Project Objectives

• Shortlist the candidates. Select commercially available, animal-free gels against seven fixed criteria. The gels must set at room temperature without UV light, be transparent enough for microscopy and plate readers, and come ready to support a wide range of cell types.
• Test them by hand first. Grow four widely used human cell types in each shortlisted gel: skin fibroblasts, and breast, lung, and colon cancer cells, all in culture medium free of animal serum. Track how well the cells survive and grow over a week, and image how they interact with the gel. Then scale the protocols down to 384-well plates, ready for automation.
• Teach the robots to handle gels. Program NEXUS’s liquid-handling robots to dispense each gel accurately, without clogging, forming bubbles, or setting too early. The protocols must be reliable enough to process at least 20 plates in a single run.
• Run a pilot screen. Test 960 compounds, a mix of approved drugs and drugs that have been through clinical trials, on all four cell types. Then repeat the screen independently to check that the results hold.
• Try it on a harder case. With Sandro Nuciforo’s group at the University of Basel, grow organoids from patients’ liver tumours in the best-performing gel, and screen the same 960 compounds.
• Share it. Publish the manual and automated protocols, a troubleshooting guide, and the raw data openly. The team will also run a webinar and offer hands-on support to other labs. The method may be added to a continuing-education course on the 3Rs at FHNW, where Eva Maria Riegler lectures.

3Rs Impact

• Removes a hidden animal product. Matrigel, the most widely used gel for growing cells in 3D, is extracted from a type of tumour grown in mice. Every batch requires animals, and manufacturers don’t disclose how many.
• Targets the stage where volumes are largest. Screening uses gel at scale. An average 3D screen at NEXUS runs on about 40 plates and uses 300–350 ml of Matrigel. That figure doesn’t include the gel customers use to develop their assays beforehand.
• Makes the animal-free route the easy option. By NEXUS’s count, more than 90% of the assays researchers bring to the facility rely on animal-derived materials. Researchers rarely switch mid-project, because redeveloping a method is slow and expensive. With validated protocols ready in advance, choosing the animal-free option no longer means extra work for the customer.
• Spreads beyond the facility. NEXUS serves universities, hospitals, and start-ups across Switzerland. Promising compounds from a screen are followed up in the customer’s own lab, so a gel that works in the screen can carry through into the experiments that follow.
• When the benefit arrives. The project itself uses no animals and spares none directly. The benefit comes once the method is in routine use. NEXUS expects to apply it at least once a year in the five years after publication, and estimates that this could spare hundreds of mice over the coming years.
• More consistent results. Chemically defined gels vary less from batch to batch than Matrigel, whose exact composition is unknown. This makes screens more reproducible.

Background

Drug discovery often starts with screening. Automated systems test thousands of compounds on cells to find the few with a useful effect, such as killing cancer cells. For decades, these screens used cells grown as flat layers on plastic. Increasingly, researchers grow cells in three dimensions instead, because 3D cultures behave more like tissue in the body. At NEXUS, 3D screens have outnumbered flat ones by two to one over the past three years.

Some cells form 3D clusters on their own, but many need a gel to grow in: a scaffold that holds them in place and supports their growth. The standard choice is Matrigel, and collagen extracted from rat tails is another common option. Both work well, but they vary from batch to batch, they are animal rather than human in origin, and producing them uses animals.

Synthetic, chemically defined gels are now available, and labs use them successfully by hand. Automating them is harder. Gels are thick, they set depending on temperature and time, and cells must be mixed evenly through them before tiny volumes are dispensed into hundreds of wells. A gel that behaves well on the bench can clog a dispenser or set halfway through a run.

NEXUS Personalized Health is ETH Zurich’s core facility for screening and lab automation. Its screening group is led by Eva Maria Riegler, who is the lead applicant, and co-applicant Stephanie Huber is one of its specialists. The team has done this before. With Sandro Nuciforo and industry partners, it built a platform for screening organoids from patients’ liver tumours, using a cellulose-based, animal-free gel, and scaled it down to 1,536-well plates. The manufacturer then closed its biomedical business, and the gel was taken off the market. Nuciforo’s group had also found that the gel didn’t suit every cell type, because some cells need to attach to the material around them in order to grow. The team sees both problems as reasons to establish several alternatives within its service.

The team expects some setbacks. It rates the risk that some gels won’t cope with automation, or that cells won’t grow well without animal-derived gel and serum, as moderate to high. The project is therefore built in stages: only gels that work by hand move on to automation, and only those that work on the robots move on to screening. Matrigel is used throughout as a comparison, to confirm that the cells are healthy.