analysis are plenty, as the challenges are still many and new opportunities will come up with available cultivation and analytical technology. Small total amounts of analytes and fluctuations in concentrations in time and space demand for specialized single cell cultivation devices and appropriate analytical techniques. The latest reports are summarized in the next sections.