A First Somatic Cell Aging Model for Systems Beyond

A Boolean network for proliferation, apoptosis, and senescence

Why we built this model

Aging doesn’t happen in one organelle or one pathway. It emerges from how nutrient sensing, stress responses, DNA damage, and inflammatory signals talk to each other over time. To even start asking serious “what if?” questions (CR, rapamycin, NAD⁺, anti-inflammatories, etc.), we need a generalizable somatic cell model, not tied to a specific tissue or disease.

The model presented here is our first computational prototype of such a cell within Systems Beyond:

• It’s mechanistic enough to include key players (mTOR, AMPK, ROS, DDR, p53, NFκB, etc.).

• It’s simple enough (Boolean logic) to explore whole-network dynamics, attractors, and basins of attraction.

• It can already express three essential fate outcomes: proliferation, apoptosis, and senescence + SASP.

This is Model v1.0: a starting point for systematic refinement, not a final “truth”.

The model structure in a nutshell

The Boolean aging network is a qualitative cell-fate simulator. It does not try to calculate concentrations or reaction rates. Instead, every node is treated as ON/OFF, and the whole cell state evolves step by step until it reaches a stable pattern, called an attractor. In our app, the model explicitly separates nodes into inputs, signaling, metabolic, stress/damage/protection, and cell-fate outputs. The final fate nodes are APOPTOSIS, SENESCENCE, SASP, and PROLIFERATION.

The big idea of the network

Our model asks:

Given a cellular environment — nutrients, growth factors, DNA damage, ER stress, inflammation, hypoxia, matrix contact, and oxidative stress — does the cell settle into proliferation, senescence, apoptosis, SASP, or a protected/adaptive state?

So the model is basically a decision machine for cellular aging.

The layers of the network are composed by the following: 1. Environmental inputs:

NUTRIENTS
GF
DNA_DAMAGE
ER_STRESS
INFLAM_ENV
HYPOXIA
MATRIX
OX_STRESS

This is the “world” the cell lives in.

• NUTRIENTS tells the model whether the cell has energetic/building resources.

• GF means growth factors are present.

• MATRIX means the cell has extracellular matrix contact.

• DNA_DAMAGE, ER_STRESS, OX_STRESS, HYPOXIA, and INFLAM_ENV are stressors.

These inputs decide whether the cell behaves like a healthy proliferating cell or like a damaged aging cell.

2. Growth-factor and adhesion signaling

The first internal decision is whether the cell receives growth permission.

GF → RTKMATRIX → INTEGRIN

If growth factors are present, RTK turns on. If the cell has matrix contact, INTEGRIN turns on. Together, these feed into the classic pro-growth machinery:

RTK / INTEGRIN → PI3K / PDK1 → AKT

This part represents the cell asking:

Do I have external permission to grow, attach, metabolize, and divide?

When this axis is ON, the model activates AKT, which then supports MTOR, BCL2, and eventually PROLIFERATION, as long as stress checkpoints are not active.

3. mTOR, metabolism, and growth capacity

The metabolic layer is the core “growth versus conservation” switch.

When the cell has:

AKT = ONNUTRIENTS = ONAMPK = OFF

then:

MTOR = ONMTORC1 = ON

This means the cell is in a growth-compatible state. MTORC1 then supports:

HIF1AGLYCOLYSISPROLIFERATION

So in our model, MTORC1 is not just a metabolism node. It is one of the key gates that allows proliferation.

But if stress activates AMPK, then AMPK suppresses the mTOR/proliferation axis. In biological terms, this represents a shift from “build and divide” toward “protect, repair, conserve.”

The stress-response system

Our model has a very important stress-damage core:

ROS
DDR
P53
P21
P16_RB
AUTOPHAGY
FOXO
NRF2
SIRTUIN
AMPK

This is the aging-relevant part of the network.

ROS as a central amplifier

ROS turns on when there is oxidative stress, mitochondrial dysfunction, or inflammatory activation:

ROS = OX_STRESS OR poor MITO_FUNC OR inflammatory NFKB activity

Once ROS is active, it feeds into many aging-relevant nodes:

ROS → NFKBROS → LKB1 / AMPKROS → DDRROS → P16_RBROS → APOPTOSISROS → SENESCENCE

So ROS is one of the main “damage-amplifier” nodes in the system.

AMPK, SIRTUIN, FOXO, NRF2, and AUTOPHAGY as protection

The protective branch works like this:

Stress → LKB1 / AMPKAMPK → SIRTUINAMPK / SIRTUIN → FOXOROS → NRF2AMPK / FOXO / NRF2 / SIRTUIN → AUTOPHAGY

This branch represents adaptive stress response.

The key protective output is:

AUTOPHAGY

In our model, autophagy protects the cell because it supports mitochondrial function and reduces damage propagation. It also inhibits the conversion of ROS into DNA damage response when active:

DDR = DNA_DAMAGE OR (ROS AND NOT AUTOPHAGY)

That is a very important design choice: ROS alone does not fully push the cell into damage response if autophagy can handle it.

The cell-fate decision logic

Our model has four final fate outputs:

PROLIFERATION
SENESCENCE
APOPTOSIS
SASP

These are not independent. They compete with each other.

Proliferation

The cell proliferates only if it has strong growth and metabolic permission, and no major checkpoint/fate block:

PROLIFERATION =RTK AND INTEGRIN AND MTORC1 AND GLYCOLYSISAND NOT P21AND NOT P16_RBAND NOT APOPTOSISAND NOT SENESCENCE

Conceptually:

The cell divides only if it has growth factors, matrix adhesion, mTORC1 activity, glycolytic capacity, and no active arrest/death/senescence program.

