Module 4 · Stress Response

The Unfolded Protein Response

When misfolded load exceeds chaperone capacity, the ER activates the unfolded protein response (UPR): three transmembrane sensors — IRE1, PERK, and ATF6 — that couple luminal protein burden to transcriptional and translational output. The UPR is adaptive at low intensity and pro-apoptotic at high, making it both a homeostatic mechanism and a decision-making device.

1. Sensing: BiP Titration

Under basal conditions, BiP binds the luminal domains of IRE1, PERK, and ATF6, keeping them monomeric and inactive. Rising misfolded load competes for BiP; BiP dissociates; the sensors oligomerise and activate. This is the competition model (Bertolotti 2000). A supplementary direct-binding model(Credle 2005, Gardner 2011) posits that IRE1 and PERK’s luminal domains directly bind unfolded peptides in their MHC-class-I-like groove; both models contribute in vivo.

2. The IRE1 Arm

IRE1α (and the tissue-restricted IRE1β) is the most ancient branch — conserved from yeast. Activation proceeds through dimerisation, trans-autophosphorylation, and activation of the C-terminal RNase domain. The signature output is unconventional cytoplasmic splicing of the XBP1 mRNA: IRE1 excises a 26-nt intron, shifting the reading frame to produce the active transcription factor XBP1s, which drives chaperone, ERAD, lipid-synthesis, and ER-expansion gene expression.

Under prolonged stress, IRE1’s RNase becomes promiscuous — cleaving a broad set of ER-localised mRNAs in a process called regulated IRE1-dependent decay (RIDD)— reducing the load on an already overwhelmed folding machinery but also eroding stress-adaptive transcripts themselves. This is one way IRE1 tips from adaptive to pro-apoptotic.

3. The PERK Arm

PERK activation results in phosphorylation of eIF2α Ser51, a global translation-attenuation switch shared with the integrated stress response (ISR; also triggered by GCN2, HRI, PKR). Phosphorylated eIF2α sequesters the eIF2B guanine-nucleotide-exchange factor, collapsing ternary-complex formation and arresting 80% of translation within minutes.

Paradoxically, some transcripts — most importantly ATF4 — are preferentially translated under eIF2α-P conditions via upstream ORF regulation. ATF4 drives amino-acid biosynthesis, antioxidant response, and autophagy genes. Sustained ATF4 induces CHOP / DDIT3, the pro-apoptotic arm of the UPR.

ISRIB (integrated stress response inhibitor, Sidrauski 2013) binds eIF2B and partially rescues translation under eIF2α-P conditions. Mouse studies show cognitive benefit in neurodegenerative models; human trials are ongoing. PERK inhibitors (GSK2606414) showed pancreatic toxicity but second-generation (PERK-low-dose) agents are in development.

4. The ATF6 Arm

ATF6 is a type-II membrane protein. Under stress, it is transported to the Golgi and sequentially cleaved by S1P and S2P proteases (the same enzymes that cleave SREBP sterol regulators). The released cytosolic domain, ATF6(N), translocates to the nucleus and induces chaperones, ERAD components, and XBP1 itself (thereby primed for IRE1-mediated splicing). ATF6 is the most transcriptional of the three arms.

Simulation: UPR Dynamics Across Stress Levels

Python

script.py52 lines

import numpy as np, matplotlib
matplotlib.use('Agg')
from scipy.integrate import odeint
import matplotlib.pyplot as plt

# Three-arm UPR simplified: IRE1, PERK, ATF6 each responding to BiP availability
# dU/dt = k_syn - k_fold * U * BiP_free/(BiP_free+K)
# BiP_free is titrated by unfolded load
# Output arms: adaptive vs apoptotic

