top of page

CD39 is expressed on functional effector and tissue resident memory CD8+ T cells

Isaacs, JF et al. (BioRxiv) DOI: 10.1101/2024.03.15.585252v1

CD39 is expressed on functional effector and tissue resident memory CD8+ T cells


  • CD8+ T cells

  • Memory T cells

  • Ecto-ATPase CD38

Main Findings

The preprint by Isaacs et al. investigates the role of the ecto-ATPase CD38 in determining the formation of CD8+ memory T cells. Previously, it has been well-known that CD38 can have immunomodulatory role by aiding in the degradation of extracellular ATP (which causes an inflammatory signal to T cells and other cells) to adenosine, in collaboration with another ecto-enzyme called CD73. CD39 has been well researched in the context of T cell exhaustion and as a marker for tumour specific CD8+ T cells. However, thus far little has been known around the role of CD38 in driving memory T cell formation.

The preprint aims to address this knowledge gap. First the authors, identify that CD39 is increased (compared to naïve CD8+ T cells) upon primary immune challenge. The authors confirm a known fact that antigen-specific CD8+ T cells downregulate CD73 at the same time. The authors find that with increasing time post-primary infection, there is a small but increasing pool of CD39+ CD73+ double expressing T cell subsets. Generally, CD39 seemed in their data to label predominantly short-lived effector cells (SLECs), while CD73 was more dominant on memory precursor cells (MPECs) based on KLRG1+ CD127- and KLRG1- CD127+, respectively.

In an interesting experiment, the authors block ex vivo CD39 using a CD39 inhibitor (namely POM-1). After priming, the cells were transferred into a mouse, which was infected on the next day. Mice receiving primed cells in the presence of the inhibitor showed an increased presence of memory T cells in the peripheral organs >30 days post infection. The authors further note that in particular tissue resident memory (TRM) CD8+ T cell populations (CD69+ CD103+) have high expression of both CD73+ and CD39+. Since CD39 is associated with T cell exhaustion, the authors ex vivo restimulated TRMs, but did not find any functional impairment. Lastly, the authors demonstrate that also bystander CD8+ T cells (in the context of a cancer mouse model, these cells were virus specific) can express high levels of CD39+ if focussed on tissue residency markers (i.e. high levels of CD69 and CD103). This opens up the possibility that CD39+ CD8+ T cells in tumours may not all be tumour specific but have a contribution of bystander specific T cells. While limited in sample size, the authors saw corroborating trends in human settings.

Overall, the study allows for a glimpse into the biology of CD39 in CD8+ T cell memory. It may be a useful starter for further more in depth analysis of the role of CD39 in these cells in different conditions.


While the study provides an interesting primer, the study may be supported by a variety of additional experiments and settings to really test the importance of CD39 for memory T cell formation. It would strengthen the authors narrative if the authors may consider additional experiment to support their claims.

  • It would be helpful to identify whether different combination of CD39 or CD73 expressing subsets (e.g. at day 10) could be adoptive transferred into different mice to see how the expression may pre-determine their memory fate.

  • It feels a little bit like a missed opportunity that the role of CD39 has not been assessed as a side-by-side comparison between an acute and chronic infection (e.g. using the LCMV model) and to specifically look at the MPECs and their fate.

  • Naturally it may be helpful if a genetic KO system for studying the role of CD39 could be used to determine its impact on CD8+ memory formation. If there are certain technical issue with this (e.g. hyperinflammatory state) then it would be good if the authors would clearly discuss this early on in their paper.

  • Currently, the most functional readouts on CD39+ memory CD8+ T cells is based on an ex vivo system. It would be good to do this rechallenge in vivo. Especially since tissue resident memory cells may actually see more ATP in the extracellular space compared to other memory cells (e.g. when the skin ruptures due to a cut). It is unclear whether different ATP concentrations were used during the restimulation and whether this could modulate their effects.

  • The authors mention in their introduction that the CD39-CD73 system can impede on the TCR signal that is received. However, the authors don’t investigate this further in their manuscript, readouts such as Nur77 expression across memory formation or CD73+ or CD39+ subpopulations feel warranted.

  • Lastly, the authors may double-check some of their statistics. In some cases (e.g. Figure 3, OT-I cells in different organs – and the corresponding supplementary), it appears that the data may not be normally distributed, yet it seems student t-tests were applied.


The study sheds new insights into the role of CD39 in normal CD8+ memory T cell formation. While showing some interesting results, it may mostly be a primer for more in-depth investigations for future studies. A main suggestion to take away from this paper seems to be that the interpretation of CD39 as a marker for exclusively tumour-specific CD8 T cells should be carefully monitored.


Reviewed by Felix Richter as part of a cross-institutional journal club between the Vanderbilt University Medical Center(VUMC), the Max-Delbrück Center Berlin, the Medical University of Vienna and other life science institutes in Vienna.

The author declares no conflict of interests in relation to their involvement in the review.

bottom of page