The title is my story’s conclusion.

We know that calcium phosphate crystals in plaque, plugs, and the vast majority of stones are hydroxyapatite, yet when we measure supersaturation with respect to calcium phosphate we measure that for brushite, the pretty crystal in the big image that begins this work.

We do it because brushite crystals may initiate calcium oxalate stone disease, so it is wise to prevent their forming.

The beautiful image is a hand colored electron micrograph of brushite crystals by Kseniya Shuturminska. I have written about brushite on this site before, but this crystal is so important yet buried in that eerie world where atoms and molecules combine, it seemed to deserve a larger place – which I have now given it.

What Brushite Is

Put simply it is CaHPO4.2H2O, a calcium phosphate crystal. It forms more readily than hydroxyapatite or calcium oxalate. 

In being a precursor form, brushite is similar to the ‘amorphous’ outer layers of bone mineral and to the first phase to deposit over plaque exposed to urine. Brushite can convert to hydroxyapatite. The work I focus on here shows it can nucleate calcium oxalate by losing surface calcium atoms to oxalate. 

Brushite Comes First in Human Urine

As he added calcium to human urine, Charles Pak found the first crystals were brushite. Of 28 urine samples 25 contained only brushite crystals.

It was this observation that led him to work out the mathematics for calculating supersaturation with respect to brushite, and show that it was higher in stone formers than in normal subjects.

He predicted, correctly, that brushite could be a critical first step in formation of kidney stones.

Work many years later revealed exactly how brushite mediates between solution chemistry and stone crystal formation.

It is this work that I will now present to you.

Oxalate Dismantles Brushite Seeds

Brushite comes first, but oxalate can extract brushite calcium atoms to build calcium oxalate crystals in its stead.

The Constant Composition System

Solutions supersaturated with both brushite (1.2 fold) and calcium oxalate (1.8 fold) were kept at constant composition by a system of sensors and servo driven pumps that could replenish what was lost as crystals took up calcium, or oxalate, or phosphate. Because the pumps are regulated to keep solution concentrations constant, their rates of replenishment measure exactly uptake of calcium, phosphate, and oxalate by crystals as they form and grow. Put simply the pump rates measure crystal growth.

Brushite Pumps

One pair: Calcium and phosphate exactly calculated to make up for growth of brushite. As brushite grows it takes up equal amounts of calcium and phosphate from the solution. A calcium sensing electrode signals the calcium pump to add more. A pH electrode senses uptake of phosphate because, as brushite forms, HPO4= is selectively removed from solutions that contain it as well as H2PO4. This drives the equilibrium: H2PO4 ↔ HPO4= + H+ to the right, so formation of brushite can be detected by a fall in solution pH.

We see dual signaling (calcium + pH) in the heavy curving line labeled “Brushite” on the figure just below. When that curve flattens, the electrodes are no longer signaling the calcium AND phosphate pumps: That means brushite is no longer growing.

Calcium Oxalate Monohydrate (COM) Pumps

One pair: Calcium and oxalate exactly calculated to make up for growth of calcium oxalate monohydrate (COM). As COM grows it takes up equal amounts of calcium and oxalate from the solution.

The sensors know to run this pair when the calcium electrode is signaling loss of calcium, and the pH electrode is signaling constant solution pH – calcium oxalate monohydrate formation does not alter pH of the solution.

We see single signaling (calcium + constant pH) in the thin curving line labeled “COM” on the figure just below.

I should note that the experimenters refer to calcium oxalate as the monohydrate. In fact they did not determine if it was the mono or the dihydrate but that does not matter to their results and I thought it best to use their nomenclature.

Likewise, I refer to pumps as in pairs, but that is only by way of explanation. There may be only one calcium pump to serve both the phosphate and oxalate pumps. It would not matter logically.

Dual Growth Conditions

When brushite and COM both grow, both sets of pumps run. The Brushite pumps replace calcium in exact proportion to phosphate (judged from pH change), the COM pumps replace calcium to keep solution calcium constant and add oxalate in exact proportion to it.

This ‘constant composition’ method was developed by George Nancollas, one of the authors. 

Added Brushite Seeds Vanish into Calcium Oxalate

Evidence from Solution Chemistry

On the left side of the adjacent panel, brushite crystals were added at about 90 minutes (x axis), and as they grew titrant was added (the curve labeled “Brushite”) containing calcium and phosphate to keep the solution steady.