def upr(y, t, k_syn):
    U, XBP1s, ATF4, CHOP = y
    BiP_total = 3.0
    BiP_free = BiP_total / (1 + U/1.0)   # hyperbolic titration
    fold = 5.0 * U * BiP_free/(BiP_free + 0.5)
    # IRE1 -> XBP1s (hormesis)
    dXBP1s = 2 * U**3/(U**3 + 1.5**3) - 0.3 * XBP1s
    # PERK -> ATF4
    dATF4 = 1.5 * U**2/(U**2 + 1.5**2) - 0.25 * ATF4
    # CHOP downstream of ATF4 (integration over time)
    dCHOP = 0.2 * ATF4**2 - 0.05 * CHOP
    # U dynamics
    dU = k_syn - fold - 0.02*U
    return [dU, dXBP1s, dATF4, dCHOP]

t = np.linspace(0, 100, 1000)
# Three scenarios
adapt = odeint(upr, [0.5, 0.1, 0.1, 0.1], t, args=(2.0,))
chronic = odeint(upr, [0.5, 0.1, 0.1, 0.1], t, args=(5.0,))
lethal = odeint(upr, [0.5, 0.1, 0.1, 0.1], t, args=(10.0,))

fig, (ax1, ax2, ax3) = plt.subplots(1, 3, figsize=(16, 5), facecolor='#0a0a1a')
for ax in (ax1, ax2, ax3):
    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)

for ax, sol, title in [(ax1, adapt, 'Adaptive (k_syn=2)'),
                        (ax2, chronic, 'Chronic (k_syn=5)'),
                        (ax3, lethal, 'Lethal (k_syn=10)')]:
    ax.plot(t, sol[:,0], color='#fbbf24', lw=2, label='Unfolded U')
    ax.plot(t, sol[:,1], color='#34d399', lw=2, label='XBP1s (adaptive)')
    ax.plot(t, sol[:,2], color='#60a5fa', lw=2, label='ATF4')
    ax.plot(t, sol[:,3], color='#f87171', lw=2, label='CHOP (pro-death)')
    ax.set_xlabel('Time (arb)', color='#cbd5e1')
    ax.set_title(title, color='#6ee7b7', fontweight='bold')
    ax.legend(facecolor='#1e293b', edgecolor='#334155', labelcolor='#cbd5e1', fontsize=8)

plt.tight_layout()
plt.savefig('output.png', dpi=120, bbox_inches='tight', facecolor='#0a0a1a')
print('UPR three arms: IRE1 (adaptive), PERK (brake + ATF4), ATF6 (transcriptional boost)')
print('Sustained PERK/CHOP tips the cell into apoptotic attractor')

Click Run to execute the Python code

Code will be executed with Python 3 on the server

5. Adaptive vs Apoptotic: The Bifurcation

The three arms activate near-simultaneously but attenuate with different kinetics: ATF6 decays fastest, IRE1 next, PERK persists longest. This kinetic staggering means that if stress resolves quickly, the cell restores protein homeostasis through the adaptive pathways. If stress is chronic, PERK/ATF4/CHOP signalling dominates:

CHOP upregulates the BH3-only proteins Bim and PUMA;
CHOP upregulates ERO1α, increasing H₂O₂ production;
CHOP upregulates GADD34 (PPP1R15A), dephosphorylating eIF2α and restoring translation of misfolded proteins — a counter-intuitive twist that accelerates cell death;
Mitochondrial permeabilisation via Bax/Bak, caspase-9 activation, apoptosis.

The bifurcation from adaptation to apoptosis is a topic of continuous current research. It is the molecular basis of many cell deaths in neurodegeneration (Module 6) and a target for pharmacological rescue.

6. Specialised UPR States

The UPR is not monolithic. Different stimuli produce different arm combinations: lipid bilayer stress activates IRE1 preferentially (Halbleib 2017); low-intensity stress tunes toward pro-survival; high-grade stress toward apoptosis. Plasma cells reprogram their UPR during antibody-secretion differentiation, activating XBP1 splicing constitutively to expand the secretory capacity. In cancer, tumours in hypoxic microenvironments are UPR-addicted — IRE1 and PERK inhibitors have emerged as an anticancer strategy (clinical trials in multiple myeloma, breast). In neurons and pancreatic β-cells, pharmacological UPR modulation is under investigation for protection in Parkinson’s, ALS, type-II diabetes, and retinitis pigmentosa.

← Module 3 Module 5: Lipids & Ca²⁺ →