This is a strict rule. Even if growth signals are present, proliferation is blocked if P21, P16_RB, APOPTOSIS, or SENESCENCE are ON.

Senescence

Senescence depends on damage plus MTORC1:

SENESCENCE =P53 AND (DDR OR ROS) AND MTORC1OR previous SENESCENCE maintained by damage/inflammation

This is one of the most interesting biological choices in our model.

It means senescence is not simply “damage.” It is:

damage + survival/growth metabolism

In other words:

If the cell is damaged but still has mTORC1-driven growth/metabolic activity, it does not simply die. It may enter a senescent arrested state.

This is a good aging-modeling idea because senescent cells are not dead. They are metabolically active, damaged, arrested, and often inflammatory.

Apoptosis

Apoptosis is activated when damage is strong, P53 is active, anti-apoptotic protection is weak, and MTORC2 is not protecting the cell:

APOPTOSIS =P53 AND (DDR OR ROS) AND NOT BCL2 AND NOT MTORC2OR previous APOPTOSIS maintained by P53

So apoptosis is our model’s “irreversible or self-maintaining death decision.”

In plain language:

If the cell detects serious damage and cannot protect itself through AKT/BCL2 or MTORC2, it dies.

SASP

SASP is driven by senescence plus inflammatory signaling:

SASP =SENESCENCE AND NFKBOR previous SASP maintained by SENESCENCE

This means SASP is not just senescence. It requires the senescent state plus inflammatory transcriptional activation through NFKB.

So our model distinguishes:

senescent but quiet

from:

senescent and inflammatory

That is biologically useful because not every senescent-like state has the same inflammatory output.

The central biological story

The core story of our model is this:

A) Healthy growth condition

When the cell has:

NUTRIENTS = ONGF = ONMATRIX = ONstress inputs = OFF

then growth signaling activates:

RTK / INTEGRIN → AKT → MTOR → MTORC1 → GLYCOLYSIS

and because damage checkpoints are OFF:

P53 = OFFP21 = OFFP16_RB = OFFSENESCENCE = OFFAPOPTOSIS = OFF

the cell reaches:

PROLIFERATION = ON

In the app’s default proliferative preset, the simulation indeed uses a low-stress healthy input state with nutrients, growth factors, and matrix contact ON.

B) Aging / chronic damage condition

When the cell has:

DNA_DAMAGE = ONER_STRESS = ONINFLAM_ENV = ONOX_STRESS = ON

then the damage layer activates:

ROS → DDR → P53 → P21

and inflammatory reinforcement appears through:

INFLAM_ENV / ROS / PKC → NFKBNFKB → ROS

This creates a damaging feedback structure:

ROS ↔ NFKBROS → DDR → P53

Depending on the balance between MTORC1, BCL2, MTORC2, and damage, the model can move toward apoptosis or senescence. In our current aging preset, the app defines a chronic-damage condition with nutrients, growth factors, DNA damage, ER stress, inflammatory environment, matrix contact, and oxidative stress ON.

Why our model is good as a complex-systems model

The interesting part is not only the individual pathways. The important thing is that our network has feedback, memory, and attractors.

The simulation keeps a history of states and checks when a state repeats. If it repeats, the model identifies an attractor, either a fixed point or a limit cycle. The app then reports attractor type, attractor length, and steps to attractor.

That makes the model a complex-systems representation of aging because aging is not modeled as one linear pathway. It is modeled as a state-space transition problem:

initial condition + environment → dynamic trajectory → attractor / fate

The cell does not merely “activate pathway X.” It moves through a network of mutually constraining decisions until it stabilizes into a fate.

The most important switches in our network

Switch 1 — Growth versus stress conservation

AKT / MTORC1versusAMPK / SIRTUIN / AUTOPHAGY

This is the growth-conservation axis.

• AKT and MTORC1 support growth.

• AMPK, SIRTUIN, and AUTOPHAGY support stress adaptation and conservation.

Switch 2 — Repair versus damage commitment

AUTOPHAGYversusROS → DDR → P53

If autophagy handles stress, damage does not necessarily propagate. If autophagy is insufficient, ROS activates DDR and P53.

Switch 3 — Senescence versus apoptosis

SENESCENCE: P53 + damage + MTORC1APOPTOSIS: P53 + damage + low BCL2 + low MTORC2

This is one of the strongest conceptual features of the model.

• Damage plus MTORC1 pushes toward senescence.

• Damage plus lack of survival protection pushes toward apoptosis.

Switch 4 — Senescence versus SASP

SASP = SENESCENCE + NFKB

So inflammatory activation determines whether senescence becomes inflammatory.

Final interpretation of our model

Our Boolean aging network represents a somatic cell as a dynamic decision system balancing growth, metabolic sufficiency, stress adaptation, DNA damage response, and cell-fate commitment. Growth factors, nutrients, and matrix contact activate the RTK/Integrin/AKT/mTORC1 axis, allowing glycolysis and proliferation when checkpoints are silent. Stressors such as oxidative stress, ER stress, hypoxia, inflammation, and DNA damage activate ROS, AMPK, DDR, P53, P21, and P16_RB. Protective nodes such as AMPK, SIRTUIN, FOXO, NRF2, and AUTOPHAGY attempt to buffer stress. If damage dominates, the system exits proliferation and settles into apoptosis, senescence, or inflammatory senescence/SASP depending on whether survival/growth metabolism and inflammatory signaling remain active.

That is the essence of our model: aging as a shift in the attractor landscape from a proliferative attractor toward damage-stabilized attractors such as senescence, SASP, or apoptosis.