We have been interested in multi-variate cytometry of the mammalian somatic cell cycle. The cell cycle can be analytically divided into a variable number of states based on the correlated levels of biochemical activities provided that the level of each activity changes in a repeating pattern and that the correlated activity patterns are not identical. For the simplest case of any single variable there are essentially 5 states to consider per oscillation: initial basal level (1), increase or decrease from basal (2), maximum or minimum (max, min) (3), and decrease or increase (4) to final basal level (5), which is not biologically equivalent to initial basal. For example, the expression of cyclins E1, A2, and B1 oscillate once per cycle, out of phase with each other. All are at "initial basal" in early G1; begin expression at different times and rise at different rates; reach max at G1/S (E1) or G2/M (A2, B1); decrease in S (E1), prometaphase (A2), or anaphase (B1); are resident at "final basal" approximately in G2 (E1), metaphase (A2), or telophase (B1) [1–4]. These expression patterns are readily discerned by immunofluorescence coupled with DNA content in flow cytometry assays . Thus, measurement of the cyclins, DNA, and a mitotic marker constitute highly informative analyses of cell cycle transition states characteristic of a specific population of somatic cells [e.g., [6–8]].
Because mitosis is characterized by abrupt, specific, sequentially timed proteolysis of substrates of the anaphase promoting complex/cyclosome (APC/C)  and an abrupt increase in kinase activities and phosphorylation of many substrates, a multi-variate cell-based approach to cell cycle analysis subdivides G2 and M in a straightforward manner [e.g., ]. G1 is also characterized by oscillating activities, but has not been analyzed cytometrically in the same manner as M. Here, we wished to extend to G1 this type of multi-variate analysis. One fundamental G1 sub-division is the kinetically defined uncommitted and committed states. The transition between these states has been labeled "R" (restriction point) by Yen and Pardee . The exact biochemical concept or nature of R is unknown, although most investigators would agree with a complex model that integrates signaling (growth and nutrition factors; cell-substrate attachment; cell-cell contact, and cell damage) at a modular level, with an "R" module containing at least the activities of D cyclin/Cdk complexes and the Rb/E2F family of transcription factors [10–14]. For a mathematical model that formally integrates a large body of information, see Novak and Tyson . For a dissenting view see .
Without an exact biochemical definition of "R", it may be possible to quantify cells that are in a pre-R or post-R state (~G1-pm and G1ps subphases of Zetterberg, ) based on measurements of specific Rb family phosphorylations; the levels of E2F transcription factors that are bound to promoters and not bound by Rb family proteins, and the levels of D and E cyclins. Previous work by Juan et al. has shown the feasibility of this approach with an antibody that binds "hypo-phosphorylated" Rb . At this time, a comprehensive analysis would be difficult to achieve since there are 3 Rb family proteins, three D cyclins using 2 cyclin dependent kinases (Cdk4, Cdk6), two E cyclins using Cdk2, and 8 E2F and 2 DP protein subunits that constitute E2F transcription factors as well as at least 5 protein inhibitors of G1 Cdks. Since much of this operates as a function of specific phosphorylations, this might be rendered simpler with probes to specific phosphorylated sites on a subset of these molecules. However, these probes are not generally available and may or may not function well in cytometric assays [e.g., see ].
In place of the more intricate and sophisticated approach outlined above, we have asked whether the bimodal distributions shown by bivariate analyses of chromatin-bound minichromosome maintenance (Mcm) proteins and DNA content marks pre- and post-R G1 cells. These patterns were first shown by Friedrich and co-workers  and are based on measuring the residue of Mcm proteins that are left behind after detergent extraction. In support of this idea, Mukherjee at al. have correlated the committed period of G1 with high levels of bound Mcm protein, hyper-phosphorylated Rb, and cyclin E expression . The approach employed by Freidrich et al. (extraction then fixation) was first described by Kurki et al.  for detecting S phase specific 'tightly' bound PCNA that was a subset of total PCNA detected by denaturing fixation. The logic is that extraction of live cells with detergent depletes cells of 'loosely' bound proteins and other molecules that can diffuse out, leaving large structures and tightly bound proteins, including those bound to chromatin. This is also close to the standard methods of creating chromatin pellets in the primary research on replication complexes.
