Module 8 · The Translation

Organoids & Regenerative Medicine

In 2009, Hans Clevers’ lab in Utrecht reported that a single Lgr5⁺intestinal stem cell could, in Matrigel with three defined factors, build a miniature gut in a dish — a hollow, polarised, villus-forming structure they named an organoid. The paper opened a decade of experimental biology: organoids from every major organ, patient-derived “tumour organoids” for precision oncology, brain organoids modelling early cortical development, and the first clinical trials using iPSC-derived cells manufactured essentially as organoid products. This module surveys the field.

1. Sato & Clevers (2009): The Intestinal Organoid

Toshiro Sato, working with Hans Clevers, embedded a single sorted Lgr5-GFP⁺cell from mouse intestine in Matrigel with R-spondin, Noggin, and EGF. The cell divided, organised into a cyst, polarised apico-basally, and built differentiated crypt-villus structures containing enterocytes, goblet cells, enteroendocrine cells, Paneth cells, and fresh Lgr5⁺ stem cells. The organoid was clonal and self-organising — no mesenchymal support, no patterning signals beyond the three soluble factors.

This was the first demonstration that an adult stem cell, taken out of its niche and supplied with the right diffusible cues, could reconstruct the architecture of its tissue. Every subsequent organoid protocol has used the same principle: thesignalling centre replaces the niche.

2. The Organoid Zoo

Within ~10 years, organoid protocols were established for essentially every organ:

Gastric (Barker 2010): Lgr5⁺stomach stem cells, pyloric vs corpus architecture recapitulated.
Liver (Huch 2013, 2015): expansion of biliary stem cells, hepatocyte organoids.
Pancreatic: acinar and ductal organoids, including patient-derived PDAC models.
Kidney (Takasato 2015): nephron segments self-organise from iPSCs.
Lung (Barkauskas 2013; Nikoli&cacute; 2017): bronchiolar and alveolar organoids; used for cystic fibrosis drug screening.
Prostate, breast, salivary, mammary: all routine by 2020.
Retinal (Eiraku 2011, Sasai lab): self-organising optic cups from ESCs; photoreceptor-bearing retinal organoids for AMD cell-replacement research.
Cerebral (Lancaster 2013): iPSC-derived self-organising cortex-like structures with ventricle-like cavities and crude cortical layering. Brain organoids have revolutionised the study of human neurodevelopment.
Cardiac (Drakhlis 2021; Lewis-Israeli 2021): self-organising heart organoids with chamber-like cavities and beating regions.

3. Brain Organoids & the Zika Problem

Lancaster’s 2013 Nature paper showed that iPSC aggregates, embedded in Matrigel and allowed to self-organise under neural induction, form structures containing multiple brain-region identities: cortex-like ventricles, choroid plexus, hippocampus, retina. A rough fetal-stage brain in a dish.

Within three years, cerebral organoids were used in two high-impact applications: (i) modelling Zika-virus-induced microcephaly (Cugola 2016, Qian 2016), which established Zika causation and identified AXL as a entry receptor; (ii) modelling genetic microcephaly (Lancaster 2013; Li 2017 MCPH1), revealing the cellular mechanism (premature neural-progenitor depletion).

Cerebral organoids remain cortically primitive — they lack vasculature, mature myelination, proper lamination, and long-distance connectivity — but they are the first system in which human cortical development can be directly observed and manipulated. “Assembloids” (Birey 2017, Pasca lab): fusion of region-specific organoids to study inter-region circuit formation.

4. Patient-Derived Tumour Organoids

Tumour organoids grown from patient biopsies preserve the mutations, phenotypes, and drug sensitivities of the parental tumour. Organoid biobanks now exist for colorectal (van de Wetering 2015), breast (Sachs 2018), pancreatic, gastric, prostate, and brain tumours. Applications:

Precision oncology: grow the patient’s tumour organoid, screen 10–100 drugs in parallel, guide clinical decisions. CF patient-derived organoids (Dekkers 2013) guide modulator selection; cystic-fibrosis trials have run with organoid-based inclusion.
Mechanistic cancer biology: test causation of individual mutations (CRISPR + organoid); study early stages of tumorigenesis that are inaccessible in patients.
Drug development: replacement or complement to 2D cell lines, which are often genetically drifted from their source tumours. Organoid panels better predict clinical drug response.