The volume of that solution is plotted along the vertical axis. Because it marks exactly how much calcium and phosphate must be added to keep the solution concentrations constant it is the exact rate at which the brushite seeds grew.

A sample of solids at SG-I contained ~40 mg, and – though not shown – would have had a molar ratio of calcium to phosphate of 1, that being the composition of brushite itself.

About 40 minutes later, the titrant solutions for calcium oxalate began (lighter curve labeled “COM”, calcium oxalate monohydrate) and rapidly rose as that for brushite fell to 0 – flat “Brushite” line.

The total solids at SG-II showed only 27 mg and the phosphate to calcium ratio was 0.72, consistent with the presence of calcium oxalate which has no phosphate. By the end of the experiment, SG-III, the solid material had a ratio of ~0, consistent with pure calcium oxalate.

Make no mistake. The sensors and pumps cannot be gotten around. Measurements shown in the paper prove the solution compositions were constant. So the “Brushite” and “COM” curves must mark growth of those crystals.

What stopped the brushite from growing?

Do not say the calcium oxalate took away so much calcium from the solution they could not grow. The pumps kept the solution calcium constant, phosphate, too.

The brushite crystals stopped taking up calcium and phosphate – the flattening of the brushite curve – because losses of calcium from their surfaces matched entry, so they could not grow and thereby donate protons to the solution and trigger the phosphate pump.

Evidence from Electron and Atomic Force Microscopy

Scanning electron microscopy (SEM) is intuitively obvious, like a light microscope but using electrons to get higher resolution – visualize smaller objects.

In panel a the solids harvested at SG-II in the picture above show brushite crystals (by SEM) roughening as they lose calcium and phosphate from their surfaces, and tiny calcium oxalate monohydrate crystals growing on them (Arrows mark COM crystals). Oxalate has taken the calcium off of brushite to make calcium oxalate. Deprived of their mates, phosphates drift into the water to live as roaming, solitary ions.

Atomic force microscopy is a lot wilder than EM. A very (very, very, very) tiny stylus sweeps over the crystal surface like a phonograph needle vibrating in the grooves of a vinyl LP record. In place of music encoded on vinyl, are the irregular planes calcium and phosphate make as they combine into brushite, or pits on those very same plates as calcium is pulled away by oxalate, and brushite dissolves.

In panel b are normal growth plates of brushite. In panels c and d pits (arrows) form, and in panel e calcium oxalate monohydrate crystals lie on the brushite surface.

Just What Have We Been Shown?

This brilliant science demonstrates what we usually see with underwater photography.

Brushite crystals, fully formed in their intricate lacework of calcium atoms and phosphate ion molecules locked together by powerful electrostatic forces, are submerged in water teeming with a rapacious predator. Oxalate ions rip calcium atoms out of the brushite structure, tear it apart, leaving behind actual holes where tissue has been eaten away to make a different crystal, the one that plagues most stone formers – calcium oxalate.

How do we know?

The Solutions Tell Us

Look back, at the vertical axis on the figure we just reviewed.

The graphs alone are enough to say the brushite grew and then stopped growing, and calcium oxalate crystals began to grow (oxalate is the only other ligand in the solution).They imply but cannot by themselves prove beyond doubt that calcium oxalate nourished itself by eating the calcium from the fabric of the brushite crystal.

Microscopy Tells Us Oxalate Destroys Brushite

The scanning EM is pretty graphic.

Brushite seeds look woebegone. Their surfaces are all roughed up. And all over it are tiny calcium oxalate crystals, making themselves out of brushite calcium.

The atomic force microscopy is even more graphic, isn’t is? There are the pits, the holes where growth plates are eaten into.

How much more do you need?

COM Replaces Brushite as a Solid

As a final proof, harvested solids turn from brushite (phosphate:calcium ratio = 1) to calcium oxalate (phosphate:calcium ratio = 0).

What Happens in Kidneys?

Oh, you will say, this is lovely but no one sprinkles brushite seed crystals into kidneys, or ureters. The science is not germane to the reality of my patient, my kidney, my husband’s kidneys.

We have more.

Same Solutions, No Seeds

I have brought the same figure back here, so you can see it along with this text. Take a look at the right side of it, where ‘N-I’ is.

