Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • ADVERTISEMENT FEATURE Advertiser retains sole responsibility for the content of this article

Sounding out cell to cell interactions for better cancer immunotherapy

T cells can be engineered to better target and attack cancer cells, and researchers are interested in technology to help pick the most sensitive receptors for the jobCredit: Shutterstock

Medical researcher, Steven Albelda, is looking for a Goldilocks receptor – one with just the right amount of sensitivity to its target. His research group at the University of Pennsylvania is developing a potential immunotherapy for lung and other thoracic cancers, based on chimeric antigen receptor (CAR) T cells.

Adoptive T cell therapies, such as those using CAR T cells, bolster the cancer-detecting abilities of a patient’s immune system. A typical CAR consists of antigen-binding and activation domains, often combined with one or two costimulatory domains. This allows it to both attach to the target antigen and activate the T cell.

CARs can either target tumour cells directly or disrupt the tumour growth process through another pathway. Albelda’s group has designed a receptor that recognizes a protein abundant on the surface of fibroblasts, which are a tumour’s ‘scaffold’ cells. By binding to a molecule called fibroblast activation protein (FAP), the CAR T cells can destroy the tumour’s framework. However, if the receptor binds too strongly to the FAP, it’s also more likely to interact with non-tumour cells that express much lower levels of the protein, which leads to higher toxicity and unwanted side effects.1

For treatments like these to work, researchers need to develop a receptor that easily locks on to its target, but is not so sensitive that it causes side effects. It’s a delicate balancing act, and testing the sensitivity of a new receptor currently relies heavily on animal studies. “If we had a better way to get an idea of how efficient a particular CAR is, and how well it binds its target, that would be useful,” says Albelda.

Adopting the best T cells

CAR T cells are gaining in popularity, and can be modified in many innovative ways. John Maher, Chief Scientific Officer of cell therapy company Leucid Bio, based in London, has created parallel CAR (pCAR) T cells, which express two chimeric receptors: one is a CAR with a single costimulatory domain; the second receptor has a different costimulatory domain. “We've made quite a number of these molecules,” says Maher, “and one question we’re pursuing is how strongly each receptor binds to its target.” That could help Leucid scientists decide which pCAR T cells to further develop.

The other main type of adoptive therapy involves engineering T cells to express tumour-specific T cell receptors (TCRs). Unlike CARs, TCRs can also recognize intracellular proteins, as they bind to peptides presented on the cell surface by major histocompatibility complex (MHC) proteins. Matthias Leisegang, an immunologist at Charité – Universitätsmedizin Berlin, studies TCR T cells. He says that another advantage TCRs have is that they are extremely sensitive2. “One MHC molecule loaded with the right peptide is sufficient to trigger a T cell.”

However, not all TCRs are equal. Different TCRs against the same antigen can have varying sensitivity to their targets. Leisegang currently has to wait weeks for the results of animal studies to find out which TCR works best3. But if there were validated, early indicators of efficacy, he would be able to reduce the number of animal experiments and quickly home in on the most suitable TCRs.

Strength in numbers

One possible predictor is avidity, a measure of the binding strength between the T cell and its target (see 'Affinity versus avidity’). But measuring avidity is not a straightforward procedure, and very few labs include it in their processes.

A new technology, z-Movi, could change that. “The way that people usually measure cell avidity is very complex and elaborate. It's typically between one effector cell and one target, so the success rate is low,” says Rogier Reijmers, Principal Scientist Immuno-Oncology at technology developer LUMICKS, in Amsterdam. “Our device allows you to analyse 100 to 300 cells in one run.”

The z-Movi device is about the size of a small benchtop centrifuge. It can measure the binding strength between cells via a technique called acoustic force spectroscopy.

z-Movi’s key component is a chip with a small cell chamber and a piezoelectric element.

Target cells, such as tumour cells, are cultured in the chip’s chamber, and incubated with T cells or other effector cells.

The chip is inserted into the device, which generates sound waves using the piezo element. The waves act as an acoustic force that pulls the target and effector cells apart4,5. At the lowest levels, the acoustic force separates the cells with the weakest connections; increasing the force progressively detaches cells with higher avidity6. This creates a unique binding curve (see ‘How z-Movi works’).

Reijmers’ team has been working with research groups to test how well z-Movi works with different cell types, including NK cells, TCR T cells and CAR T cells.

How z-Movi works (A) T cells are added to the flow channel of the chip and adhere to a monolayer of target cells. The piezo element on top of the chamber generates an acoustic force (green waves) to lift the T cells from the monolayer. (B) The acoustic force linearly increases over time. (C) Observed through the microscope, fluorescent T cells are initially seen scattered uniformly, but are pulled into nodes as the acoustic force exceeds their binding strength. (D) Two distinct avidity curves distinguish pools of low (a-c) and high (d-f) avidity T cells.

In Leisegang’s lab, z-Movi proved it could distinguish between TCR T cells with high or low sensitivity to their targets. Leisegang was able to find the TCR-sensitivity predictor his team was looking for. “Our goal was to see to what extent the TCR has an impact on the binding between the T cells and the tumour cells,” Leisegang explains. “The z-Movi device gave us important information about the binding strength.”

At UPenn, Albelda’s group used z-Movi to measure the binding strength of their CAR T cells. They compared their newly designed CAR with F19, an existing CAR against the same FAP target, and the z-Movi measurements clearly distinguished the two.

“The F19 CARs stuck quickly and took a lot of force to get off,” says Albelda, “whereas our new CAR required much less force to knock it off.” To ensure that the receptor would be less likely to cause side effects, they also tried the experiment with different target cells, expressing much lower amounts of FAP. The F19 cells still stuck to these targets easily, but the new CAR T cells did not — exactly what they were hoping to find.

Maher has also used z-Movi to test his pCAR T cells. “We found that we’re able to correlate the performance of CARs and pCARs with the binding strength of T cells to target-expressing cells.”

What’s more, it only takes a few hours from seeding the first cells to analysing the output of a z-Movi experiment. Speed is particularly appealing to companies such as Leucid Bio. “A lot of people are making CARs, so we need to be able to process things as quickly as we can,” says Maher. “With a technology like this, we can whittle down candidates rapidly.”

Click here for more information on z-Movi from LUMICKS


  1. Majzner, R.G. & Mackall, C.L. Nature Medicine 25: 1341–1355 (2019).

    Article  Google Scholar 

  2. Huang, J., et al. Immunity 39: 846–857 (2013).

    Article  PubMed  Google Scholar 

  3. Obenaus, M. et al. Nat Biotechnol 33: 402–407 (2015).

    Article  PubMed  Google Scholar 

  4. Sitters, G. et al. Nature Methods 12: 47–50 (2015).

    Article  PubMed  Google Scholar 

  5. Kamsma, D., et al. Cell Reports 24: 3008–3016 (2018).

    Google Scholar 

  6. Bunse, L., Green, E.W. & Platten, M. in Methods in Enzymology (eds Galluzzi, L. & Rudqvist, N-P) Ch. 20 (Academic Press, 2019).

    Google Scholar 

Download references


Quick links