Estimator Reading · Tier-3 · Draft

Logic Mass in the Living Cell

m^L

What it costs to make a specialized cell into something else, and what that cost is made of

Logic mass measures how expensive it is to reconfigure a sovereign structure. The cell is the one place this site treats as a sovereign candidate, and biologists have spent twenty years measuring what it takes to turn a specialized cell into a stem cell. Each measurement moves in the direction the kernel predicts, if these laboratory observables are valid estimators of the kernel's quantities. The clearest match is stored history. Two findings go beyond the kernel. Reprogramming shows up as a random waiting time, closely tied to cell division, and the experiments do not show that the waiting is itself the cost. And much of the difficulty sits in a few named restraints, each of which can be taken away, which suggests the difficulty has a shape: a cell may be cheap to move in one direction and costly in another.

What the kernel says logic mass is

The definition is in the kernel at § 8.4 (LOGIC-MASS). Quoted as written:

Logic Mass is the resistance of a sovereign recursive structure to admissible reconfiguration under recurcline pressure. Logic mass quantifies how much transaction cost is required to produce equivalent structural change in an attractor; high logic mass means change is expensive, low logic mass means change is cheap.

It rises with recursive depth, with feedback interlock density (“more closed loops → more cross-constraint → more resistance”), with constraint accumulation (“more invariants → less freedom”), and with history, because “accumulated α-trace contributes to logic mass.”

It exists only in one place: “m^L(X, t) is defined only for sovereign attractors during their bootstrapping interval,” and it “is undefined for: non-sovereign configurations, attractlets, pre-ignition states, and post-β-loss states.”

And it is never read off directly:

Metrics estimate m^L; metrics do not define logic mass. Operational metrics for m^L may include interlock density measurements, basin-depth estimators, T₃/T₄ historical cost integrals, and α-trace richness measures. None of these is logic mass; each is a substrate-specific estimator.

Five rules this page follows

Estimators only. The page reports measurements of cost, interlock, basin depth and stored history. It never gives “the logic mass of a cell.”

The cell stays a candidate. The Living Cell finds that a free-living cell meets all four sovereignty conditions. It keeps the label “candidate” because one question is open: whether a cell that divides ends, giving two new attractors, or continues in one of its daughters. This page carries the same label. The question matters more here than anywhere else, because reprogramming takes many divisions. If each division ends one attractor and starts others, no single attractor goes through the whole change, and the measurements below would describe a lineage of cells rather than one cell's logic mass.

Cost, never force. The kernel lists “causal-resistance attribution” as a misuse: logic mass “accounts for cost; it does not actively resist.” This page says what reconfiguring a cell costs. It never says logic mass stops a cell from changing.

No ranking across cells. Logic masses of different attractors “are not directly comparable without explicit normalization.” A young donor's cell and an old donor's cell are different attractors. The page can say one population costs more to reprogram than another. It cannot rank their logic masses.

Admissible reconfiguration is T₄. The kernel's reconfiguration fee (§ 9.5) covers “restructuring internal topology, feedback routing, or attractor configuration while preserving identity continuity.” It is told apart from rupture by one test: “T₄ preserves the four SOV conditions throughout; T₅ violates at least one of them.”

When reprogramming counts as reconfiguration

When a skin cell is turned into a stem cell, the cell keeps running the whole time. It keeps its membrane, keeps making its own parts and keeps paying its own running costs. What changes is which stable pattern of gene activity it holds. If the cell keeps all four sovereignty conditions throughout the change, the change is a T₄ event, the kernel says the attractor after it “is the same sovereign attractor with restructured topology,” and these measurements become candidate estimators of logic mass. The experiments show that cells stay alive and end up converted. They do not show the four conditions holding at every step, and the cell's sovereignty is itself unresolved, so every reading below carries that condition.

Division raises a second condition. Reprogramming takes weeks and many rounds of cell division, and Hanna and colleagues found the number of divisions to be “a key parameter.” A T₄ change has to happen within one attractor. If a dividing cell continues as one of its daughters, the cell that ends up reprogrammed is the same attractor that started, and the reading stands. If each division ends the mother and starts two new attractors, the change is spread across a chain of different attractors, none of which goes through all of it, and the waiting time describes a lineage. The framework has not settled which, so every reading below also depends on that answer.

Not every cell makes it through. Soufi and colleagues note, citing earlier work, that cell death occurs in the first 48 hours of reprogramming. A cell that dies loses its boundary, and that is rupture (T₅). The colonies counted in these experiments are the cells that survived, so the readings apply only to cells that stayed intact through the change.