Solution Behavior

The solution was as in the left side, SS of 1.2 for brushite, 1.8 for calcium oxalate. Pumps were ready, electrodes measuring. Nothing happened for nearly 1400 minutes – that is about one day.

Then, the pumps turned on, the ones for brushite – pH was falling along with calcium. Brushite grew merrily but by 100 minutes you can see the COM pumps were also running. By N-II brushite growth was over, and COM was taking off.

It is the same as before except instead of adding brushite seeds you just wait. Brushite forms all by itself. Then oxalate takes its calciums away to make calcium oxalate.

Analyses of the Solids Formed

The solid phase at N-I were pure brushite by x ray diffraction. Their ratio of phosphate to calcium was 1. No doubt pure brushite. Just as in urine, brushite is the first crystal to appear on the scene.

By N-II, x ray diffraction showed brushite + COM, and the ratio was 0.67. By N-III, COM predominated with some brushite. The ratio was 0.16. By N-IV, pure COM by X-ray and the ratio was 0.

Brushite formed, oxalate ate it all up to form calcium oxalate.

Pictures Proving Cannibalization

I have lavished much white space on this SEM picture.

Inside the larger ring is crystal without phosphate, just calcium. It is from the N-II solids that x ray diffraction proved was both brushite and COM. This portion is pure COM – no phosphate.

The inside core, within the smaller circle, is brushite. It has calcium and phosphate in nearly equal measure.

This is the perfect finish to the proof that oxalate attacks brushite from the outside, rips calcium off ifs surface to make itself, and gradually replaces the brushite, here reduced to a core.

You may wonder why there is no peak for oxalate. The reason is that its atoms (carbon, oxygen, and hydrogen) cannot be detected by this kind of analysis. So the partner to calcium is missing. Since there are no choices of partner except phosphate and oxalate, we can be sure the upper portion is indeed calcium oxalate.

What Does This Mean?

An Overview

It means what Charles Pak first envisioned.

As kidneys conserve water, their natural function, they concentrate calcium, phosphate, and oxalate, raising supersaturations for brushite and calcium oxalate. Brushite comes first, grows for a while. Then, oxalate begins to eat it up and make itself.

Brushite comes first because it is a more simple crystal, forms more readily. In fancy talk it has a lower energy of formation, needs less supersaturation, forms faster.

For patients and their physicians it means that urine brushite supersaturation is very important, perhaps most important of all. If it is below 1, the urine is deprived of a useful – perhaps essential pathway to calcium oxalate crystal formation.

What About Hydroxyapatite?

The phosphate in stones from the vast majority of calcium oxalate stone formers is not brushite, it is hydroxyapatite. So is the mineral in plaque, and tubule plugs. The story is not so different. Hydroxyapatite forms on brushite like calcium oxalate, and gradually eats it up, so all you have at the end in hydroxyapatite, no brushite. This is amply proven in experiments other than the one I made this article out of.

But to show you that work would make this article overly long.

What About Plaque as a Base for Calcium Oxalate Stones?

It is a different story. Plaque forms in kidneys. Kidneys make it.

When plaque is exposed to urine, urine calcium and phosphate ions create a very early phase of phosphate crystal over it, not exactly brushite but a crystal even less fully formed. Calcium oxalate grows over it, perhaps by oxalate pulling calcium out of the phosphate crystals. That has proven hard to study in the laboratory. Could calcium oxalate grow over plaque if brushite SS were well below 1? No one knows.

What About Plugs as a Base for Calcium Oxalate Stones?

Plugs form in the most terminal portions of collecting ducts, in a fluid virtually identical to the final urine. One can find calcium oxalate stones growing over them like with plaque. The details at the growth surface have not been studied as closely as for plaque.

How kidneys make plugs is not certain. I presume brushite forms and converts to hydroxyapatite, but that has not been shown directly. Surely brushite itself is scant, almost never found, but we have not looked for it in extremely microscopic amounts, either, especially in the smallest plugs.

Will tubules plug if brushite SS is kept well below 1? No one knows.

What About Stones Forming In Urine Like in the Reaction Vessels We Just Spent So Much Time On?

Plaque and plugs have taken up a lot of attention this past decade, and crystallization in urine a lot less.

Suppose crystals form in urine, why care? They can just migrate out, so tiny no one knows about them.

But maybe that is wrong. Even tiny crystals could lodge up in the crevices where papilla meet the medulla. Maybe in some people tiny crystals aggregate to make larger masses that will get stuck and grow. We need larger numbers of surgical observations, unselected, to get an idea of what fraction of stones are accounted for by plaque and plugging.

