Dr. James Bryers is a Professor in the Department of Bioengineering at the University of Washington. A central focus of his research program is on the use of precision biomaterials to drive tissue regeneration, cell reprogramming, and resolve inflammation. We sat down with Dr. Bryers to discuss the history and future of biomaterials, based on his own research.
Can you tell us about the history of biomaterials and tissue engineering?
Historically, biomaterials heal poorly. This includes those used in solid hip replacements, contact lenses, dental materials, etc. All of them, regardless of whether they’re made from soft tissue or solid material, never really recapitulate what they’re replacing. They have limitations in terms of life span as well. If you get a hip replacement at forty, you’re going to need to get another one when you’re about seventy. There are also limitations in terms of disease and type of injury that biomaterials can replace. Devices can’t treat Parkinson’s disease, or replace something like a liver. It’s these limitations that caused the field of tissue engineering to emerge around twenty years ago, with the hope of being able to regenerate things like livers or heart valves instead of using mechanical replacements.
The initial naive approach was to make a scaffold that would support cells in the hopes that they would become your new liver or whatever. About twenty years of research followed, during which porous materials were created with little molecules placed on them intended to make target cells happy and grow. A lot of the research was done ex vivo, where the cells were cultivated within scaffolds, and the cell-seeded scaffold then placed inside a patient. A variety of different approaches were tried. Some failed, and some worked but didn’t perform as well as others. Researchers discovered that it’s not as simple as just putting cells in a three-dimensional construct and expecting them to become a complete organ. This led to more research. And one of the big questions was the effect of porosity. Does the pore size of the cell scaffold actually have anything to do with healing?
To test this, I would imagine that you’d need to create scaffolds with pores of a consistent size. But my understanding is that this isn’t a trivial thing to do. How did you go about addressing that issue?
You’re right. About 15 years ago, most porous polymer scaffolds could be made in only a handful of different ways. The problem with these techniques was that the pore size distribution was really variable. If you were trying to obtain 100 µm pores, you’d wind up with pores that were anywhere from 50 – 150 µm.
What was needed was a technique to fabricate a scaffold with pores that were all the same size, such that one could really evaluate the effect of pore size on healing. Dr. Andrew Marshall, a PhD student of my colleague at UW, Prof. Buddy Ratner, came up with a technique to make scaffolds where every pore is uniformly controlled [narrow pore size distribution] throughout the scaffold and the pore interconnects are also uniform in size, with both parameters being adjustable. They created scaffolds with 20, 40, 80, and 100 µm pores, and tested the effect on healing. It turned out that everything healed when pores were all around 40 µm. Blood vessels appeared and tissue was regenerated. Remarkably, this occurred no matter what polymer was used, although there are subtle differences in immune response depending on the polymer selected.
Do you know what it is about the 40 µm pores that led to healing?
When we were first looking into answering this question, we hypothesized that the macrophages that invade the scaffold soon after implantation promote an inflammatory response that somehow leads to healing. And when we followed the cells over a period of time, we did see large numbers of macrophages throughout the scaffolds. But as time progressed, tissue and vessel cells appeared, which was strange considering that the scaffold was full of macrophages the previous day. We wondered whether the macrophages were differentiating again, or whether they weren’t even macrophages to begin with (even though they displayed classical macrophage cell markers). Perhaps they were macrophage progenitors that were still on their way to becoming a macrophage. Since this early work, researchers have found all kinds of stem cell-like markers on macrophage cells, so it’s entirely possible that the “macrophages” that we observed initially could be capable of differentiating into a different cell type, if coaxed.
To test this, we made a double transgenic mice strain expressing fluorescent green protein in all myeloid cells (macrophages, dendritic cells, neutrophils) and fluorescent red protein in all non-myeloid cells. These reporter genes are chromosomally based, so the cells will continue to fluoresce regardless of what type of cell they become. We then used these mice to test the scaffolds. For the first one to four weeks we observed green and red cells cohabiting within the 40 µm pore scaffolds. In scaffolds with larger or smaller sized pores, we initially observed the same pattern of red and green cells early after implantation, but this would be followed by a spacial migration of red cells out of the scaffold, while the remaining green myeloid cells would grow larger and fuse, eventually becoming foreign body giant cells.
