As many estimable colleagues have noted, guanine's weird. It's not that soluble, which isn't too much of a problem when it's a part of your very soluble DNA, but from an origin-of-life perspective, that's a hassle. Envisioning a primordial-soup scenario with guanine is tricky, because of this low solubility. Further complicating the problem is guanine's prodigous tendency to associate with itself:

Monomers of G will do this, forming gels in solution (which exhibit repeating units of the above structure). We knew about guanine's propensity to self-associate before we even had cracked the structure of DNA.

So why do we care about ancient history? Well, you have a bunch of repeats of the sequence d(TTAGGG) at the end of your genome. Your DNA polymerases can't copy all the way to the end of your gene, so you lose some of this every time your cells divide. Once your cells have divided a certain number of times, you run out of this TTAGGG "telomere" sequence, and they die. More likely, you and I will be hit by buses or have heart attacks well before then. If you get to the age when you're worried about running out of telomere, count yourself lucky!

In solution (and people still fret about whether this is the case in the cell, but it looks pretty likely), telomeres fold into the G-tetrad/G-quadruplex structure above.

Let's go back fifty years or so and talk about a young woman named Henrietta Lacks (known for awhile as Helen Lane for reasons of anonymity). She was unfortunate enough to contract cervical cancer. Her cervical cancer cells, as well as many other cancers, express an enzyme called telomerase. These copy the telomere "tails" on DNA, negating the Hayflick-type senescence (cell death) associated with telomere shortening.

Telomerase can't extend telomeres if your telomeres are folded into quadruplexes. We have drugs that bind Watson-Crick DNA by virtue (in part) of being about the size of a W-C base pair, like ethidium:

Ethidium binds and stabilizes W-C duplexes. A small molecule drug of about the same size as a G-tetrad can do the same. Enter telomestatin:

Telomestatin is believed to inhibit telomerase by binding to G-quadruplex forming sequences (i.e., telomeres), inhibiting replication, and inducing Hayflick-type senescence of cancer cells. Your normal cells, not expressing telomerase, are not as affected. Telomerase is an interestingly selective target and the subject of loads of research. It is my favorite cancer target.

Note that if you're a biologist, you probably have worked with some of the young Ms. Lacks' cells. You know them as HeLa cells. Her cells were actually taken without her consent and commercialized - the issue was addressed in John Moore v. Regents of the University of California, which is one of the seminal "do your cells belong to you" cases. Think of Henrietta next time you grow up a HeLa culture!