Investigating cell cycle control mechanisms in CD8+ effector T cells

Abstract

CD8+ T cells have an important role in adaptive immunity. By possessing cytotoxic potential, CD8+ T cells are key effectors of the adaptive immune system, targeting diseased cells for destruction. CD8+ T cells are a key focus in the growing field of immuno-oncology because of their role in tumour cell killing. Naïve CD8+ T cells exist in a state of quiescence. Upon antigen recognition and co-stimulation, the CD8+ T cell escapes from this state of dormancy and towards a complex programme of cellular growth, cell cycle entry, and differentiation, leading to the production of short-lived effector T cells (SLECs), and memory progenitor cells (MPECs). The memory T cell pool which remains shares a lot of characteristics with the naïve cell including a capacity to remain dormant, engage in self-renewal, and maintain proliferative potential and multipotency upon restimulation, in a manner very similar to stem cells. Another characteristic shared with stem cells is a capacity for rapid proliferation. Like stem cells, at the peak of expansion, CD8+ T cells are also predominantly found in S phase, however it is not currently known whether their cell cycle regulation is similar to stem cells. There is also a good body of evidence showing a correlation between the rate of proliferation and the route of differentiation in CD8+ T cells. Cells which rapidly re-enter S-phase post mitosis favour differentiation towards the SLEC phenotype, while cells which stall in G1 phase favour the MPEC phenotype. However, the exact nature of this correlation, and how such a switch in cell cycle regulation is governed during the expansion phase is not well understood. This is in part because CD8+ T cells are resistant to cell cycle inhibiting agents, making the synchronisation studies seen in classic cell cycle analysis experiments not possible to do with CD8+ T cells.

In order to investigate how the cell cycle is regulated in CD8+ T cells, we used PRIMMUS (Proteomic analysis of Intracellular iMMUnolabelled cell Subsets), a method which utilises FACSorting to conduct proteomic analysis on sorted populations from asynchronous CD8+ T cells, thereby allowing a method to identify the cell cycle regulated proteins in CD8+ T cells without the need for synchronisation.

Our analysis identified 160 proteins which had a strong periodicity rating, indicating they were likely to be cell cycle regulated. The majority of these identities consisted of mitotic proteins such as those involved in kinetochore assembly. In Embryonic Stem (ES) cells, rapid proliferation is facilitated by the inhibition of an E3 ubiquitin ligase known as the anaphase promoting complex cyclosome (APC/C) due to high expression of the pseudo-substrate Emi1. This prevents APC/C from binding with co-activator Cdh1, and thus preventing the degradation of many S phase promoting mechanisms, especially proteins which are involved with origin licensing such as Cyclins E and A, which stay constitutively active. This results in a shortened G1 and G2 phase, with the majority of stem cells found in S phase more than any other cell cycle phase. In CD8+ T cells we noticed that many targets of APC/CCdh1, including DNA replication enzymes and origin licensing components and regulators, showed either very low periodicity or very little fluctuation over the course of cell cycle, indicating a loss of cell cycle regulation. We also observed that Cyclin A was found to fluctuate far less in CD8+ T cells than has been observed in the somatic Nb4 cells, and that Emi1 had a very high level of periodicity. When comparing this dataset with other studies on asynchronous populations for ES and somatic cells, we found a greater overlap of processes shared between CD8+ T cells and ES cells than with somatic cells which suggested CD8+ T cells may share stem-like qualities in their cell cycle regulation. We therefore hypothesised that Emi1 has a role in defining cell cycle rate and by extension, the eventual differentiation fate.

To test this hypothesis, we conducted another proteomic analysis comparing IL-2 treated T effector cells to IL-15 treated memory T cells. Within these two phenotypes, we further divided them into G1 delayed and actively cycling cells. Absolute protein content was generally higher in effector cells. But when examining APC/C substrate abundance relative to total protein content, we found that G1 delayed memory T cells had a higher proportion of APC/C substrates compared with G1 delayed effector T cells. This was not the case with actively cycling cells, which showed similar proportions of APC/C substrates between effector and memory. This could imply that while memory T cells are encouraged to exit cell cycle, they retain much of the machinery to re-enter, maintaining a higher state of cell cycle “readiness” than arrested effector T cells possess. We also noted that Emi1 levels were consistent between memory and effector cells, with the true correlating factor of Emi1 expression being whether the cell was G1 delayed or actively cycling, with actively cycling cells expressing a high quantity of Emi1, while G1 delayed cells expressing very little.

To examine the role of Emi1 on CD8+ T cells more directly, we induced a knockout of Emi1 by Crispr. This knockout of Emi1 lead to T effector cells undergoing “re-replication”, in which the genome is duplicated a second time before the cells enter M phase. The cells that did not experience re-replication, experienced an apparent block to cell cycle re-entry. While re-replication was present in memory T cells, the effect on cell cycle re-entry did not appear. The re-replication is likely to have occurred as a consequence of insufficient APC/CCdh1 inhibition. This results in the degradation of origin licensing regulating proteins like Cyclin A and geminin, resulting in inappropriate origin licensing, which is essential for commencement of DNA replication. However, if the cell becomes deficient in Emi1 before origin licensing occurs, it remains within G1 and becomes unable to progress, a phenomenon observed in Emi1 disruption within transformed cancer cells and stem cells. That we saw no increase in G1 phase in knockout memory T cells could imply that Emi1 was already significantly reduced in memory cells, rendering the knockouts impact on G1 phase bellow detection.

We therefore conclude that CD8+ T cells achieve a rapid proliferation phenotype via stem cell-like regulation of cell cycle, enabling steady re-entry into S-phase mediated by high levels of Emi1, and that IL-15 treated cells shift their cell cycle behaviour, reducing Emi1 and thus reducing their proliferative activity. Because we noticed in our dataset the fluctuation of deubiquitinase Usp1 mimics the pattern of Emi1 fluctuation, we hypothesise that SLEC associated transcription factor Id2 and MPEC associated transcription factor Id3, proteins which Usp1 preserves, may be the link between cell cycle speed and memory phenotype, as Id2 and Usp1 are both APC/CCdh1 substrates, but Id3 is not. We suggest a model then in which Emi1 mediated APC/CCdh1 inhibition promotes both rapid proliferation and survival of Id2, driving cells towards the effector cell fate. While in the absence of Emi1, cells remain in G1 where APC/CCdh1 is active, degrades Id2, allowing Id3 to promote the memory cell fate.