# Cells Control Their Size: A Measured Eukaryotic Cell

> Many of a cell's processes depend on concentrations, amount divided by volume, so growth dilutes anything whose amount does not keep pace. This lesson presents the leading, still-debated model in which budding yeast reads its size from the dilution of a brake called Whi5; shows why most of what a growing cell makes must keep pace with its volume, as the fission-yeast nucleus does; and asks where a cell can fall behind: a single genome appears to become limiting for transcription in greatly enlarged cells, and surface area grows more slowly than volume when a cell of fixed shape gets bigger. It works the surface arithmetic to the radius where supply meets demand, shows how plant cells add volume as vacuole instead of contents, and ends with a compact reference to animal, plant and fungal parts, with measured shares tied to named cells.

Course: Cell Biology (Seeing and testing cells). Level: core university (third year). About 25 minutes.
Lesson page (link this one when citing): https://stemhop.com/lessons/measured-eukaryotic-cell/

Builds on: Size, Scale and Surface · The Cell Cycle and Mitosis · Organelles as a System

## What you'll learn

- Explain the leading, debated model in which budding yeast reads its size from the dilution of the brake Whi5, and use it to predict what a cell born small or large, or a cell given a different Whi5 dose, does at Start.
- Explain why a growing cell must make RNA and protein in step with its volume, why their steady concentrations carry no size signal, and what happens to its cytoplasm when it grows too large for its synthesis.
- State a measured compartment share together with its cell type, explain why a share measured in one cell type does not transfer to another, and describe how fission-yeast experiments showed that the nucleus scales with the cell rather than with its DNA.
- Calculate surface area, volume and their ratio for a model spherical cell, find the radius at which a stated uptake just meets a stated demand, and say what that arithmetic leaves out, including how a plant vacuole changes it.
- Compare animal, plant and fungal cells by wall, plastids and vacuole, and say what a result in a model organism, such as humanized yeast, does and does not show.

**In 1977 Leland Hartwell and Michael Unger followed budding yeast as it divided unequally, into a large mother cell and a small daughter. The daughter waited longer before starting its next cycle, as if it first had to reach a size.**

The first film above followed one ratio, amount over volume; it leaves out the evidence behind each step, and where a cell’s volume goes. The sections below give the evidence for the Whi5 ruler, for contents keeping pace, and for the two costs, the genome and the surface; the second film then follows the compartments that share the volume, how root cells add vacuolar volume, and a short reference to animal, plant and fungal parts.

> **The one idea to carry through this lesson.** Many of a cell’s processes depend on concentrations: amount divided by volume. Growth raises the volume, so anything whose amount does not keep pace is diluted. In the leading model, budding yeast reads its size from one such dilution, of a brake called Whi5. Most other contents a growing cell must make in step with its volume. Two things can fall behind and set costs of being big: a single genome, which appears to serve only so much cytoplasm, and the surface, which grows more slowly than the volume when a cell keeps its shape.

A note on units: cell volumes are given in femtolitres (fL) or cubic micrometres (µm³), and the two are the same, 1 fL = 1 µm³.

## Reading size: the brake that growth dilutes

The first film’s ruler needs a few terms, then its evidence. The budding-yeast cycle runs through G1 (growth before the DNA is copied), S (DNA synthesis), G2 and M (mitosis). **Start**, the first film’s green line, is the point in G1 where the cell commits to a full cycle of copying its DNA and dividing (the mammalian version, the restriction point, is in [cell-cycle control](https://stemhop.com/lessons/cell-cycle-control/)). Whi5 is a protein that holds back the transcription factor (a gene-switching protein) that turns on the genes for Start; as Whi5’s concentration falls, the brake weakens. A rule under which cells divide at about the same size, whatever size they were born, is a **sizer**.

The first film’s red dots stand for a measurement. In 2015 Kurt Schmoller, Jan Skotheim and colleagues tagged Whi5 with a [fluorescent protein](https://stemhop.com/methods/fluorescent-protein-tagging/) and measured it in single living cells. Whi5 was made mostly in S, G2 and M, in an amount that hardly depended on the cell’s size. A small daughter therefore began G1 with about as many Whi5 molecules as a larger cell, at a higher concentration. That premise does all the work: if the dose at birth grew in proportion to birth size, every cell would start at the same concentration, and dilution would tell it nothing about its size.

