Full transcript
0:03- First I'd like to thank C&EN
0:05for giving me this honor
0:07and since I'm the last one in the session, I thought I could go over time a little bit
0:13and try hard to make it a full session,
0:15as Lauren has always wanted.
0:17So today I'm going to talk about some of the work
0:19that we have done in the past years,
0:22engineering living systems.
0:24And a little bit of my background:
0:25So I did my undergraduate in biomedical engineering
0:28in Southeast University in China.
0:32And since then, I have not changed my major.
0:35And I came to the States for my master at Wash U in 2008,
0:42exactly 10 years ago,
0:44after which I actually moved with my PhD advisor,
0:47professor Younan Xia to Georgia Tech,
0:50again in the Department of Biomedical Engineering,
0:53jointly with Emory University.
0:56And about five years ago, I arrived at Boston
0:58to do my postdoc as Lauren just mentioned under
1:03supervision of Professor Ali Khademhosseini
1:06which is where I stayed as faculty,
1:11which is in Division of Engineering Medicine
1:13at Brigham and Women's Hospital
1:15which is part of Harvard Med School and basically
1:18essentially a biomedical engineering division hospital.
1:22So, today I'm going to give you a like,
1:25briefly talk about my trajectory in the past 10 years
1:28in engineering function of human tissues and tissue models
1:31using a combination of technologies and strategies,
1:35ranging from biomaterials and chemistry of materials
1:38to biofabrication and biomedical devices
1:42and all the way to bioanalysis.
1:45So the story starts from a lack or a severe lack
1:49of donor organs for transplantation.
1:52And we know that many people die
1:55from the lack of donor organs, not even mentioning
1:58the high cost associated with the processes.
2:03So to this end tissue engineering,
2:06the concept of tissue engineering, was raised
2:09about three decades ago, which essentially allowed us
2:13to generate functional human tissues outside human body,
2:18which we can use to transplant back into the human system
2:21for regeneration purpose.
2:23And tissue engineering is essentially a
2:26multidisciplinary technology,
2:30which combines different components
2:33for generating functional tissues and organs
2:36we can eventually use for transplantation.
2:38And one of the most important component
2:40is the scaffolds.
2:41So basically it's the like structural
2:45material that gives you the ability to host the cells
2:50in three dimensions and allow them to function.
2:54And among the different scaffold systems,
2:57pore scaffolds is especially important because
3:00they have interconnected pore that allows
3:03cells to communicate and also allow the diffusion of nutrients
3:05and the bioactive molecules.
3:07However, one of the issues with
3:09conventional fabrication strategies
3:13for scaffolds lies in the inability to reproduce
3:17the pore structures, right.
3:18So if you look at scaffolds fabricated using,
3:22like, across different batches,
3:24they have four different structures, right?
3:26Making it very difficult to engineer tissues
3:29that are having uniform qualities.
3:31So this basically
3:35prompted my research in my graduate school,
3:40which was to generate these kind
3:42of inverse opal scaffolds using a specialized strategy,
3:45giving us the chance to generate fabricate scaffolds
3:49of uniform, like with uniform pores.
3:52And to do this we basically used the microfluidic device
3:56to produce uniform microparticles, which we can use
3:59to pack together to form the lattice, right.
4:03So once you have the lattice you can infiltrate
4:05the secondary pore material into the space
4:08between the microspheres and by removing the microspheres,
4:11eventually, you can form this inverse opal scaffolds
4:14with very uniform pore sizes.
4:17So, as I mentioned, then you can start to use
4:21this inverse opal scaffold with very controllable properties
4:27for engineering functional tissues that are uniform across
4:29different properties and by controlling, for example,
4:32the microspheres or the sizes of microspheres
4:35that we make for templating purpose,
4:37we can make scaffolds off of different pore sizes,
4:40but in each scaffold,
4:41the pore size is actually very uniform.
4:43We can show also the, for example,
4:46annealing temperature of this lattice
4:50and make sure that you can get scaffolds
4:53having different window sizes
4:55which are basic interconnecting
4:58like interconnecting cavities between the different pores.
5:03And we can also modulate the chemistry or the material
5:07properties, to be able to generate the surface sizes
5:10like scaffolds with different surface sizes,
5:13to also modulate tissue behaviors.
5:15So now we can start to engineer
5:19the different tissue types
5:20for regeneration purpose of the study
5:23of the biological behaviors of cells
5:25and tissues within this kind of scaffolds, right?
5:27So these range from, for example vascularization,
5:29cell migration, cell body production scaffolds,
5:32and also generating patches that we can use
5:35for wound healing purposes directly on the patients.
5:40And take vascularization as an example, right.
5:43So these pore sizes, the scaffolds with different pore sizes
5:47allows us to basically modulate how the vasculature,
5:50or the host tissue, is responding to the scaffolds.
5:53For example, the surface pore size actually
5:55determines how the infiltration efficiency of the vessels
6:00from outside of scaffolds into the scaffolds,
6:03and then the pore size that I mentioned,
6:05actually determines the time of progression
6:08of the vessel infiltrating vessels between
6:11before they actually encountered a barrier.
