Thread by @BettinaWeigelin

It’s exciting to finally share and discuss our study on additive cytotoxicity, a crowd-based killing mechanism of cytotoxic T cells (CTL) to target cancer!
Highlights in the thread, free access @NatureComms here: nature.com/articles/s4146…
1/17
#immunology #microscopy

Initially, we aimed to study modulators of CTL serial killing capacity. But live-cell imaging of various mouse and human model systems, incl. OT1 CTL targeting OVA-expressing cancer cells surprisingly showed, that CTL are pretty bad killers and most contacts were non-lethal. 2/

We doubted our ability to culture cytotoxic T cells.... but in co-cultures with fibroblast-like cells, the same OT1 CTLs proved to be efficient serial killers. 3/

Did the CTL fail to recognize the cancer cells? Or were the cancer cells resistant?
We introduced different optical reporters into melanoma cells to directly visualize CTL-mediated damage. 4/

We found that CTL frequently induced sublethal damage, from which the melanoma cells (repeatedly) recovered.
The video shows the transient loss of nuclear integrity during CTL attack, indicated by the leakage of NLS-GFP into the cytoplasm. 5/

CTL-mediated killing is typically considered to be a binary yes/no process. But we even found sublethal DNA damage induced by CTL followed by melanoma cell recovery. 6/

Recovery rates after single contacts were high (> 90% for perforin events) and recovery times depended on the type of damage. Perforin pores were repaired within minutes, damage to the nuclear lamina within minutes to hours, and DNA damage took several hours to dissolve. 7/

This led to the question: Does damage accumulate if multiple CTL attack sequentially?
Video footage and analysis of many apoptosis events showing the involvement of multiple CTLs before tumor cell death suggested just that. 8/

Or, were most sublethal CTL attacks irrelevant, and only occasionally, a rare, very strong hit from a single super-killer completed the job? 9/

Statistical analysis by @JohannesTextor showed that the impact of a CTL hit depended on the presence of prior perforin events while simulation of stochastic killing did not resemble the actual data. This indicated that damage from sequential CTL attacks indeed accumulated. 10/

The concept of additive cytotoxicity implies that killing efficacy depends strongly on the local CTL density to achieve a high frequency of serial hits. What does this mean for CTL effector function in the tumor microenvironment, where the CTL : tumor cell ratio is often low? 11/

We used intravital microscopy to monitor CTL density in tumor subregions over time and identified the invasive tumor front as particularly sensitive for high CTL swarming, multi-hit activity and tumor cell eradication. 12/

Time-lapse imaging of individual CTL-tumor cell interactions in the living tumor showed that, similar to the in vitro data, sequential CTL interactions preceded tumor cell death. 13/

Comparison of the frequency of perforin events in the invasive front of the tumor with tumor cells located in the main tumor mass further confirmed that invasive tumor cells accumulate more perforin events per hour. 14/

In summary, we show that CTL induce sublethal damage that is integrated in the target cell over time. Consequently, the efficacy of CTL killing in tumor tissue depends strongly on local CTL density and migratory capacity, which jointly mediate frequent sublethal hits. 15/

The integration of damage and recovery in the tumor cell also implies that CTL killing is a tunable process that can be targeted by immunotherapy. Further, requiring a high local CTL density may act as a filter and limit unintended tissue damage from rare mistargeted CTL. 16/

And just like T cells, scientists need teamwork to succeed – Many thanks to the great teams at @radboudumc, @WSIC_Tuebingen, @MDAndersonNews, the support from @NWONieuws, @dfg_public, @CoE_iFIT & the great microscopes from @LeicaMicro, @LaVision_BioTec, @miltenyibiotec
/TheEnd