**Predict.** A strain carries two copies of the *WHI5* gene instead of one, so each newborn cell gets about twice the usual dose. Do these cells pass Start at their normal size, a smaller size, or a larger one? And what would deleting *WHI5* do?

> **Check your prediction.** Larger. In the idealized model, if the Whi5 dose stays fixed during G1 and Start occurs at the same Whi5 concentration, doubling the dose doubles the volume at that threshold. Experimentally, adding *WHI5* copies made cells larger, though not by a guaranteed twofold amount. Cells lacking Whi5 pass Start small.

The first film used the simplest case, a hard line that every cell crosses at the same size; here is the fuller picture. The budding-yeast sizer is weak and noisy: cells pass Start over a range of sizes, though the chance rises as they grow, so a cell born small makes up only part of its shortfall in one G1. And how much of size control Whi5 dilution explains is still argued (Going deeper).

Fission yeast, a rod that grows at its ends, reads size another way, before mitosis. It divides only once it reaches a set size, so a cell born small grows more first, a cell born large grows less, and sizes at division come out similar. Its mutants that divide too early or too late are the size mutants in the design problem below.

## Keeping pace: RNA and protein

The ruler works because Whi5 does not keep pace; nearly everything else has to. The rate of a reaction depends on the concentration of the molecules that carry it out, so a growing cell has to make more RNA and protein to keep their concentrations steady. In most cells it does: total RNA and protein rise roughly in step with volume, because larger cells transcribe and translate faster. The first film’s flat orange gauge showed concentration only: the total amount still grows with the cell, but the concentration carries no size signal.

## What can fall behind: the genome

The first film’s first cost, the genome, ended on a question. Its swelling cells come from a 2019 study by Gabriel Neurohr, Angelika Amon and colleagues: held in G1, budding-yeast cells grew from about 65 fL to about 600 fL in six hours, and their cytoplasm was diluted.

**Predict.** After six hours these cells are both much bigger and much longer arrested. How could you tell whether the dilution comes from their size or from the time spent arrested?

> **Check your prediction.** Hold some cells for the same time but let them grow less. The team did this by limiting glucose during the arrest, so these cells spent as long arrested but stayed smaller. If dilution comes from time, they should be diluted too; if from size, they should be spared. The comparison pointed to size: it was the enlarged cells that failed to scale up synthesis.

Why the genome? RNA is copied from DNA, so a cell with one genome copy has a fixed set of templates. At very large sizes the nuclear DNA itself appears to become limiting for transcription; if so, one genome can serve only so much cytoplasm. Cells that copy their genome without dividing, or carry extra genome sets, tend to be larger, which fits that idea. In enlarged human fibroblasts (connective-tissue cells) the same study could not measure protein and RNA directly, but found a sign of reduced crowding that fits dilution. Recent work suggests that excess size itself helps drive senescence, the permanent exit from division. A cell can be too big for its own synthesis machinery, not only for its membrane.

## What falls behind: the surface

The first film’s second cost, now with numbers. For its sphere going from 5 to 10 µm in radius, A = 314.2 → 1256.6 µm² and V = 523.6 → 4188.8 µm³, so A/V = 3/r falls from 0.6 to 0.3 µm⁻¹. The surface still grows, only more slowly than the volume while the shape stays the same. A rate on each side, uptake J per unit of surface and use q per unit of volume, shows where that bites.