Perhaps quite a few are left unaccounted for. An open question.

Perhaps in those unaccounted for, stones crystals form in urine and grow into clinical menaces.

Effects of Uric Acid and Citrate

Solution Chemistry

Both parts of this graph are like the right hand portion of the figure I already have shown twice. In both experiments, the solution of 1.2 and 1.8 SS brushite and calcium oxalate was let stand until it chose to begin forming crystals.

The pumps and sensors were on throughout.

Uric Acid

If 0.05 mmol/l of uric acid was added to the solution – way too little to form crystals – brushite began forming by about 900 minutes instead of the nearly 1400 minutes without it.

Thereafter, calcium oxalate began to form only 30 minutes after brushite, much faster than without uric acid in the solution.

So uric acid in solution shortens the time needed for brushite to crystallize and also shortens the time needed for calcium oxalate be begin growing over brushite.

As a reasonable estimate human urine contains about 0.5 to 4 mmol/l of uric acid species. Typically there are as a mix of sodium, potassium and ammonium salts with only small amounts of the acid itself. 


At the same concentration, 0.05 mmol/l, citrate prolonged the onset of brushite crystallization to nearly 3000 minutes. Thereafter, brushite grew until halted by calcium oxalate crystallization. The interval between brushite nucleation and the start of calcium oxalate growth was lengthened to about 200-240 minutes. In general human urine contains about 0.5 to 4 mmol/l of citrate.

Appearance of Crystals

Calcium oxalate crystals that form on brushite are small and well separated when viewed by SEM (Panel a). Those grown with uric acid 0.05 mmol/l in the solution (Panel b) are a lot bigger because smaller crystals have aggregated together. This kind of aggregation is seen in human urine according to the authors of this paper.

By contrast, the crystals grown with 0.05 mmol/l of citrate in the solution (Panel c) are small and better dispersed.

Clinical Meaning


Citrate is an established treatment for reduction of calcium oxalate stone recurrence. In general, it is thought to act by inhibiting formation of calcium oxalate nuclei and their subsequent growth.

This work says about the same but shows that citrate acts at both of the steps: It slows nucleation of brushite, and slows formation of calcium oxalate out of the calcium in brushite.

The concentration of citrate needed is far below those in urine. Perhaps that is because this system lacks the many complexities of urine so citrate effects are more visible.

Uric Acid

Long ago, I proposed that abnormally high urine uric acid excretion raised risk of calcium oxalate stones, and that lowering it with the drug allopurinol seemed to reduce new stone formation. Subsequently, Bruce Ettinger published a double blind random allocation prospective trial that showed allopurinol did indeed reduce new calcium oxalate stone production in patients with elevated urine uric acid excretion rates.

But how uric acid might promote calcium oxalate stones was not clear. Our group and others like us thought perhaps crystals of uric acid might act for calcium oxalate like brushite does, but we could not prove it. Likewise for other ideas including adsorption and removal of crystal inhibitors from urine, and “salting out” of calcium oxalate salts by urate salts.

Lacking a clear mechanism, allopurinol has languished as a stone treatment despite having a superb trial to support it. It shows an odd role science plays in medical practice. Physicians, myself included, seem to need an explanation for how a treatment works before we are willing to use it with enthusiasm, even when trial evidence is excellent.

These experiments may have disclosed yet another way uric acid could indeed foster calcium oxalate stones. They suggest a way to explore the matter further, and also suggest we may have under-used allopurinol and done less than full justice to the wonderful trial Bruce Ettinger did for us decades ago.

The Strangeness of Brushite Stones

Here and there we find patients whose stones contain brushite – in part or altogether, and frankly they seem to me a mystery worth solving. How does so vulnerable a crystal survive among such predators as oxalate and hydroxyapatite? What saves its calcium atoms from the oxalate molecules that roam about in the urine? Why does it not lapse into hydroxyapatite?

Something must stabilize brushite, protect it from dissolution. Brushite is not safe in urine or any other solution that has a significant calcium oxalate supersaturation and a pH high enough to permit hydroxyapatite to form.

I will not take this up here, but plan to, one day. It baffles me, astounds me, really. Yet I not rarely care for brushite stone formers, knowing nothing of why they exist.

