Today we talk about… mating! Whoa -- did you just see that spike in the stats page? Haha, okay, but first you have to sit through the usual genetic lesson. Here it goes: we start talking about the major histocompatibility complex, or MHC.
I've mentioned many times in my previous posts that in order to trigger an immune response you have to make sure that the immune system recognizes the antigen, or "foreign" object. Antigens are made of proteins that, once inside the cell, are chopped into bits and pieces. The bits and pieces are then transported to the cell surface and "presented" to T-lymphocytes, which in turn recognize the bits as a "red flag" (as in, "ALARM! The cell has been infected with foreign and dangerous object!"), and destroy the cell. MHC molecules have the function of grabbing the bits of proteins inside the cell and presenting them to the cell surface. This is a very important step in the immune response, because without this "presentation" the immune system is unable to recognize the antigen.
Now, the genes that code the MHC molecules are highly polymorphic: what this means is that it's a DNA region that varies greatly across individuals. Why? Because the broader range of MHC molecules we have, the better chances to recognize even the rarest antigen that invades our body. Natural selection favors variety in the MHC genes. Having different alleles for these genes is beneficial for disease resistance. Basically, it makes our immune system stronger.
How do we get different alleles? Remember, we inherit one copy of each gene from our mother, and one from our father, and we get discordant alleles when the mother's copy is different from the father's. Studies have shown that mating occurs preferentially between MHC discordant alleles . And how do we detect people with different MHC alleles than ours? By sniffing. Seriously.
This is actually an old study (1995), and many of you may already know about it. Wedekind et al.  enrolled a sample of male and female students. They had the male participants wear the same T-shirt for two consecutive nights, and then they asked the ladies to rate the "odor pleasantness" of six T-shirts, chosen so that three came from MHC-similar males, and three from MHC-dissimilar males. A curious trivia is that the women in the study used nasal sprays to enhance their sense of smell, and they had all read Suskind's novel "The Perfume." Interestingly, the researchers found a correlation between odor preferences and discordant MHC.
Let's look at the genetics behind the scenes. The genes that regulate olfactory receptors are all over our genome. However, Ehlers et al.  found a very large cluster (36 genes) very close to the HLA complex on chromosome 6. This entire region is in strong linkage disequilibrium, which is a very complicated way geneticists use to say that basically we tend to inherit these genes together. You know how chromosomes split before they turn into oocytes or spermatocytes? Well, the splitting is not completely random, and it turns out that these olfactory receptor genes are highly correlated to the HLA genes (they tend to "stick" together), which would explain how odor preferences would correlate to discordant MHC types.
Wow, I've managed to weave a link through the first three items in my schizophrenic title. Now to the fourth one: birth control pills. Well, the Wedekind study found that the correlation varied depending on the women's hormonal status and, furthermore, the trend reversed for women on the pill. In other words, women on the pill seemed to prefer the T-shirts from MHC-similar men.
Okay, all of the above made a fantastic punchline, but… how about the caveats?
For starters, Roberts et al.  in 2008 repeated the exact same experiment Wedekind et al. did and, alas, couldn't reproduce the results. What they did find was that single women seemed to prefer the odor from MHC-similar men, while women in a relationship preferred odors form MHC-dissimilar men.
By the time I came to the end of the paper, they had done so many tests that any p-value they found would have to be taken with a grain of salt. And in all fairness, what they propose to measure in these studies are variables extremely hard to quantify. The likeability of a certain odor varies not only from person to person, but from day to day. Have you ever shopped for a fragrance? After spraying a few testers, don't they all smell the same?
As for the biology caveats (and for this part I have to thank my dad!):
1) In mammals sexual stimuli are regulated through pheromones. While generic smells are perceived through the olfactory receptors in the epithelium inside the nasal cavity, pheromones are sensed through the vomeronasal organ. The two are regulated by different genes.
2) In humans the vomeronasal organ has lost its function, mostly because it got replaced by the fact that we can see colors. This has caused a shift: most sexual stimuli in humans are perceived through sight, and the genes regulating the vomeronasal organ have become pseudogenes (non-coding).
3) Lastly, how we react to odors is not simply a genetic behavior, but it is highly correlated to the environment. In fact, it changes throughout our lifetime as we "learn" to like certain smells more than others, and this is due to the way our brain changes and reacts to the environment.
Nonetheless, these are certainly interesting experiments as they point to "trends" in human behavior. They give some insights on how much we are driven by hormonal changes, and on the complex ways physiology and environment weave together into making who we are.
 Milinski, M. (2006). The Major Histocompatibility Complex, Sexual Selection, and Mate Choice Annual Review of Ecology, Evolution, and Systematics, 37 (1), 159-186 DOI: 10.1146/annurev.ecolsys.37.091305.110242
 Wedekind C, Seebeck T, Bettens F, & Paepke AJ (1995). MHC-dependent mate preferences in humans. Proceedings. Biological sciences / The Royal Society, 260 (1359), 245-9 PMID: 7630893
 Ehlers A, Beck S, Forbes SA, Trowsdale J, Volz A, Younger R, & Ziegler A (2000). MHC-linked olfactory receptor loci exhibit polymorphism and contribute to extended HLA/OR-haplotypes. Genome research, 10 (12), 1968-78 PMID: 11116091
 Roberts SC, Gosling LM, Carter V, & Petrie M (2008). MHC-correlated odour preferences in humans and the use of oral contraceptives. Proceedings. Biological sciences / The Royal Society, 275 (1652), 2715-22 PMID: 18700206