Richard E. Humphries

Unravelling the chemical basis of competitive scent marking in house mice

Major urinary proteins (MUPs) in the urine of male house mice, Mus domesticus, bind the male signalling volatiles 2- sec -butyl-4,5-dihydrothiazole (thiazole) and 3,4-dehydro- exo -brevicomin (brevicomin) and slowly release these volatiles from urinary scent marks. To examine the role of urinary proteins and volatiles, either attached or unattached to the proteins, in competitive scent marking, we fractionated urine from isolated male BALB/c laboratory mice, Mus musculus, by size-exclusion chromatography into three pools. Pool I contained all of the urinary proteins and their bound ligands while pools II and III contained lower molecular weight components including unbound signalling volatiles. In experiment 1, pools I-III were streaked out on to absorbent paper (Benchkote) and introduced into enclosures housing single wild-caught male mice, together with a clean control surface. Each male was tested with fresh stimuli and with aged stimuli deposited 24 h previously. Only pool I stimulated significantly more countermarking and investigation than the control, attracting mice to investigate from a distance even when the rate of ligand release was considerably reduced after 24 h. Experiment 2 examined responses to pool I when this was fresh, aged by 7 days, or had been mixed with menadione to displace ligands from the proteins. Although all three protein stimuli were investigated and countermarked more than a clean control, the aged and menadione-treated pool I stimulated the strongest responses, despite containing the lowest levels of thiazole and brevicomin. Thus competitive countermarking is stimulated by proteins or by nonvolatile protein-ligand complexes in male urine, while release of volatile ligands attracts attention to a competitors scent marks. Copyright 1999 The Association for the Study of Animal Behaviour.

The ownership signature in mouse scent marks is involatile.

Male house mice advertise their territory ownership through urinary scent marks and use individual–specific patterns of major urinary proteins (MUPs) to discriminate between their own scent and that of other males. It is not clear whether recognition occurs through discrimination of the non–volatile proteins or protein–ligand complexes (direct model), or by the detection of volatile ligands that are released from MUPs (indirect model). To examine the mechanism underlying individual scent mark signatures, we compared investigatory and countermarking responses of male laboratory mice presented with male scent marks from a strain with a different MUP pattern, when they could contact the scent or when contact was prevented by a porous nitrocellulose sheet to which proteins bind. Mice investigated scent marks from other males whether these were covered or not, and biochemical analysis confirmed that the porous cover did not prevent the release of volatiles from scent marks. Having gained information through investigation, mice increased their own scent marking only if they had direct contact with another males urine, failing to do this when contact was prevented. Individual signatures in scent marks thus appear to be carried by non–volatile proteins or by non–volatile protein–ligand complexes, rather than by volatiles emanating from the scent.

MHC odours are not required or sufficient for recognition of individual scent owners.

To provide information about specific depositors, scent marks need to encode a stable signal of individual ownership. The highly polymorphic major histocompatibility complex (MHC) influences scents and contributes to the recognition of close kin and avoidance of inbreeding when MHC haplotypes are shared. MHC diversity between individuals has also been proposed as a primary source of scents used in individual recognition. We tested this in the context of scent owner recognition among male mice, which scent mark their territories and countermark scents from other males. We examined responses towards urine scent according to the scent owners genetic difference to the territory owner (MHC, genetic background, both and neither) or genetic match to a familiar neighbour. While urine of a different genetic background from the subject always stimulated greater scent marking than own, regardless of familiarity, MHC-associated odours were neither necessary nor sufficient for scent owner recognition and failed to stimulate countermarking. Urine of a different MHC type to the subject stimulated increased investigation only when this matched both the MHC and genetic background of a familiar neighbour. We propose an associative model of scent owner recognition in which volatile scent profiles, contributed by both fixed genetic and varying non-genetic factors, are learnt in association with a stable involatile ownership signal provided by other highly polymorphic urine components.

Polymorphism in major urinary proteins: molecular heterogeneity in a wild mouse population.

Major urinary proteins (MUPs) are present in high levels in the urine of mice, and the specific profile of MUPs varies considerably among wild-caught individuals. We have conducted a detailed study of the polymorphic variation within a geographically constrained island population, analyzing the MUP heterogeneity by isoelectric focusing and analytical ion exchange chromatography. Several MUPs were purified in sufficient quantities for analysis by electrospray ionization mass spectrometry and MALDI-TOF mass spectrometry of endopeptidase Lys-C peptide maps. The results of such analyses permitted the identification of three new MUP allelic variants. In each of these proteins, the sites of variation were located to a restricted segment of the polypeptide chain, projecting to a patch on the surface of the protein, and connected to the central lipocalin calyx through the polypeptide backbone. The restriction of the polymorphic variation to one segment of the polypeptide may be of functional significance, either in the modulation of ligand release or in communication of individuality signals within urinary scent marks.

Urinary Lipocalins in Rodenta:is there a Generic Model?