> **Equation card: The largest radius at which a spherical cell's surface can take in as much of a nutrient as its volume uses.**
>
>     r_max = 3J / q
>
> In words: The largest radius is three times the uptake per unit of surface, divided by the use per unit of volume.
>
> - r_max: the radius at which supply just meets demand (µm)
> - J: the most nutrient the membrane can take up, per unit area per second (molecules µm⁻² s⁻¹)
> - q: the nutrient each unit of volume uses per second, evenly through the cell (molecules µm⁻³ s⁻¹)
>
> Sanity check: Units: J/q leaves µm, a length, as a radius must be. The 3 comes from A/V = 3/r for a sphere.
> Try one number Invent two rates; they are made-up inputs, not measurements. Let J = 20 molecules µm⁻² s⁻¹ and q = 6 molecules µm⁻³ s⁻¹, with the whole surface taking up nutrient and nothing else limiting the cell. Then r_max = 3 × 20 / 6 = 10 µm. The ratio of capacity to demand is (J/q) × (3/r) = 10/r. At 5 µm it is 2 (supply 6,283 molecules per second against demand of 3,142), and at 10 µm it is exactly 1. Change J, q or the shape and the answer changes.
> Where it comes from optional Supply meets demand when J × A ≥ q × V, with A = 4πr² and V = 4πr³/3. Divide both sides by J × V: A/V ≥ q/J. For a sphere A/V = 3/r, so 3/r ≥ q/J, which gives r ≤ 3J/q.

**Predict.** Keep J = 20, but let the cell’s demand double to q = 12. Where is the new rmax, and what does a 10 µm cell now face? Then check with the sliders below.

> **Check your prediction.** rmax halves, to 3 × 20 / 12 = 5 µm. A 10 µm cell now takes in only half of what it uses: the capacity-to-demand ratio is 5/r, which is 0.5 at 10 µm. Doubling demand moves rmax exactly as halving uptake would.

*Interactive on the lesson page:* Cell size: supply against demand

So the sphere gives a cost that grows with size (the simulation calls it a scaling penalty), not a fixed largest size. The first film’s sphere was also the simplest shape. A flat or long thin cell keeps much more surface per volume than a sphere, and diffusion time grows with the square of distance, which also favours cells where no point is far from the surface. “Eukaryotic cells are 10 to 100 µm across” is a guide, not a bound: a 65 fL yeast cell has a sphere-equivalent diameter of about 5 µm. Geometry sets a cost on being big; the coupling of growth to division chooses the size.

So far the cell has been one bag of contents, but its volume is split among compartments, and the share each one takes can be measured.

The second film picks up here: what share of a cell each compartment takes, and how a plant cell grows by adding vacuole.

**Film transcript: Where the Volume Goes**

### Film transcript: Where the Volume Goes

A cell looks like one bag. But where does all its volume go? Chapter one followed amount over volume as a yeast cell grew. Now ask what share each compartment takes, and whether that share holds during growth. A membrane-bound compartment is a region enclosed by a membrane. The nucleus, which holds most of the DNA, is one.

Divide its volume by the cell's, and you get its share. Here is one carefully measured cell line: hamster V79, growing in a dish. The nucleus fills about 27% % of the cell. Mitochondria, another compartment, fill about 5%. Now a fission yeast, a rod that grows at its ends. Its nucleus is about 8%. percent. In this yeast, the nucleus grows with the cell, so its share holds as the cell grows. Two careful measurements, two very different answers. So which number is the nuclear share of a cell: 27%, % or 8%? Pause and decide.

Both, each for its own cell. There is no universal nuclear fraction. A share means something only with its cell type attached, so carry the cell type with every number. One question stays open. Does the nucleus follow the DNA inside it, or the cell around it? The design problem below lets you test it.

Plant cells change their shares as they grow. Take the outer layer of a root in Arabidopsis, a small lab plant. In young cells, the vacuole fills about 30 to 40%. percent. As the cells lengthen, vacuolar occupancy rises to about 80 to 90%. percent. Across the comparison, cell volume is about 14-fold greater. But space outside the vacuole is only about two- to threefold greater. Most of the added cell volume is vacuolar. The smaller rise outside it requires relatively little new RNA and protein. This can ease a biosynthetic cost of being big.

Which cells have which parts? Animal, plant and fungal cells share a plasma membrane, nucleus, mitochondria, ER, Golgi, and cytoskeleton. For these core parts, focus on three differences: walls, plastids, and vacuoles. A root cell deep in soil usually has plastids, but no chloroplasts. Fission yeast keeps its nuclear share as it grows.

Shares are cell-specific; elongating Arabidopsis root cells add mostly vacuolar volume.