The push comes from outside. The experimenter adds four transcription factors, Oct4, Sox2, Klf4 and c-Myc (together “OSKM”), and the cells live on supplied serum. That does not disqualify the reading. T₄ asks only that the sovereignty conditions hold throughout; it does not ask where the push came from. The sovereignty test for the boundary is a question of capability: what counts is whether the cell “retains the recursion to remake” its boundary, “not what X happens to import.” And the kernel's rule is to name the structure before classifying anything. Name the cell as the structure, and the added factors and the serum sit outside it as supply and disturbance.

The cost of reconfiguring a cell

It succeeds rarely. In the first human experiments, Takahashi and colleagues reported that “from 5 × 10⁴ fibroblasts, we observed ∼10 hES cell-like colonies”: about 10 in 50,000, or roughly 0.02% (our arithmetic). Six separate experiments gave 5 to 11 colonies each. The colonies appeared around day 25 and were picked at day 30 (Takahashi et al. 2007). A later group measured 0.05% under similar conditions (Soufi et al. 2012).

But almost every cell gets there eventually. Hanna and colleagues kept the factors switched on and waited. Reprogramming turned out to be “a continuous stochastic process where almost all mouse donor cells eventually give rise to iPS cells.” By 18 weeks, more than 92% of their wells had produced reprogrammed cells, and the waiting times fit a gamma distribution (Hanna et al. 2009).

The waiting is tied to cell division, but not only to it. Speeding up division, by blocking the p53/p21 pathway or adding Lin28, sped up reprogramming in proportion, and the authors conclude that “the number of cell divisions is a key parameter.” One factor, Nanog, sped it up without changing the division rate, so division count is not the only thing that sets the pace.

What this gives the page: reprogramming difficulty shows up in practice as a random waiting time, often closely tied to the number of cell divisions. The experiments do not show that the waiting is the cost. A rare transition with a small chance at each division produces a long wait without anything being paid down, much as a lock with a billion combinations takes a long time to open by chance with no work stored in the failed tries. So the waiting time is an observable of reconfiguration difficulty, and whether it measures transaction cost is open. It does fit the kernel's rule that logic mass accounts for cost and does not forbid change: almost every cell converts in the end.

A caution from the original paper. Takahashi and colleagues offered three explanations for the low rate. The first was that “the origin of iPS cells may be undifferentiated stem or progenitor cells coexisting in fibroblast culture.” If so, the rate would only measure how rare those cells are. Hanna and colleagues' result answers this: almost all the donor cells reprogrammed in the end, so the low rate reflects waiting time across the whole population.

Named restraints: one brake can carry most of the difficulty

Remove Mbd3. Mbd3 is part of a repressor complex called Mbd3/NuRD. The same factors that switch on the stem-cell network also recruit this complex, which “potently restrains the reactivation” of the genes they target. Depleting Mbd3 turned a slow, rare process into “deterministic and synchronized iPS cell reprogramming (near 100% efficiency within seven days from mouse and human cells)” (Rais et al. 2013).

Remove Ink4/Arf. The Ink4/Arf locus holds genes that stop cell division. “Genetic inhibition of the Ink4/Arf locus has a profound positive effect on the efficiency of iPS cell generation, increasing both the kinetics of reprogramming and the number of emerging iPS cell colonies” (Li et al. 2009).

What this gives the page: much of the difficulty can sit in one or two identifiable restraints. Take away Mbd3 and a change that took weeks and rarely succeeded becomes near certain within a week. That is strong evidence of localized resistance to reconfiguration. It is only candidate evidence for the kernel's interlock estimator, which counts closed loops (“more closed loops → more cross-constraint → more resistance”). A brake is not automatically a loop. The loop Rais and colleagues describe is started by the added factors themselves, which recruit Mbd3/NuRD; it is not shown to be one of the loops that hold the specialized cell in place. Ink4/Arf is a checkpoint that stops cell division, further still from a closed loop. Calling either an interlock would need a map of the regulatory network showing it takes part in the closed loops that keep the specialized state.

Stored history: the cell remembers what it was

The old identity survives the change. Reprogrammed cells taken early “harbour residual DNA methylation signatures characteristic of their somatic tissue of origin, which favours their differentiation along lineages related to the donor cell, while restricting alternative cell fates.” The memory “could be reset by differentiation and serial reprogramming, or by treatment of iPSCs with chromatin-modifying drugs” (Kim et al. 2010).

It fades with time. Cells reprogrammed from skin, blood and muscle showed “distinct transcriptional and epigenetic patterns” according to where they came from, and “continuous passaging of iPSCs largely attenuates these differences” (Polo et al. 2010).