Eventually you traced the healing properties of the scaffold to exosomes, correct?
Yes! Serendipitously, since the red and green fluorescence markers in the transgenic mice are membrane sequestered stains, any extracellular vesicles (e.g., exosomes, microvesicles) that arise from a parent cell retains that fluorescence. So, the first year that we were doing these experiments, we kept seeing all this red and green “pixie dust” (~100–200 nm in size) flying around under microscopic examinations of the scaffolds. I’d hired a postdoc at the time who’d previously worked with exosomes, and she was the one that identified that we were indeed seeing exosome generation. We then were able to isolate the red or green exosomes and carry out outer membrane analysis (membrane proteins) and vesicle content analysis (RNAseq, proteomics).
In addition to looking at the RNA in these exosomes, did you examine the proteome?
Oh yeah. We did everything! That’s why we’re broke. These analyses are expensive!
We performed RNA-Seq and looked at the proteome of all the 40 µm- versus 100 µm-pore size scaffold generated exosomes, separated into red (non-myeloid) or green (myeloid) sub-groups. When we compared miRNA found in the exosomes generated by 40 µm versus 100 µm scaffolds, we found significantly more miRNAs in the former. The composition of the miRNAs and proteomes were dramatically different in both groups (red vs. green) of exosomes as well.
When we did a PANTHER data search, we found that a significant number of 40 µm-generated miRNAs were involved in angiogenesis and neovascularization. In fact, we can take exosomes generated from the scaffolds, harvest them, and when we use them in a very simple angiogenesis tube or scratch assay we observe vessel tube formation.
What’s the link between pore size, macrophages, and exosomes?
We think the small 40 µm pores promote the initial invading cells to cohabitate (keep in mind that cells are only about 20 µm in diameter). The cell population in a 40 µm pore scaffold demonstrate a very unique combination of healing macrophage (called M2s) and very specific regulatory T cells. Cell populations in the 100 µm scaffolds do not resemble this pattern, being exclusively pro-inflammatory macrophage (M1s). We think this exclusive population distribution is what’s causing the formation of these specific exosomes. I suspect it’s mechanical, caused by cell membrane receptor sensing of their surroundings.
If the exosomes are the actual agent of regeneration, could they be used as a cell-free therapy to treat damaged tissue?
That’s what we’re trying to test now. When we identified the 25 most abundant miRNAs from the RNA-Seq of the exosomes, we PANTHER searched them, and discovered that about five to ten of them had something to do with neovascularization or angiogenesis. We have currently chosen five of them to create synthetically, will insert them into artificial liposomes, and repeat our in vitro angiogenesis assays. Keep in mind that the exosomes also contain proteins within them, which may lead to a difference in angiogenesis using the liposomes versus native exosomes.
One of my students, Thomas Hady, is currently working on cultivating these exosomes from 40 µm scaffolds, harvesting them, and using them to coat the pores of a 100 µm pore size scaffold, which normally does not heal, and will see if there is any healing.
If it does heal at the 100 µm pore size, is there any advantage to using scaffolds of that size versus other sizes?
The process of creating scaffolds of uniform specific pore size is quite tedious. If you don’t care about pore size variability, however, scaffolds can be made cheaply and fast, but these are more susceptible to poor healing performance. So, it would be great if we could quickly make scaffolds with a distribution of pore sizes, coat the pore surfaces these exosomes or miRNAs, and have them heal. There are also certain situations, like spinal cord injuries or lung perfusion ischemia, where you don’t necessarily need to regenerate tissue or need a scaffold, but you want to decrease chronic inflammation. In these cases, you would potentially just need a suspension of exosomes.
Thank you for taking the time to discuss your research with us, Dr. Bryers!