s=spicule. Fig. 1e from Imsiecke et al. 1995. Click image for source.

Let's say you're a small cell engaged in heavy manufacturing. Like most animal cells, you are coated only in a thin membrane made a double layer of fluid fat-like molecules. The thing you make is a giant, pointy glass rod twice your size. Would you expect to survive this process? Well, if you're a cell called a sclerocyte, yes you can.

Last month I wrote about the astounding skeletal architecture of rock sponges. In the course of writing that post I discovered that the special cells that make sponge spicules – the siliceous building blocks of the sponge skeleton -- have some pretty amazing tricks up their sleeve for assembling the bewildering array of forms.

Scientists were curious about the methods of these cells, called sclerocytes. Do they do the work of crafting spicules solo, or do they require “helper” cells to assist them with their momentous task? And what happens to them once they produce a finished and intimidatingly spiky spicule? Do they fall on their own spines?

At least one previous scientist had assumed they simply died. But this much was known: that sclerocytes build a scaffolding known as the axial filament, and then grow the spicule by depositing silicon onto this filament, thickening it and elongating it.

To find out what happens, back in 1995 a team of scientists from Brazil coaxed gemmules -- asexual reproductive structures by which this sponge survives winter – of the freshwater sponge Ephydatia muelleri to hatch in culture (until a few years ago, I had no idea there was such a thing as a freshwater sponge. There's a picture of one at the end of my old post "Sponges: The Original Animal House").

They then "mechanically dissociated" these structures into individual cells, which then reaggregated (as sponges are famous for doing) within 24 hours into bunches of 20 to 4000 cells. These five day old bunches of cells contained a few sclerocytes, already identifiable by the presence of an axial filament or growing spicules. They were released at day 22 when the entire aggregate of cells wandered away. Left behind by the busy beaver-like sclerocytes was a pile of spicules glued to the floor of the cells' growth chamber, but not to each other as they would be in a real sponge skeleton.

Sponge cell aggregate, left, and pile of spicules they left behind glued to the floor of their container. Fig. 1 b and c from Imsiecke et al. 1995. Click image for source.

Meanwhile, a few sclerocytes had wandered off on their own. They also made spicules, but didn't even start until 30 days in culture had passed.

The scientists could clearly see the organic axial filament form in culture, just as it does in wild sponges.

n=nucleus, af= axial filament. Fig. 3b from Imsiecke et al. 1995. Click image for source.

The axial filament is an organic structure whose composition I still have not been able to discover, other than in a paper from 1972 that concludes a "substantial" portion of the filament is made of protein, but may also be partly made of carbohydrates.

As silicon was pasted on to this filament, the spicule grew by about one to ten micrometers per hour, the rate seen in nature. Even as its glass shell grew, the filament was still visible in electron micrograph cross sections after removing the silicon with uber-nasty, glass-dissolving hydrofluoric acid. In the image below, the location of the spicule is indicated by the empty white space marked "s". Interestingly, the axial filament is triangular in cross section, and in this case, off center.

af= axial filament, which has a diameter of 200-500 nanometers. s=location of former spicule. cb= collagen bundle, P=pinacocyte. Fig. 3a from Imsiecke et al. 1995. Click image for source.

As the spicules grew, they began to "protrude" the cell's membrane, a bit like a baby bump, were the baby an enormous, double-sided spike. Let's not consider that analogy too carefully.


s=spicule, f=filopodium. Fig. 1f from Imsiecke et al. 1995. Click image for source.

Consider further that the spicules made in nature are over twice the size of those made in culture: 350 micrometers in the wild compared to 150 micrometers in the lab. Ouch.

Like a baby, they remained covered by cell membrane until they reached full size. In about 5% of the cells, a second spicule began forming while the first was still nearby.

s1=fully developed spicule, just released s2=newly forming spicule, n=nucleus. Fig. 1g from Imsiecke et al. 1995. Click image for source.

This clearly shows that sclerocytes not only survive their ordeal, they go on to make more than one spicule -- sometimes way more than one.

The scientists would often find bunches of 8-10 identical spicules, like a pile of cute little sponge presents. It appears as if a single cell just kept making and depositing them in the same place and then abandoned them all when the herd moved on.

Fig. 4 f and g from Imsiecke et al. 1995. Click image for source.

In a charming hint that even a single tiny cell living as part of an animal can have a quirky "personality", some of these piles had identically irregular spicules all made with a weird bump in roughly the same place, as you can see in the image above right.

Sometimes, the cells made spicules with two arms. The scientists interpreted these as errors in axial filament formation, where a second filament was begun while the first was still under construction and accidentally got stuck to the first.

Fig. 4 d and e from Imsiecke et al. 1995. Click image for source.

At right are spicules where for some reason, the cell glued nearly all of the silicon to one spot in the center of the axial filament, creating a spindle-like structure.

The axial filament was visible both inside the cell prior to silcon molding and encased in spicules after release, here in an irregularly shaped spicule next to a normally-shaped one.

af= axial filament visible in malformed spicule. Fig. 4a from Imsiecke et al. 1995. Click image for source.

When the cells were grown in cultures that lacked any silicon, all they could make was an axial filament about half as long as a finished laboratory-sized spicule, which they went on to release just the same. Sad cells.

S=sclerocyte, af= axial filament. Fig. 5 from Imsiecke et al. 1995. Click image for source.

The scientists learned another interesting thing from these experiments. Since lone wolf sclerocytes were still capable of making spicules, the scientists deduced that no "helper cells" are necessary. All the information and resources necessary to construct a spicule, they inferred, is contained in a single cell.

The authors also offered a fascinating speculation regarding spicule formation. Many of you may be familiar with the famous Hox (or homeotic) genes that control the development of most animal embryos, telling the developing embryo which way is up and down, left and right, and where to put the arms, legs, eyes, and every major anatomical feature. Sponges, it turns out, have similar homeotic genes. But sponge bodies have no symmetry, organs, or major body parts whatsoever. What if, the scientists suggest, sponge homeotic genes are used to guide the formation of the blizzard of intricate and precise spicule forms?

Demospongiae spicule diversity.png

"Demospongiae spicule diversity" by Rob W. M. Van Soest, Nicole Boury-Esnault, Jean Vacelet, Martin Dohrmann, Dirk Erpenbeck, Nicole J. De Voogd, Nadiezhda Santodomingo, Bart Vanhoorne, Michelle Kelly, John N. A. Hooper - Van Soest RWM, Boury-Esnault N, Vacelet J, Dohrmann M, Erpenbeck D, et al. (2012) Global Diversity of Sponges (Porifera). PLoS ONE 7(4): e35105. doi:10.1371/journal.pone.0035105. Licensed under CC BY 2.5 via Wikimedia Commons.

That's a big what-if, and a subject for a future study, they say.

That also leaves me with a few still-unanswered questions. How do spicules that aren't just straight rods form? Is the axial filament formed in the intricate shapes they will eventually become, or is the intricate silicon deposition of more complex spicules guided in some other way? What prevents the giant death spike of a spicule from rupturing the cell membrane? Is the cell skeleton of microfilaments, intermediate filaments, and microtubules somehow involved? Chime in if you know!


Imsiecke G., Marcio Custodio, Radovan Borojevic & Werner E. G. Müller (1995). Formation of spicules by sclerocytes from the freshwater spongeEphydatia muelleri in short-term cultures in vitro, In Vitro Cellular Developmental Biology, 31 (7) 528-535. DOI: http://dx.doi.org/10.1007/bf02634030