Some of it does not fade. Soufi and colleagues, citing Lister et al. (2011), report that 22 of the silenced regions described below “fail to lose the H3K9me3 mark” in human reprogrammed cells, compared with embryonic stem cells. (Lister et al. was not read for this page.)

What this gives the page: the closest match on the page between a kernel term and a measurement. The record of the old identity is carried through the reconfiguration, it biases what the cell can become next, and most of it wears away over many divisions. On this site's terms it is α-trace.

Age: older cells cost more, for a specific reason

“Organismal ageing upregulates the Ink4/Arf locus and, accordingly, reprogramming is less efficient in cells from old organisms, but this defect can be rescued by inhibiting the locus with a short hairpin RNA” (Li et al. 2009).

What this gives the page: the cost rises with age, as the kernel would expect from accumulated history. But the paper traces the age effect to one brake that can be switched off, which is the named-restraint result again. The page does not claim that age raises logic mass through the stored record in general, and it does not rank the logic masses of young and old cells.

Basin depth: part of the sides of the valley is in the chromatin

Regions the factors cannot reach. Soufi, Donahue and Zaret mapped where the four factors bind in human fibroblasts during the first 48 hours. They found “264 contiguous swaths of the fibroblast genome of 2.2 megabase average size that were refractory to OSKM binding at 48 hr.” These regions hold genes the stem-cell state needs late in the process, including NANOG, DPPA4 and SOX2. Their main feature is the H3K9me3 mark, a signature of tightly packed, silenced chromatin, about 10-fold more enriched there than the related H3K9me2 mark.

Remove the mark and the regions open. Knocking down the enzymes that lay down H3K9me3 let Oct4 and Sox2 into these regions within 48 hours, with no change at sites just outside them. Knocking down any of these enzymes “significantly accelerated the appearance of newly reprogrammed colonies,” and knocking down SUV39H1/H2, which removed the most H3K9me3, “significantly enhanced the efficiency of reprogramming.”

The authors' reading is that the regions “impede the pace of reprogramming,” and that the impediment “is overcome” in the minority of cells that convert (Soufi et al. 2012).

What this gives the page: a candidate molecular contribution to the effective sides of the valley in Waddington's landscape. Basin depth belongs to the dynamics of the whole cell, including its chromatin, its network of transcription factors, its signaling and its metabolism. H3K9me3 contributes to the barrier; it is not shown to be the barrier. The regions slow reprogramming and are crossed by the cells that make it, which again fits logic mass as a cost. They are also a third named restraint that can be taken away, alongside Mbd3/NuRD and Ink4/Arf.

The older experiment, and why it is history here

In 1962 John Gurdon moved nuclei from fully specialized tadpole gut cells into frog eggs whose own nucleus had been removed. Of 726 gut-cell nuclei, 10 gave normal feeding tadpoles; the paper rounds this to “1½ per cent,” and the exact ratio is 1.4%. Nuclei from early embryos gave 100 normal tadpoles from 279 transfers (36%). Serial transfers showed that “at least 7 per cent” of the gut nuclei still held everything needed for a normal tadpole, and 24% once transfers that failed for technical reasons were set aside. His conclusion was that “the differentiation of a cell cannot be dependent upon the incapacity of its nucleus to give rise to other types of differentiated cell” (Gurdon 1962).

Why these numbers are not estimators. In nuclear transfer the donor cell is taken apart and only its nucleus goes into the egg. The donor cell loses its boundary, which under the kernel ends it, and the embryo is a new attractor: “Identity is not inherited across β.” Nuclear transfer is not a T₄ reconfiguration of the donor cell, so it cannot estimate the donor's logic mass. It belongs here as history. It showed that a specialized cell's nucleus keeps its full capacity, so whatever makes a cell costly to reprogram is not missing genetic information.

Read together

Kernel estimator	What was measured in the cell	Direction
Reconfiguration difficulty (candidate for T₃/T₄ cost)	About 0.02% to 0.05% success; weeks to first colonies; almost every cell converts eventually	Long random waiting time; not shown to be cost paid
Candidate interlock contribution	Removing specific restraints (Mbd3/NuRD, Ink4/Arf) makes reprogramming much faster and more frequent	Localized resistance; whether the restraints are closed loops needs a network-level map
α-trace richness	Methylation memory of the tissue of origin, fading with passaging; some silenced regions never lose their mark	History carried through the change
Basin depth (candidate contribution)	264 H3K9me3 regions that exclude the factors; removing the mark speeds reprogramming	Deeper basin, slower reprogramming

Each measurement moves in the direction predicted if these biological observables are valid substrate-specific estimators of the corresponding kernel quantities. The strength of the mapping varies. The α-trace reading is strong. The basin-depth reading is plausible. The interlock reading needs the network shown. And the waiting time cannot yet be told apart from a low chance of a rare transition, so it is not yet a reading of cost.