From the film

- A membrane-bound compartment's share is its volume divided by the cell's, and it belongs to one cell type: the nucleus is about 27% of a cultured hamster V79 cell but about 8% of a fission-yeast cell.
- There is no universal nuclear fraction, so every measured share should carry its cell type.
- In fission yeast the nucleus grows with the cell, so its share holds as the cell gets bigger.
- As Arabidopsis root cells lengthen, vacuolar occupancy rises from about 30 to 40% of the cell to 80 to 90%, so most added cell volume is vacuolar, with comparatively little added cytosol.
- For the core parts compared here, animal, plant and fungal cells differ in walls, plastids and vacuoles; a root cell deep in soil usually has other plastids but no chloroplasts.

## Where the volume goes: compartments and their shares

The first film’s purple block was DNA alone; the second film adds the nuclear envelope around it. The second film left one question open: does the nucleus follow its DNA or its cell? Many other organelles also scale in size or number with the cell, in ways that depend on the organelle and cell type, but the nucleus is where this was tested most cleanly.

### Design problem

Does DNA content or cell size set the size of the nucleus? You work with fission yeast and have size mutants that divide at very different sizes, a way to raise DNA content without division (switching off Cdc13, the cyclin the cell needs to enter mitosis, so the genome is copied again and again), a fluorescent nuclear-envelope marker, and images that outline each cell. Design the experiment and say what each model predicts before you look.

> **One way to do it.** First, measure nuclear and cell volume in each cell, across the widest size range you can get from the mutants. If cell size sets nuclear size, the nuclear-to-cell ratio stays constant as cells get bigger. If DNA alone sets it, nuclear volume stays roughly fixed in cells with the same DNA content, so the ratio falls as cells grow.
>
> Second, change DNA content without changing the rule for cell size. Switch off Cdc13 so the genome is copied several times, and compare cells with 2C and with much higher DNA content (C is the DNA of one haploid genome). The DNA model predicts a relatively larger nucleus in the high-DNA cells; the cell-size model predicts the same ratio.
>
> Controls: measure wild-type cells in the same conditions, use the same marker and thresholds for every strain, and measure enough cells per strain to see the spread.
>
> Neumann and Nurse did this in 2007. The ratio stayed roughly constant, near 0.08, across a 35-fold range of average cell sizes, and a 16-fold difference in DNA content (2C against 32C) left relative nuclear size unchanged. In cells with several nuclei, a nucleus surrounded by more cytoplasm grew faster. The answer in fission yeast is cell size, probably through something in the cytoplasm.

Here are the two measured shares behind the second film, each with its cell type and method.

- **Chinese hamster V79-S171 cells, log phase, in culture:** mean volume 710 µm³, nucleus 190 µm³, total mitochondria 37.5 µm³, about 90 mitochondria per cell (an estimate from electron-microscope sections that depends on how a mitochondrion is counted; see Going deeper). In 1972 D. W. Ross and H. C. Mel took whole-cell volume from electronic cell sizing and apportioned it among nucleus, cytoplasm and mitochondria by [electron microscopy](https://stemhop.com/methods/electron-microscopy/).
- **Wild-type fission yeast, growing:** nuclear-to-cell volume ratio 0.080 ± 0.013, measured by Frank Neumann and Paul Nurse with a fluorescent nuclear-envelope marker.

**Reading a volume table.** From the V79 means:

> nuclear fraction = 190 / 710 = 0.27; mitochondrial fraction = 37.5 / 710 = 0.053

So in fission yeast the nuclear share holds as the cell grows, while the hamster and yeast measurements show that shares differ between cell types. Carry the cell type with each number.

Treated as a sphere of the same volume, a V79 cell has r³ = 3 × 710 / (4π) = 169.5 µm³, so r = 5.5 µm and the diameter is about 11.1 µm. That is a calculation, not a measured diameter; cells spread on a dish are flatter. How these volumes are measured comes in [microscopy and quantitative images](https://stemhop.com/lessons/microscopy-and-images/).

## Plants add volume mostly as vacuole, not cytoplasm

The second film showed the synthesis side in Arabidopsis root epidermis (the root’s outer cell layer). Vacuolar occupancy rose from about 30 to 40% in young cells to 80 to 90% after elongation. Across that comparison, cell volume was about 14-fold greater, but the space between vacuole and cell boundary was only about two- to threefold greater. The surface side is yours to work out.