6:14And lastly, the window size, which is the size of
6:18the interconnection between different pores actually
6:22determines the ease of penetration of vessels
6:24into the secondary structures
6:25or interior structure of the scaffolds.
6:28And eventually by this way we can control how the tissue
6:31is responding to the scaffolds that
6:33we implant into the human system.
6:35And histology results show that basically we can
6:39like using scaffolds with different pore sizes
6:41we can have different vascularization patterns
6:45within the scaffolds, giving us the chance to precisely
6:49regenerate the tissue defects in the human system.
6:53However, the histology results that I just showed
6:57have been collected using conventional strategies
7:03based on for example, the cryotome or sectioning,
7:06which is a labor-intensive process and also
7:10is time-consuming, right.
7:13So that's why a secondary part
7:15of the PhD work was trying to
7:19develop different imaging technologies to be able
7:23to allow us to characterize the engineered tissues
7:26and the qualities in a non-invasive
7:28and in a continuous manner.
7:31And in particular we, in collaboration
7:33with Dr. Lihong Wang who was in Wash U
7:36and who is now in Caltech actually, right now.
7:38We specifically focused on this photoacoustic imaging
7:42which is a combination of light and ultrasound imaging,
7:46which gives us not only the penetration depths
7:48but also the resolution that we need
7:50for characterizing tissues and as such we can
7:52now monitor these engineered tissues and their qualities
7:56in a non-invasive manner continuous in the human body,
7:59ranging from high resolution imaging of the single cells,
8:02to cells on the scaffolds to scaffolds themselves
8:05and how the scaffolds are interacting
8:07with tissues in the human body.
8:09And again, taking the scaffolds
8:11with different pore sizes example,
8:13we can instead of having the histology slides
8:17that I showed like just before,
8:20we can also use this photoacoustic imaging
8:23to noninvasively monitor how
8:25the vascularization is going on
8:26in the human system, actually in the animal
8:29in this case, for scaffolds with different pore sizes
8:33that were implanted.
8:34And we did see a differential response
8:37of the of the vascularization in scaffolds
8:38with smaller pore sizes and larger pore sizes.
8:43Alright, so when I came to Boston I learned another
8:47challenge of, like, facing the whole
8:50I guess healthcare field,
8:53which is drug development,
8:54so with all the pharma companies from Boston.
8:57And I learned that the drug development process
9:02is highly costly and taking extremely long time
9:07and without much of the high efficiency.
9:11So the primary reason for that is conventional models
9:15that people use for developing the drug molecules
9:18are either based on planar static cell cultures
9:21or animal models, both of which don't really
9:26recapitulate the human system, right.
9:28Because as we know the human systems are dynamic,
9:31three-dimensional, and these cannot be recapitulated
9:34by the 2-D cell cultures and animals,
9:37they are physiologically relevant,
9:38but they are having genetic differences.
9:40So many of the drugs that are tested [to be] efficient
9:44or safe in animals, they don't really
9:46translate into the human body.
9:48So this is why I then use my tissue engineering background
9:52that I accumulated in the PhD studies,
9:56to basically generate this now in vitro
10:01models of human tissues on these little microfluidic chip devices
10:06for better drug development and also personalized medicine,
10:11when I came to Boston.
10:13And these are essentially miniaturized versions
10:15of the human tissues and organs that you can make
10:19on the little devices, which allow us
10:22to reproduce the dynamic environment
10:25of the human system.
10:26And most importantly we can connect this different organ
10:29on the chip devices together in a way that
10:31the different organs connect in the human system.
10:33So we can also start to study this,
10:35their interactions outside human body for a more accurate
10:40drug testing and development and eventually,
10:44I think as Roxanne [Kieltyka] also mentioned,
10:46we can potentially combine with, for example,
10:48like stem cells, so we can eventually start
10:51to also do process medicine as well using these devices.
10:55So again, as I mentioned, one
10:57of the most important components,
11:00is the bioreactor systems,
11:02which are those microfluidic units
11:03that are hosting the organoids that we generate.
11:06And this is one of the simplest form of bioreactor
11:09which has only a simple chamber where we can put
11:12the micro tissues in for dynamic flow culture
11:15and we can make variations as well.
11:16For example, can add a layer in the middle
11:18and put endotheliel cells on top.
11:20So, now we can form very nice layer of endothelium
11:24on the membrane so we can now study also the,
11:26for example, transport of the drugs through
11:28the endothelial barrier and how they interact
11:31with the surrounding tissues.
11:32And as we know the human systems are dynamic where,
11:36for example, in your vessel system,
11:38they're not really static with the heart beat.
11:41So the vessels are constantly dilating and constricting.
11:45So we can also build these mechanical cues onto these
11:48little microfluidic devices so we can model
11:51the contractions of the vessel wall
11:54and study their effects on the cells
11:58or vascular cells seated on top.