It is increasingly clear that mediation of chemical signals is not the exclusive domain of low molecular volatile or water soluble metabolites. Pheromone binding proteins play an important role in mediating the activity of low molecular weight compounds, while proteins and peptides can also act as information molecules in their own right. Understanding of the role played by proteins in scents has been derived largely from the study of Major Urinary Proteins (MUPs) in the mouse (Mus musculus domesticus) and the rat (Rattus norvegicus). As part of an ongoing programme to explore the diversity and complexity of urinary proteins in rodents, we have applied a proteomics-based approach to the analysis of urinary proteins from a wider range of rodents. These data suggest that many species express proteins in their urine that are structurally similar to the MUPs, although there is considerable diversity in concentration, in sexual dimorphism and in polymorphic complexity. This is likely to reflect a high degree of species-specificity in communication and the information that these proteins provide in scent signals.

Heterogeneity of Major Urinary Proteins in House Mice: Population and Sex Differences

House mice (Mus domesticus) and rats (Rattus norvegicus) secrete large quantities of protein into their urine as the normal condition (Finlayson and Baumann, 1958). These proteins are major urinary proteins (MUP) and alpha2u-globulins respectively. Both are small proteins that have been assigned to the lipocalin protein family on the basis of sequence data (reviewed by Flower, 1996) and X-ray crystallography (Bocskei et al., 1992). As such, both display a typical lipocalin beta-barrel structure enclosing an internal cavity in which volatile ligands may bind.

Information in Scent Signals of Competitive Social Status: The Interface Between Behaviour and Chemistry

From an evolutionary viewpoint, signals generally should be reliable or honest (Zahavi, 1987; Johnstone, 1997). Animals can gain a number of advantages from advertising high competitive ability to potential mates and to other competitors, particularly males which often compete strongly for mating opportunities (Andersson, 1994). Animals often prefer high quality mates that will increase the fitness of their offspring, both because of genetic benefits (through good genes or Fisherian selection) and because parents of high competitive ability often provide better resources and protection. Competitors will also gain an advantage if potential challengers withdraw from, or otherwise avoid, aggressive encounters with an opponent of high fighting ability. There is thus strong selection pressure on signallers to advertise high social status and competitive ability to others. However, receivers will only gain an advantage from responding to such signals if these are reliable indicators of the signaller’s competitive ability. Females that mate with low quality males that dishonestly signal high competitive ability will gain no advantage for their offspring, while males that withdraw from agonistic encounters with poorer competitors will be disadvantaged. There is thus strong selection on receivers to respond only to honest signals that are resistant to cheating, and therefore for high quality animals to provide such reliable information in signals as these will be effective in attracting mates and deterring competitors.

Mice, Mups and Myths: Structure-Function Relationships of the Major Urinary Proteins

The presence of high concentrations of protein in the urine of many mammals is often a pathological event reflecting a failure of the renal mechanisms to prevent passage of plasma proteins through the glomerular filter, or for recovery of protein in the proximal tubule. In adult humans, for example, the normal urinary protein concentration is of the order of 50μg/ml, a value that serves to emphasise the remarkable protein output in the urine of mice, and to a lesser extent rats. The average protein concentration in the house mouse (Mus domesticus) reaches between 200 and 2000 times that of a normal human, at concentrations of the order of 30mg/ml. Over 99% of this protein comprises the Major Urinary Proteins, a group of 18–20kDa proteins, synthesised in the liver, secreted into plasma and subsequently passed through the glomerular filter into the urine as it is elaborated.

The “scents” of ownership

Scent marks need to encode a reliable signal of ownership to inform about the specific owner. Although tests of discrimination have told us much about the abilities of animals to detect differences in conspecific scents deriving from a wide range of sources, we know little about the components involved in scent ownership recognition or individual recognition when animals meet because this requires functional tests of individual recognition. Ownership signals in scent marks need to be stable and persistent, ideally genetically determined and sufficiently polymorphic. Recent work from our laboratory, using functional tests of scent mark recognition, suggest that the pattern of polymorphic MUPs in the urinary scent marks of male house mice provides an ownership signal. The ownership signal is involatile, requiring investigatory contact with the scent source, and involves either involatile complexes between MUPs and their bound odorants or the MUPs themselves, probably detected through the vomeronasal system. However, mice also detect non-MUP related differences in urinary volatiles. We propose a model of learnt association between involatile and volatile components that would allow mice to recognize previously encountered volatile profiles from familiar individuals or animals of the same sex without requiring close contact investigation. Investigation of fresh scent marks deposited around the environment would allow animals to update the proposed association between an individual’s stable involatile profile with any changes in its volatile profile. Further research is required to test this model and to establish its generality in other mammalian species.

The Role of Urinary Proteins and Volatiles in Competitive Scent Marking Among Male House Mice

Male house mice (Mus domesticus), like many other male mammals, advertise their competitive dominance and ability to defend territories by depositing numerous urinary scent marks throughout their territory (reviewed by Rails, 1971; Johnson, 1973; Gosling, 1982, 1990; Hurst, 1987). Male mice also increase their rate of marking near any competing scent marks from other males, a behaviour termed counter-marking (Hurst, 1990, 1993; Hurst and Rich, 1999). Because only those males successfully dominating their territory can ensure that their own marks are always the freshest and predominant in the area, other males can use the temporal and spatial deposition dynamics of male scent marks to assess territory ownership and competitive ability (see Hurst et al., this volume). Perhaps more importantly, female mice can also use these scent marks to assess the quality of potential mates, preferring dominant male territory owners that counter-mark scent mark challenges from competitors and which ensure that their own marks are always the freshest (Rich and Hurst, 1999).