**Predict.** Treat the growing cell as a sphere whose volume rises 14-fold. What happens to membrane area per unit of volume? And per unit of cytoplasm, if the cytoplasm grows only two- to threefold?

> **Check your prediction.** A sphere’s area grows as its volume to the power two-thirds, so a 14-fold rise in volume gives 142/3 ≈ 5.8-fold more area. Per unit of volume, membrane therefore falls to 5.8 / 14 ≈ 0.4 of its starting value: the surface cost is still there. Per unit of cytoplasm it rises about two- to threefold (5.8 divided by 2 to 3). A real root cell elongates rather than swelling as a sphere, which keeps even more surface. For the cytoplasm, the vacuole turns the surface cost around.

The vacuole cannot do this alone. The cell wall has to loosen so the cell can stretch, and vacuole expansion and wall loosening must be coordinated. Which one leads, and exactly what the vacuole contributes, is still being worked out.

## Same parts, different shares

This last section is a reference tool, not a new step: the shares above only mean something if you know which parts each kind of cell has.

A eukaryotic cell keeps most of its DNA inside a membrane-bounded nucleus; mitochondria, and plastids in plants, carry small genomes of their own. Its interior is divided into compartments, regions enclosed by a membrane. The space inside an organelle is its lumen; the fluid outside the organelles but inside the plasma membrane is the cytosol.

Animal, plant and fungal cells share a plasma membrane, a nucleus and a cytoskeleton, the network of protein filaments that shapes the cell and moves things within it. All three have an endoplasmic reticulum (ER) and Golgi apparatus, which sort and transport proteins headed for secretion, and mitochondria, which carry out [oxidative phosphorylation](https://stemhop.com/lessons/oxidative-phosphorylation/). The bilayer that bounds all of these is covered in [membrane structure and function](https://stemhop.com/lessons/membrane-structure-function/), and the broad picture of the cell in [the introduction to biochemistry and cell biology](https://stemhop.com/lessons/intro-biochemistry-cell-biology/).

The differences are where students slip:

| Part | Animal cell | Plant cell | Fungal cell (yeast) |
| --- | --- | --- | --- |
| Wall | none | cellulose fibres in a matrix of hemicelluloses and pectin | β-glucans, chitin and mannoproteins (*S. cerevisiae*) |
| Plastids | none | chloroplasts in photosynthetic cells; in nonphotosynthetic tissues, leucoplasts without pigment or, in many fruits and flowers, pigmented chromoplasts | none |
| Vacuole | small or absent | often large, can fill most of the cell | acidic, with degradative and storage roles |

Chloroplasts carry out the [light reactions](https://stemhop.com/lessons/light-reactions/), but only in cells that photosynthesize. A root cell deep in soil usually lacks chloroplasts, though it still contains other plastids. Their number is a measured share too: wild-type Arabidopsis leaf mesophyll (Wassilewskija) holds a mean of 83 chloroplasts per cell, a figure for one tissue of one plant. Yeasts are unicellular fungi, and budding yeast has the same core organelles as your cells, which is why it is so useful.

*Figure (AI-generated illustration).* Animal, plant and fungal cells share a plasma membrane, nucleus, ER, Golgi, mitochondria and cytoskeleton. The slips are in the wall, the plastids and the vacuole, and a root cell is usually a plant cell with no chloroplasts. Image: three schematic cells side by side. An animal cell with no wall, a plant cell with a cellulose wall, green chloroplasts and a large vacuole, and a yeast cell with a wall and a small vacuole; all share a nucleus, ER and Golgi, mitochondria and cytoskeleton.

## Going deeper

### Sizers, adders and timers

A sizer can correct a size error within one cycle. An adder adds about the same volume each cycle whatever the birth size, so a size error shrinks only over several generations. A timer keeps cycle time constant; on its own it cannot correct size in cells that grow exponentially, though it can hold the sizes a sizer has set. Budding yeast shows a weak sizer in G1 and a timer for the rest of the cycle, so its whole cycle behaves roughly like an adder; fission yeast behaves as a sizer overall. Many cultured mammalian cell lines behave close to adders; one proposal, not settled, is that cancer mutations disrupt their G1/S control. Whi5 works much like the mammalian brake Rb, and plants use dilution too: in Arabidopsis the inhibitor KRP4 is diluted with growth.