Normal mammalian cell replication involves mechanisms to bias faithful duplication of the genome once per mitotic cell cycle. The sub-system responsible for preventing re-replication consists of a protein complex that is built on origins-of-replication (ORI) and "licensed" for initiation of DNA synthesis. A licensed pre-replication complex (pre-RC) is built on existing complexes of ORC proteins by sequential binding of Cdc6 then Cdt1/RFL-B, which then "loads" Mcm proteins. Licensing is complete after the 6 Mcm proteins are loaded onto chromatin as a functional but inactive hexameric complex. Cdt1 is rate limiting for this process. In replicating cells, licensing occurs as a continuous process from late mitosis, after loss of mitotic Cdk activity, through G1 [20, 21].
In higher eukaryotes somatic cells, the rates of licensing are not known, however, one model for actively dividing cells (based on a zone within the Chinese hamster DHFR locus) suggests that licensing begins in mitosis but origin site specification occurs in G1 . Models for cell cycle re-entry suggest that quiescent cells lack bound Cdc6 and Mcm proteins and therefore license in late G1 [e.g., [18, 23]]. On cell cycle re-entry, Mcm loading appears to be dependent on cyclin E and Cdc7 expression and activities , which peak late in G1. The rate-limiting step for initiation may be unwinding of DNA by the helicase activity of the complex composed of Mcm2 through 7 . Once DNA synthesis has been initiated, irrevocable rescinding of the license occurs by several mechanisms, including removal of Mcm proteins from the chromatin, that prevent reassembly of pre-RC in S phase. Reloading does not occur until the licensing process resumes in mitosis in cycling cells or in G1 in stimulated quiescent cells. This system has been reviewed in detail [e.g., see [20, 21, 24]].
The reported cytometric pattern of Mcm2 and DNA content for CV-1 cells provides a visualization of Mcm binding in G1 and Mcm removal from chromatin during S phase . The expression of detergent resistant Mcm2 was bimodal in G1 with low and high expressing cell clusters. The high expression cluster level was coincident with the G1/S border, suggesting that cells enter S phase only after the bound Mcm complexes have reached a maximum level (max). In the published patterns, the levels in S decreased to a minimum at the S/G2 border consistent with the known cell cycle related expression [20, 21, 24, 25].
Since origins of replication are not well mapped in higher eukaryotes, the timing of licensing and the rate of Mcm loading in actively dividing cells has not been comprehensively studied. The work of Friedrich et al. suggests that G1 cells cluster into two distinct groups with the highest cluster ~10 fold greater than the lower cluster. Since the frequency of cells existing in each cluster is proportional to the time spent at that state, CV-1 cells may pass through two G1 loading periods with a rapid transition between the two states. However, this type of pattern would also be consistent with exponential loading. While over-interpretation is not useful, the quantitative bimodality of G1 cells, if substantiated, creates a biochemically related division of G1 that could be the beginning of subdividing G1 into meaningful biochemical states that would increase the power of multi-variate analyses of the cell cycle.
Since the simplest interpretation of the CV-1 G1 patterns are an early and late subdivision, our goals were to 1) determine whether the same pattern exists in human somatic cells, and 2) determine whether the low and high clusters correlate with pre- and post-restriction point states. The second goal is an experimentally more complicated approach to verifying the early → late sequence, but if true, has the value of giving the measurement of tightly bound Mcm proteins in G1 two meanings. The first is as a measure of the functional state of the Mcm loading sub-system within the cellular/environmental context, and the second is as a surrogate marker for commitment to cell cycle progression.
Because a bivariate analysis of any immunoreactivity versus DNA content leaves an uncertainty at the G1/S interface proportional to the coefficient of variation (CV) of the two measures, we enhanced the probability that cells would be identified as either G1 or S by co-staining for tightly bound PCNA. S phase is associated with the release of the Mcm proteins and PCNA binding to replication forks (reviewed in: [26–28]). Tightly bound PCNA has also been shown by cytometry to correlate with incorporation of BrdU .