12:01And we can also form unconventional microfluidic devices.
12:06For example, the vessels in the human body are circular.
12:09So, we can also make these circular devices
12:13which are then like endothelialized
12:17with these vascular cells.
12:18So we can use them to connect different organoids
12:21to really mimic the function of the vessels
12:23connecting different organs in the human system.
12:26And one of the recent developments in the lab is again,
12:31so, the previous ones are based on the
12:35elastomer materials which are having issues
12:38when testing drug molecules, because
12:42they tend to absorb the small molecules,
12:45especially those that are hydrophobic.
12:48So, we have been devoting our efforts on generating
12:51also thermoplastic materials for microfluidic chip fabrication
12:56so we can avoid those kind of issues.
12:58And we were able to generate all those
13:00different types of microfluidic units for controlling purposes
13:03and build them into the bioreactors for that purpose.
13:08So now we can also start to
13:12multiply the complexity of the microchip devices
13:15and everything is automatically controlled
13:20using our controllers.
13:22So I guess the last part I'm going to really
13:26briefly talk about is the biofabrication.
13:28So now we have the bioreactors.
13:29Next problem is how do you actually generate
13:32three-dimensional tissues that are functional
13:34within these microfluidic devices for drug testing?
13:37So as we know the human systems are complex.
13:41So here shows some exemplary organs, which are usually
13:46comprised of little tissue building units at the micro scale,
13:50which are, again, very complex.
13:52So the challenges here are how do we reproduce the complexity
13:58of human tissue microarchitecture
14:01and how do we mimic the interactions
14:03between the different cell types.
14:05So in this way we actually used bio-printing to generate
14:09our micro-scale tissue models in the chips.
14:13So one strategy that we did earlier was the
14:15multi-material extrusion bio-printing
14:18in which we combined multi bio-inks into a single nozzle
14:21and by controlling how the nozzle moves around
14:23and by ink-jet deposition we can start to generate
14:26human tissue-like, organ-like structures,
14:29at miniaturized scale using this bio-printing strategy.
14:33And recently we have also extended this multi-material
14:36printing into another strategy called,
14:39stereolithography printing, which is a layer-by-layer
14:44printing system and we incorporate the microfluidic device,
14:47to allow the system to be able to also produce
14:49multiple materials that are mimicking our microtissues
14:53directly on the chips.
14:55And we have been focusing on vascular bio-printing
14:58as well, because vessels are essential
15:00and one of the most important components of all tissue types.
15:04So with sacrificial bio-printing we can make
15:06these little channels which we can then endothelialize
15:09to model the human vessel system.
15:10And we can then start to model,
15:13for example, the vascular disorders,
15:14we can put whole human blood into the channels
15:17and let it clot in there.
15:18So now we can form thrombus in the channel smaller
15:21than the human thrombus, in the human vessels, right.
15:23So if we do histology, we can see that these models are
15:26the model that we have engineered was very similar to
15:29the real human thrombus giving us the capacity
15:32to model, to study the pathology
15:36and also develop the treatments.
15:39And we can also start to model micro-vasculature
15:43on a scale of a few tens of micrometers using
15:46other sacrificial bio-printing strategies.
15:49And not only we have vascular structure that are hollow,
15:52we have other structures that are hollow as well
15:53in human body, for example, the mammary ducts, right.
15:56So in this case we actually used the sacrificially printed
15:59hollow channels to model breast cancer
16:01and how the cancer cells were proliferating
16:04in the ducts.
16:05And when they became confluent, how they started to migrate
16:08into the surrounding matrix, modeling this, like, reproducing this
16:12ductal carcinoma in situ
16:14to invasive ductal carcinoma process.
16:18And we can also directly print hollow fibers
16:22using the coaxial multi-layered nozzle,
16:24allowing us to directly print a multi-layered vessel bed,
16:29that is profusable.
16:30We can also complex this with multi-material bio-printing
16:34for producing the transitional structures
16:36that we have in the human systemb
16:37because the vessels, we know that
16:39they're not really uniform, right?
16:42A lot of vessels are coming from multi-layer
16:44to single layer and going back to the multi-layer.
16:47So we can also do this perfectly.
16:50We can also print for example bioelectronic devices
16:53using bio-printing that are not only conductive
16:55but also are biocompatible so we can now start to model,
17:00for example, bioelectrically conductive devices.
17:04So I think we have also basically been able to develop all
17:10the sensors for monitoring to show, like, organoids
17:13on the chips as I showed in the beginning
17:16with the tissue monitoring.
17:19So this allows us to basically look
17:22at how the organoids are behaving or interacting
17:24with drug molecules in the system.
17:26Alright, so I think with that with that slide,
17:28I'm going to thank all the funding agencies
17:33and collaborators across the world
17:35and also the great lab members across several generations,
17:38and of course my mentors,
17:40PhD advisor Professor Younan Xia,
17:43postdoctoral advisor Professor Ali Khademhosseini,
17:46and all the collaborators.
17:48Thank you very much.