What this suggests about the descriptor

This section records an observation. It coins no construct and changes nothing in the kernel.

The kernel writes logic mass as one number per attractor, m^L(X, t). The cell suggests that reconfiguration difficulty has a shape. Reprogrammed cells carry a memory of their tissue of origin that “favours their differentiation along lineages related to the donor cell, while restricting alternative cell fates” (Kim et al. 2010): the same cell is cheaper to move toward its old lineage and costlier to move toward others. And each named restraint acts on a particular part of the process: removing the H3K9 mark opened the silenced regions and left the sites just outside them unchanged. Written loosely, the cost of a change from X to Y₁ can be much smaller than the cost from X to Y₂.

That looks less like a single mass and more like a cost that depends on the direction of the change, something like m^L(X, ΔX). The kernel's own wording already leans that way: it measures the cost “to produce equivalent structural change,” which only makes sense for a named change. If the scalar is kept, two cells could have similar overall difficulty and very different directions of cheap and costly change, and the scalar would not show it. The question is recorded for the kernel's author as an open candidate.

What this page does not claim

It does not measure logic mass; it reports estimators, as the kernel requires. It does not settle that the cell is sovereign; it inherits “sovereign candidate” from The Living Cell. It does not compare the logic masses of different cells or populations. It does not establish that the waiting time before reprogramming is transaction cost. It does not show that Mbd3/NuRD or Ink4/Arf are closed regulatory loops. It does not settle whether reprogramming across many cell divisions is one attractor's T₄ change or a change spread across a lineage. It does not claim a law linking the estimators to each other; the link between trace density and logic mass is stated on this site as a correlation, and this page adds nothing stronger. And it coins no construct: the case of a cell reconfigured by supplied factors is handled with the kernel's existing T₄.

Sources

Takahashi, K., Tanabe, K., Ohnuki, M., Narita, M., Ichisaka, T., Tomoda, K., and Yamanaka, S. (2007). Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell 131, 861–872. doi:10.1016/j.cell.2007.11.019. Full paper read.
Hanna, J., Saha, K., Pando, B., et al. (2009). Direct cell reprogramming is a stochastic process amenable to acceleration. Nature 462, 595–601. doi:10.1038/nature08592. Abstract and author manuscript read.
Rais, Y., Zviran, A., Geula, S., et al. (2013). Deterministic direct reprogramming of somatic cells to pluripotency. Nature 502, 65–70. doi:10.1038/nature12587. Abstract read.
Li, H., Collado, M., Villasante, A., et al. (2009). The Ink4/Arf locus is a barrier for iPS cell reprogramming. Nature 460, 1136–1139. doi:10.1038/nature08290. Abstract read.
Kim, K., Doi, A., Wen, B., et al. (2010). Epigenetic memory in induced pluripotent stem cells. Nature 467, 285–290. doi:10.1038/nature09342. Abstract read.
Polo, J. M., et al. (2010). Cell type of origin influences the molecular and functional properties of mouse induced pluripotent stem cells. Nature Biotechnology. doi:10.1038/nbt.1667. Article page summary read.
Soufi, A., Donahue, G., and Zaret, K. S. (2012). Facilitators and impediments of the pluripotency reprogramming factors' initial engagement with the genome. Cell 151, 994–1004. doi:10.1016/j.cell.2012.09.045. Full paper read.
Gurdon, J. B. (1962). The developmental capacity of nuclei taken from intestinal epithelium cells of feeding tadpoles. J. Embryol. Exp. Morph. 10(4), 622–640. Full paper read.
Not read: Takahashi, K., and Yamanaka, S. (2006), Cell 126, 663–676, the first mouse paper; Lister et al. (2011), cited through Soufi et al.
Kernel: LOGIC-MASS § 8.4, SOV § 7.2 and T₄ § 9.5, quoted from Kernel v2.39.0.

Logic mass is a Tier-3 persistence descriptor, defined only for a sovereign attractor during its bootstrapping interval. This page reports substrate-specific estimators from published cell-reprogramming experiments and reads them against the kernel's definition. The cell's standing as sovereign is carried over from The Living Cell as a candidate. This page coins no construct and modifies no canon.

← Back to the metrics