As for many PhD students, my first year was tough. The focus was to spin artificial spider silk which worked for a while but would stop working for no apparent reason after long hours in the lab. The precious protein that I had expressed and purified for weeks to spin the artificial silk would become useless within a few hours. When I examined the protein, I realised that it looked like a gel and not a liquid anymore. How did that happen? I knew from the literature that others have made hydrogels from spider silk proteins but elaborate protocols needed to be followed for that to happen. It didn’t happen just like that. Being frustrated and curious, we decided to follow up on it.
Our first guess was that the protein got warm during the spinning process, perhaps the warmth of my hands or the pump for extruding the protein in the spinning bath caused gelation. We put some the protein in the 37°C incubator and sure enough, after 30 minutes, it had gelled. Again, comparing that to other protocols, this just seemed too easy. Next we were curious to find out what is causing gelation.
Spider silk proteins, also called spidroins, are generally composed of three parts:
- NT: an N-terminal non-repetitive domain
- Rep: a large repeat region
- CT: a C-terminal non-repetitive region
The three parts have different functions for silk formation. NT is highly soluble and is responsible to keep the spidroins in solution before the silk is spun. The spider keeps a liquid protein stock in their glands that they use for spinning. The repetitive region is aggregation-prone and is responsible for the outstanding mechanical properties of spider silk, i.e., its high toughness, a unique combination of high extensibility and strength. Due to its aggregation prone nature, we use 2 repeat regions in our artificial set-up instead of up to 100 as in the spidroins. CT, aka the trigger, is believed to trigger silk formation by forming amyloid-like structures in response to a decreased pH and shear-forces which helps the repetitive region to unfold.
Fig. 1: Schematic representation of NT2RepCT. Adapted from Arndt et al., 2022.
Considering the natural function of the three different domains, we all put our bets on either CT or 2Rep being responsible for gelation. Both the repeat region and CT are known to go from soluble to amyloid-like structures during the spinning process, making them good candidates for being responsible for hydrogel formation. Also, other scientists had similar thoughts in the past and formed hydrogels from spider silk proteins using the repeat region or CT. NT on the other hand, being known for its great stability, was used to develop a solubility tag for recombinant protein production. Hydrogel formation likely requires rather large conformational changes which we know Rep and CT could do but nobody has ever observed that for NT.
From the way I’m writing, you can probably guess that we were in for a big surprise because it was actually NT that was able to form hydrogels. At the same concentration as NT2RepCT, NT only needed to be incubated for 10 minutes at 37°C. That is fast! Plus, both NT2RepCT and NT hydrogels are transparent which is great for applications involving microscopy. As a side note, CT also formed hydrogels, but gelation was difficult to control.
Fig. 2: Gelation of different constructs.
Now we really wanted to know what was happening to NT. We performed FTIR- and NMR- spectroscopy on NT before and after gelation and saw that this super soluble protein is forming beta-sheets. TEM showed that there are fibrils and they’re even binding to ThT. Checking a lot of hallmarks for amyloid-like structures.
Fig. 3: Spider silk protein hydrogels are made-up of fibrils. Adapted from Arndt et al. 2022.
NT has been studied excessively throughout the years. It’s been known as this protein that is super easy to handle, it expresses a lot recombinantly, it doesn’t aggregate and is in general just very stable. A co-worker once forgot a diluted sample on his bench for a few weeks (!) and when he checked, it was still fine in solution and the structure had not changed. So how can we explain this amyloid-like behavior of NT? NTs’ high methionine content is unique and makes its native fold rather fluid. Salt, low pH and methionines to leucine mutations stabilize NTs’ fold and we found that all these can prevent gelation. Thus, NT’s fluid fold is needed for gelation.
Fig. 4: NT fused to GFP results in functionalized hydrogels. Adapted from Arndt et al.
Lastly, we show that hydrogels can also be functionalized. We fused green fluorescent protein (GFP) and an enzyme called purine nucleoside phosphorylase (PNP) to NT and found that they can be expressed at high yields, form hydrogels, and GFP and PNP remained highly active (>70% and 65% for NT-GFP and NT-PNP, respectively).
What had been started as something annoying (not being able to spin my fibers due to gelation of the protein) has developed into a super fun project with lots of surprises.