Pheromone transport and release
Organisms have developed a range of strategies for the release of pheromones. The highly volatile pheromones utilised by Lepidoptera are usually dispersed by wing fanning. Some species of insects have specialised glands (e.g. hair pencils) with a large surface area, which can be inverted to present the inner secretary surface to the outside world. In fish, free steroids are released from the gills which have a large surface area washed by a continuous flow of water. The release of steroids in comparatively small volumes of urine or through the skin, requires that they be much more soluble and so they are conjugated as sulfates or glucuronides.
In mammals, evidence is emerging that specialised proteins are used to transport odorous compounds. The first indications came from work on the Syrian golden hamster, which is unusually dependant on olfaction for normal sexual behaviour 73. The night before oestrus the female hamster lays down a trail to her underground burrow using a watery vaginal secretion which attracts the male. This vaginal deposit contains more than 200 compounds, but the major pheromonal activity is apparently due to dimethyl disulfide 74 75, These results have been called into question by later studies which showed that although males investigate vaginal odours more frequently than females, there is no gender difference in the investigation of dimethyl disulfide 76. Moreover dimethyl disulfide is extremely common in nature; it deters feeding on brassicas by sheep 77, acts as a nipple attachment pheromone in pup rats 78 and is the principal malodorant in human tooth disease.
(Figures reprinted from previous page for ease of reference)
Whatever lures the male golden hamster to the females nuptial burrow, the later stages of courtship and copulation are induced by a protein (aphrodisin 79) found on the genital region of the female, which is licked by the male and detected by the vomeronasal organ 80. The behaviour is sufficiently stereotypical that an extract painted onto the hind quarters of an anaesthetised male will induce mounting attempts by another male.81. Aphrodisin 82 is a member of the lipocalin (lipocalycin) family of 20KDa, soluble extracellular proteins 83. This is a diverse family which includes odorant binding proteins 84 (OBP; in the nasal mucus), major urinary protein (MUP), retinol and retinoic acid binding proteins, b-lactogloblin, a-1-microglobulin and quiescence specific protein 85.
Another indication that lipocalins are involved in the function of pheromone-like compounds came from the finding that the human armpit odour component, trans-3-methyl-2-hexenoic acid (16) is bound to two proteins (25 and 46KDa) which appear to be involved in transportation from the site of biosynthesis 86. The smaller of these is apolipoprotein D, but with a different glycosylation pattern to plasma apolipoprotein D 87. Both of these proteins are also lipocalins. There is no known role for plasma apolipoprotein D, although it binds progesterone and pregnelone in vivo and there are indications that it might bind bilirubin and cholesteryl esters in vivo. The boar pheromones, androstenol 13 and androstenone 14 are released from a 20 KDa carrier protein (pheromaxein) which is produced in the submaxillary glands 88. Very little is known about the structure, but the molecular weight is consistent with a lipocalin 89.
Apparently the lipocalins are transport proteins for small hydrophobic molecules, although in many cases the ligand has not been identified. No enzymatic activity has been demonstrated for any lipocalin, except for the facile isomerisation of PGH2 by PGD synthase to PGD2 which acts as a sleep promoter 90. The structures have low sequence homology, but essentially identical 3-dimensional structures, in which an a-helix and two orthogonal b-sheets form a b-barrel with a hydrophobic interior, which is the putative binding pocket 91. At one end of the b-barrel there is a three residue invariant sequence located at the start of the A b-strand and another at the end of the F b-strand. These may be part of a receptor for common cell surface receptors, whereas plausibly the other, more variable, end of the b-barrel binds the ligand 92. A lipocalin with 31% homology to apolipoprotein D is expressed by E. coli in the stationary growth phase 93. It is tempting to suggest that its role, is to bind ligands which signal that conditions have improved sufficient for growth and division to resume.
(Figure reprinted from previous page for ease of reference)
Dehydro-exo-brevicomin (8) and sec-butylthiazoline (9) are both selectively bound by mouse major urinary protein (MUP), which is expressed in the liver 94. The X-ray crystal structure has been determined and shows a typical lipocalin structure 95. MUPs differ between different strains of mice, but the variations are generally conservative (eg Lys to Gln or Glu) 96. N-Terminal sequences of mouse OBPs are very similar to those of the MUPs 97, which is significant because lipocalins usually have low sequence homology. The idea that similar if not identical proteins are involved in pheromone release and detection is appealing on the grounds of molecular economy 98. However this story has a further twist. It was noted earlier that when young female mice are housed together away from adults, development into puberty is retarded. This effect cannot be overcome by dehydro-exo-brevicomin (8) or sec-butylthiazoline (9), however MUP alone or the hexapeptide, N-Glu-Glu-Ala-Arg-Ser-Met (which is a truncated MUP) accelerates puberty as assessed by the weight of the uterus 99. This is a startling result wholly at odds with the transport protein rationale. Moreover it should be noted that despite over 10 years of work with aphrodisin no ligand has been identified. This raises the question: what role if any does dehydro-exo-brevicomin (8) or sec-butylthiazoline (9) play in this system; is the volatile ligand an attractant for the involatile lipocalin; in other words a pheromone's pheromone?!
The use of common compounds for signalling between organisms raises crucial questions about selectivity. In some cases there is none. Cross species mating attempts have been observed between Arctiidae (Tiger and Footman moths) and between other moths 100 but presumably not (successfully) between moths and elephants! In other species slight shifts in the composition of a blend of pheromones result in large differences in attraction 101. All that can be said with certainty is that this will be a fertile area for further investigation.