Lecture 7 - High-throughput, Optimized Platforms for 3D Bioprinting of Kidney Organoids

current kidney disease

850 million people globally

large number of patients

are not diagnosed until end stages

difficulty with chronic kidney disease

high heterogeneity

organoids

recapitulate the key structure and functional features of human actual kidneys as a simplified version

2d cell culture

often failed to recapitulate the complex 3d architecture and cell-to-cell interactions

animal models

interspecies differences

spheroids

simple homogeneous cells and lack spatial organization seen in actual tissues

organoids

miniaturized tissue-like structures derived from the patient’s stem cells

biomaterial

lack of tissue specific bioactive materials for kidney organoids

poorly defined

2d stem cell culture

tumor mimetic

3d cell culture

variable

organoid assembly

platform

requires extensive optimization of printing parameters

matrigel advantages

biocompatibility, ability to support cell adhesion, proliferation, and differentiation

matrigel disadvantages

undefined competition, batch to batch variability

existing organoid culture methods are

labor intesive and low throughput

how organoids are generated

all the stem cells are detached and distributed into small microtube → centrifuge and aggregate

bioink from

ecm of porcine kidneys

pig kidney chosen

most similar to human organ size

decellularization

removing cells from the tissue or organ, removing only the extracellular membranes which can support the new cell growth

perfusion

use the organ’s blood vessels to deliver the resorbent agents detergent and help preserve 3d structures

emergent and agitation

e chop all the tissues and then soak into a solution and gently stir until they remove all the blood and cells

kidney derived ecm

can promote the maturation and vascularization of human kidney organs

ecm biomaterials used alone

lack the sufficient tunability of mechanical properties for 3D bioprinting

photocrosslinkable bioink through

metacrylation of kidney descent

freeze the fresh kidney tissue at -9-

easier slicing

pig cells removed

using an optimized detergent solution such as SDS or Tritonex

convert pig ecm

perform solubilization step resulting in a dECM biomaterial rich in collagen

due to its high collagen content

DCM solution undergoes gelation as the temperature rises, like for example

thermophoresinking characteristic

coat the surface to culture the organism on top of that

two critical drawbacks

poor printability and difficult to control the mechanical property

incorporate the metacrylate synthesis

photocrosslinkable — when the specific light wavelength shoots the material, it becomes gelation

check wether all the pig cells are completely removed

H&E staining

photocortical material

rheology test

SEM analysis showed that the lower biomaterial concentration led

to a larger pore sizes

because the metacrylic anhydride is chemical

so it's pretty toxic to the cells

So I dialysis all the DCM materials

to remove all the toxic things

embedded the human embryonic kidney cells

into the DCM metacrylate material, and then core sinking, and then culture those scaffolds, and then see whether the cells are happy or not.

day 30, cell proliferation become plateau after one month

indicating that all clusters of cells have merged together.

multilayer 3D scaffold bar printing

via DLP-based SLI bar printer

tested the developed biomaterial

using a piston-driven extrusion bioprinter

thinning behavior of bioink

when I perform the printing, the material needs to go through the needle and it should be liquid from a gel state

stepper motors drives the piston

to dispense the bio-ink

to control the droplet volume

also optimized the relationship between the stepper motor's rotational frequency and also speed and the resulting droplet volume

confirmed that the GFP-TAG33 fibroblast cells encapsulated

in both 2x2 lattice structure and also the droplet shape

machine learning optimization

development of a bioprinter capable of high-throughput image data collection along with the advanced platform that optimizes printing parameters using regression-based machine learning

pre-printing stage

optimizing the printing parameters before doing actual printing begins

feature extraction in machine learning

manual

feature extraction in deep learning

automatic

a regression-based supervised machine learning

at the same time, the deep learning to compare which algorithm shows the best result

for machine learning models

applied decision tree, random forest, and polynomial regression

deep learning

multi-layer perception and long short-term memory

use these five algorithms to

compare their functionality and performance to predict the droplet volume

movable stage into 3D bioprinter

allowed images and videos to be collected efficiently

this system could print the cellular droplets as small as

0.1 microliter with high precision

algorithm with three different image processing steps

to measure the droplet volume using the finer extracted images

after calculating the drop-down volume

I used the resulting numerical data and then the corresponding parameters as input for machine learning algorithms

my first step was to identify

optimal hyperparameters for training

performed the hyperparameter optimization

before actual training

deep learning models such as MLP and LSTM show that relatively strong performance

with low mean square error compared to the decision tree random forest of whole NML regression

machine learning models also showed good predictive ability

with MSE values before 1

the purple dots, the data points are distributed linearly around the diagonal line

indicating their good model performance

although I tested droplet volume ranging from 0 to 30 microliter

our study focuses more on the lower volume range, particularly 0 to 5 microliter, because normally organoids are very small

competition time

compare the training time and the testing time across five different models

the five algorithms that are embedded

give you the predicted volume, how much the kidney organoids, the volume will be, and with the prediction accuracy as well

validated the system

confirming cell proliferation within the bio-printed 3D cellular droplets

kidney organoids is very time-consuming and labor-intensive

centrifuge all the small microtubes and transfer each cell aggregates to the transfer membrane one by one so normally the technician can make four kidney organoids

kidney organoids were generally using very, very high cell concentration

they use 500k and then minimum 200k

develop a cost-effective bioprinter

easily accessible to most of the research lab

reduce the labor-intensive nature of conventional organized production

and also to enable the efficient large-scale production

reduce up until

8,000 nephroprogenital cells to make the kidney organoids

maximize the number of organoids produced per membrane

differentiate the intermediate mesoderm from day zero to seven

improve the repeatability between organoids

explore whether it is possible to create mature organoids using a very small number of nephroprogenital cells

prevent the drying out after finishing printing

medium was added directly below the transfer insert

kidney contains about

one million nephrons

The bowl shape we call the

cospuscle

closer tubule is we call

proximal tubules

porosites labeled

with the MPHS1

proximal tubules stained with

LTL

CD31 staining

confirmed that the presence of the complex blood vessels

ECAD

osteotubules

normally the patient takes 12 hours

this biopsy took just a couple of minutes

through our human patient-led FTC kidney organoids

achieved a significant reduction in labor, time, and time, and enabled production of more organoids in a just single trans-folding ring

developing the multi-organ spinal chip

induced activity in the kidney organ side and then investigate how the cytokines secret from these injured kidney organoids affect to the engineered heart tissue

heart dECM based hydrogels like what i've done with the kidney dECM hydrogel

encapsulate the cardiac organoids into the heart dECM hydrogels