### Debates over what is measured

Whi5 dilution is the leading model, not a closed case: later measurements disagree about how far Whi5 concentration falls during G1 and how much of size control it explains (for example, Litsios and colleagues in 2019 and Barber, Amir and Murray in 2020, with follow-ups from the Schmoller and Skotheim groups). Fission yeast reads several inputs at once. One is the activator Cdc25, whose concentration rises as the cell grows, the mirror image of Whi5. Whether it measures its length, surface area or volume is debated; a sensor at the cell cortex built around the protein Cdr2 has been proposed to read surface area.

Growth rate feeds in too. In bacteria and yeasts, cells that grow faster, in richer media, are larger; a similar trend is reported in mammalian cells, but it is weaker and depends on how growth is changed.

### The enlarged-yeast experiment

The arrest used a [temperature-sensitive](https://stemhop.com/methods/temperature-sensitive-mutants/) *cdc28-13* allele. In primary human fibroblasts (IMR90) enlarged by senescence induction, the evidence was faster diffusion of tracer nanoparticles, the indirect sign of reduced crowding mentioned above.

### Counting organelles

The count of about 90 mitochondria per V79 cell depends on the counting rule, since mammalian mitochondria can also join into branched networks. In budding yeast, image analysis shows that mitochondria can form one interconnected network or many separate tubules, so the count depends on the rule: counting each connected structure as one object makes a single network one mitochondrion. This lesson gives no census of organelle numbers across animals, plants and fungi.

### Model organisms and the limits of transfer

This lesson’s numbers come from two yeasts, a hamster line and Arabidopsis; each model answers some questions and not others.

**Budding yeast** (*Saccharomyces cerevisiae*) is fast and easy to manipulate genetically ([CRISPR and functional genomics](https://stemhop.com/lessons/crispr-and-functional-genomics/)). In 2015 Aashiq Kachroo, Edward Marcotte and colleagues tested human orthologs (the matching human genes) in place of 414 essential yeast genes, and 176 (43%) restored growth in their own assays; adding earlier published cases brought the rate to 47%. That shows a conserved function in those genes. It does not show that the rest are unconserved, since failure can have several causes, and growth rescue says nothing about tissue roles.

*Figure (AI-generated illustration).* With the yeast copy of an essential gene shut off, cells with an empty plasmid cannot grow, while cells given the human ortholog can. Growth rescue shows a conserved function; failure alone does not say why. Image: two side-by-side frames of budding yeast whose own copy of an essential gene is shut off. With an empty plasmid a single cell does not grow; with a plasmid carrying the human ortholog the cells bud and grow.

**Fission yeast** (*Schizosaccharomyces pombe*) grows at its ends, which makes size easy to measure.

**Cultured mammalian lines** give precise numbers, like the V79 volumes, but a line is not an animal, and a line’s name is not a guarantee. Uri Ben-David and colleagues characterized 27 strains of the breast cancer line MCF7 in 2018 and found rapid genetic diversification: at least 75% of the compounds that strongly inhibited some strains were completely inactive in others. Report the strain and conditions.

**Flies** (*Drosophila*): a 2000 survey by Fortini and colleagues found apparent fly counterparts for 178 of 287 human disease genes, and a larger 2001 search by Reiter and colleagues matched 714 of 929 (77%). Conservation is gene by gene: no fly counterpart of α-synuclein was found, yet flies made to express human α-synuclein are used to model Parkinson’s disease.

**Worms** (*C. elegans*) are transparent and mostly self-fertilizing, so cells can be watched in a living animal and a mutation breeds true without a mate. Sydney Brenner’s 1974 paper linked about 300 EMS-induced mutants to about 100 genes, from visible defect to mapped gene. Human effects need more evidence.

**Plants** (*Arabidopsis*): the *arc6* mutant, with one to three chloroplasts per mesophyll cell against 83 in wild type, and the root vacuole measurements show control of chloroplast number and vacuole size that animal cells do not offer. Leaf and root data are not interchangeable.

---

From STEMhop (https://stemhop.com), free lessons by Christopher Deppmann. Lesson text © the authors. Practice questions: https://stemhop.com/lessons/measured-eukaryotic-cell/