Risk of Forming Stones Begins at a Brushite SS Above One

Prediction from Basic Science

The present work I have summarized, and the many other demonstrations of brushite as a crystal precursor predict that brushite supersaturation should predict stone forming. In fact, given brushite is like tinder and supersaturation its match, one would expect risk to increase at a SS just above 1.

That appears to be the case.

Empirical Evidence the Prediction is Not False

I have shown this figure elsewhere, and only bring it back here for convenience.

The vertical axes show the relative risk of becoming a stone former among women (red) and men (blue) in three large cohorts followed over many years as part of a long term effort to establish risk factors for a multiplicity of diseases.

Over time, some people began forming stones, and 24 your urines were obtained. As controls, 24 hour urine samples were also obtained from a properly chosen set of well matched people who did not form stones.

Daily excretion of calcium, oxalate, citrate, and water – urine volume, were predictive of stone risk. In all four cases, risk rose significantly as the excretion rate rose (calcium and oxalate) or fell (Citrate and water – urine volume).

Risk rose with SS for calcium oxalate as well, as one might expect. But the units here are odd. They represent the actual SS divided by the mean SS among a group of about 50 normal people (SS mean = ~3). So a value of 1 is about a SS of 3 fold and the first risk group is therefore SS from 3 to 6.

For SS with brushite (CaP) SS values are not so corrected, and risk begins as SS equals or exceeds 1. This is shown by the height of the bars that register the mean relative risk, and by the fact that in 2 of the 3 cohorts the lower 95th percentile for relative risk lies about 1.

Reservations and Science

I have shown you basic science pointing to a role of brushite to initiate the common stone, and to empirical science in the form of epidemiology supporting brushite as a risk factor for onset of new stones.

Urine Chemistry is Far More Complex

The complex environment of urine that contains about1800 peptides, many of them affecting crystals, may mute the effects of brushite or even obliterate them. It is not possible to measure in urine what I have shown you here. For example in whole urine, crystals induced by adding excess oxalate are not affected by seeds of brushite. One might note that initiating crystallization by adding oxalate is almost certain to complicate effects of phosphate crystals, as it creates by definition sufficient supersaturation to initiate calcium oxalate crystals de novo.

The Miscreant Crystal May be Another Phosphate Phase

The Nancollas group has reported that amorphous calcium phosphate phases may be critical as opposed to brushite, citing that urine may not be supersaturated with brushite as a general rule. This latter is contradicted by the epidemiology findings just above. The relative risk of stone forming with higher CaP supersaturation arises from higher values in those who became stone formers.

However, the way we calculate brushite supersaturation may not include all possible kinds of calcium phosphate phases, so our measurements could be imperfect. They could over estimate “true” brushite supersaturation.

Medicine Requires Action Despite Uncertainty

I cannot resolve these matters, but that does not affect my main point. Time and more experiments will eventually sort out the details.

Right now clinicians do not have measurements for supersaturation of amorphous calcium phosphates, so whatever they do cannot matter in clinical practice. What we do measure is a risk factor for stones, and what basic science we have points to brushite as a good source of trouble.

As I have said, no risk, some gain, suffices.


I want to thank Dr John Asplin for his careful reading of this article. The material is very complex and it is reassuring to have someone as expert as he as a reviewer. I have incorporated his additions.

Professor George Nancollas

George Nancollas invented the constant composition technique that drove this research. I knew him well, and heard portions of this research long before the final publication.

George was a distinguished chemist utterly devoted to the problems of biological crystallization.  

At the ROCK society, our small group of like minded scientists interested in stone disease, George was a favored guest, brilliant speaker and discussant, and lovely person as well. Howsoever lacking a sophistication to match his, we were nevertheless politely welcomed to chat with him about what he did and knew, and felt – so long as he was around – somehow smarter about the whole matter.

He was aware of the brushite/calcium oxalate issue and also of the uric acid connection to calcium oxalate stones. It was because of clinical investigators like myself and others that he did the uric acid portion of the present work.

He died Jan 11 2016. His 420 publications remain for any to use that have wit and desire enough, and a taste for their ascetic beauty.

Ashes denote that Fire was —
Revere the Grayest Pile
For the Departed Creature’s sake
That hovered there awhile —

Fire exists the first in light
And then consolidates
Only the Chemist can disclose
Into what Carbonates.

Emily Dickinson

Scroll to Top