Rebecca Morelle, BBC News 9 Jun 09;
An unmistakeable hiss fills the air.
I peer down, and spot a plastic container.
The source of the noise scuttles into view - a cockroach. In fact, the box is filled to the brim with them, with a few other writhing and wriggling creepy-crawlies thrown in for good measure.
"Dinner," explains my host, pointing to the dozens of very large, rather hairy and extremely leggy tarantulas eyeing this bug banquet.
Welcome to the Spiderlab.
Tangled web
Sara Goodacre, who is showing me around her unusual laboratory based at Nottingham University, is fascinated by arachnids.
"They are just so neat," she says.
And her team is especially interested in tarantulas, thanks to the incredible webs that they spin.
Because spider silk is so strong - stronger even than steel, lightweight and very, very stretchy - researchers envisage a huge host of commercial applications, ranging from medical sutures to clothing.
But to achieve this means having access to large amounts of the material.
And unlike silkworms, spiders do not make good factory workers - thanks in part to their tendency to eat each other if kept in close confines. So the ultimate aim is to find a way of making artificial silk.
Much research so far has focussed on orb weavers - spiders that belong to the Araneidae family, which includes some of the species that you might find at the bottom of your garden - mainly because the silk that they produce is especially tough and elastic.
But although scientists have been able to decode some of the orb weavers' genes that are responsible for silk and even create genetically modified goats that produce the silk protein in their milk, creating a useable material has proved very tricky.
So the Nottingham University researchers are trying a different tactic.
Dr Goodacre says: "Tarantula silk really is one of the unexplored areas - we don't know what it is made of or how useful it might be to us.
"Nobody has really looked at these spiders. It might just be that the silk is very, very different from that in the other groups studied this far.
"But we just don't know."
There are about 900 species of tarantula (most belong to the Theraphosidae grouping), and they inhabit a very different position on the arachnid family tree compared with the well-studied orb weavers.
Dr Goodacre explains: "It is of the order of several hundred million years ago that tarantulas branched off from the other spiders.
"And they have been separate since then - so [in terms of the silk] evolution has had a fair amount of time to produce something different."
And not only is their silk unlike that of other spiders, there also appears to be a lot of variation amongst the silk types produced by the many different tarantula species.
So in the Spiderlab, goliath bird eaters - the biggest of all the tarantulas, the highly aggressive Tanzanian orange baboon tarantula and the more docile Chilean rose tarantula, which can be kept as a pet, can be found sitting alongside each other - albeit separated by glass, to avoid any spider-eat-spider incidents.
There are about 30 of them, although that number might soon rise after we witnessed a pair mating: two sets of eight legs entwining to become 16, their bodies merging into a weird spidery mass, until the male makes a very, very quick exit, nearly running out of the tank.
There's clearly no spider pillow talk for a tarantula trying to avoid the fangs of a ferocious female.
Silk varieties
By looking at these different species, Dr Goodacre is aiming to find out why one type of spider has managed to produce so many varieties of silk.
She says: "We have some species that are ground dwellers that lay silk above the ground, others actually dig down and build a silken burrow, whereas others climb up trees to make their webs.
"What we are trying to understand is whether they are all using the same kind of silk and do they use it in the same way?"
The first task for the team has been to try to uncover the genes that might be involved in making the silk.
So far, PhD student Jon Bull has isolated genetic material from the genes that are expressed in tarantula spinnerets - the body parts that turn the liquid silk molecules into a fibre.
The next step is to find out exactly what these genes are, how many there are, how they are linked to one another and what they make.
Dr Goodacre explains: "We know a little bit about the genes for silk in orb weaver spiders, but at the moment we know virtually nothing about the genes that make silks in anything else.
"The jury is really out on whether the silk genes that you find in a tarantula are really very similar to those you find in an orb weaver, given that the end product looks quite different."
Coming unstuck
Even if the team is able to uncover the genetic secrets of tarantula silk, learn about its complex chemical properties, and even go as far as synthesising their own silk protein - they will still face a major challenge that has slowed or even halted many other artificial silk projects: spinning it into a useful product.
Spiders pump silk protein through their spinerettes, transforming it from a messy solution to the thin fibre that makes up their intricate webs.
Dr Goodacre says: "It is analogous to having wool on a sheep's coat and making it into a yarn - you have to know how to spin it properly."
The spinerettes are able to carefully control the acidity, pressure and concentration of the silk protein, to allow a spider to spin its web with such apparent ease.
But replicating this complex process in the lab is proving far from easy.
It seems that nature's solution is not quite a simple as it first appears - and the days of being able to create vast amounts of this useful product may be some way off.
Dr Goodacre admits the team is at the beginning of a long road of research, but is still confident that tarantulas could hold the key to unravelling some of the secrets of spider silk.
She says: "If you are trying to predict what would be a good silk to make, rather than think of all of the alternatives you can make yourself, why not just look at what other spiders like tarantulas already do?
"Several million years of evolution is likely to have generated some pretty good solutions."