Simulation: Cell-Therapy Clinical Pipeline

Python

script.py32 lines

import numpy as np, matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt

# Current pipeline of stem-cell therapies (approximate, 2025)
indications = ['RPE for AMD', 'Dopaminergic (PD)', 'Islet (T1D)', 'Cardiomyocytes (MI)',
               'Photoreceptors', 'Oligodendrocyte (SCI)', 'T cells (CAR-T)', 'HSC gene therapy']
# Phase 1=preclinical, 2=Ph1, 3=Ph2, 4=Ph3, 5=approved
phase = [4, 3, 3, 3, 2, 3, 5, 5]
source = ['iPSC/ESC', 'iPSC', 'iPSC/ESC', 'iPSC', 'iPSC', 'ESC', 'Autologous T',
          'Autologous HSC']

fig, ax = plt.subplots(figsize=(12, 6), facecolor='#0a0a1a')
ax.set_facecolor('#111827'); ax.tick_params(colors='#cbd5e1')
for s in ax.spines.values(): s.set_color('#334155')
ax.grid(True, color='#334155', alpha=0.3, axis='x')

colors = ['#2dd4bf' if p>=5 else '#fbbf24' if p>=4 else '#fb923c' if p>=3 else '#f87171' for p in phase]
y = np.arange(len(indications))
ax.barh(y, phase, color=colors, edgecolor='#fda4af')
ax.set_yticks(y)
ax.set_yticklabels([f'{i}\n({s})' for i, s in zip(indications, source)], fontsize=9, color='#cbd5e1')
ax.set_xlabel('Clinical stage', color='#cbd5e1')
ax.set_xticks([1,2,3,4,5])
ax.set_xticklabels(['Preclin', 'Ph 1', 'Ph 2', 'Ph 3', 'Approved'], color='#cbd5e1')
ax.set_title('Stem-cell-derived therapies: clinical pipeline (approx 2025)',
             color='#fda4af', fontweight='bold')

plt.tight_layout()
plt.savefig('output.png', dpi=120, bbox_inches='tight', facecolor='#0a0a1a')
print('HSC gene therapies: first wave approved (Casgevy 2023, Lyfgenia 2023)')
print('iPSC-derived cell therapies: RPE, cardiac, dopaminergic leading Phase 2/3')

Click Run to execute the Python code

Code will be executed with Python 3 on the server

5. iPSC-Derived Cell Therapies in the Clinic

Leading clinical programmes (2025):

Retinal pigment epithelium (AMD)

Takahashi’s Kobe trial (2014) was the first-in-human iPSC therapy — a 65-year-old woman’s own iPSCs differentiated to RPE, transplanted as a subretinal sheet for wet age-related macular degeneration. The treated eye stabilised; no tumorigenesis observed. Subsequent allogeneic HLA-matched iPSC RPE trials (Sumitomo, 2017; Lineage Cell Therapeutics OpRegen, 2024) in Phase 1/2 for dry AMD show encouraging safety and preliminary efficacy.

Dopaminergic neurons (Parkinson’s)

Japan’s CiRA (Takahashi Jun, Kyoto) began Phase 1/2 in 2018; preliminary data (2024) showed graft survival, dopamine restoration on PET, and symptomatic improvement in several patients. BlueRock Therapeutics (bemdaneprocel) and Aspen Neuroscience running parallel US trials.

Pancreatic islets (T1D)

Vertex’s VX-880 (ESC-derived allogeneic islets, encapsulated/immunosuppressed): first patient became insulin-independent in 2021 (phase 1/2). Expansion to unencapsulated version now in Phase 2. Parallel efforts at Sigilon, Sernova.

Cardiomyocyte patches (post-MI)

Sawa (Osaka): iPSC-derived cardiomyocyte sheets transplanted onto ischemic myocardium. First patient 2020. Murry lab (Seattle) demonstrated engraftment and coupling in primates; human trials starting.

T-cell products

CAR-T cells — autologous T cells engineered with a chimeric antigen receptor — are not stem-cell therapies in the strict sense but occupy adjacent regulatory and manufacturing ground. Kymriah (tisa-cel, 2017), Yescarta, Carvykti are approved. iPSC-derived off-the-shelf NK and T products (Fate Therapeutics, Century) are in trials.

6. Manufacturing, Safety, and Scalability

iPSC-derived therapies face shared translational challenges: tumorigenicity (residual undifferentiated pluripotent cells form teratomas — requires rigorous QC), genetic stability (iPSC lines accumulate chromosomal abnormalities; cGMP lines undergo karyotyping and hotspot sequencing), immune matching (autologous avoids rejection but is too expensive; allogeneic HLA-matched or hypoimmune-engineered lines are the pragmatic frontier), and manufacturing cost (currently >$100 000 per patient; must fall 10× for wider use). The field is industrialising at speed; iPSC master-cell-bank platforms at Fujifilm, Lonza, and others are now reaching the scale where a single clinical lot serves thousands of patients.

← Module 7 Module 9: Ethics & Clinical Trials →