Diversity of major urinary proteins (MUPs) in wild house mice

Major urinary proteins (MUPs) are often suggested to be highly polymorphic, and thereby provide unique chemical signatures used for individual and genetic kin recognition; however, studies on MUP variability have been lacking. We surveyed populations of wild house mice (Mus musculus musculus), and examined variation of MUP genes and proteins. We sequenced several Mup genes (9 to 11 loci) and unexpectedly found no inter-individual variation. We also found that microsatellite markers inside the MUP cluster show remarkably low levels of allelic diversity, and significantly lower than the diversity of markers flanking the cluster or other markers in the genome. We found low individual variation in the number and types of MUP proteins using a shotgun proteomic approach, even among mice with variable MUP electrophoretic profiles. We identified gel bands and spots using high-resolution mass spectrometry and discovered that gel-based methods do not separate MUP proteins, and therefore do not provide measures of MUP diversity, as generally assumed. The low diversity and high homology of Mup genes are likely maintained by purifying selection and gene conversion, and our results indicate that the type of selection on MUPs and their adaptive functions need to be re-evaluated.

incubated at 37°C overnight. The digest was stopped by addition of 20 % trifluoro acetic acid (TFA, 1 μL). Mass spectrometric protein identification of shotgun samples was based on a nano high performance liquid chromatography electrospray quadrupole time of flight mass spectrometer system nanoHPLCESI-QTOF-MS, TripleTOF 5600, Sciex, USA). Peptides were separated with an Ultimate 3000 RSLC using a pre-concentration trap column (PepMap100 C18) and a nano separation column (Acclaim PepMap RSLC 75 μm x 25 cm, nano Viper C18, 2 μm, 100 Å) (Dionex, The Netherlands). Peptides were separated with a gradient from 4 % B (80 % acetonitrile (ACN) with 0.1 % formic acid) to 35 % B in 120 min and then up to 90 % B in 15 min followed by a washing step with 90 % B for 10 min. Mobile Phase A consisted of H2O with 0.1 % formic acid. We performed instrument calibration using a ß-Galactosidase digest (Sciex, USA) before each sample injection. The mass measurement accuracy was <2 ppm RMS, a mass tolerance of 50 ppm was used for autocalibration of the instrument. Detection was carried out on a high resolution QTOF mass spectrometer online coupled to the LC with an ESI source. Due to high MUP homology, data were recorded from m/z= 250 to 1500 to also cover differences in short terminal fragments for protein differentiation. A Top 25 method was chosen with the following parameters: used. We performed fragmentation based on a data dependent acquisition strategy by using a unique peptide list.

Method C: MUP identification after two-dimensional gel electrophoresis (2D-PAGE)
In this study, 2D-PAGE was conducted using a combination of native IEF as described in Method B (see above) and sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE), thus separating proteins first by their pI and subsequently by their molecular weight (MW). This setup is different from classical two-dimensional electrophoresis which uses denaturing/reducing IEF in the first dimension. However, we chose this setup to ensure comparability to 1D patterns of Method B.
For the first dimensional run, 5 mm strips were cut from narrow range dry plates, rehydrated and proteins (5 μg) separated as described 5 . Post separation, strips were frozen until use. For SDS-PAGE, separation gels (140x140x1.5 mm) were prepared with a 10-20 % T gradient in the upper half and 20 homogenous gel in the lower half, for improved separation of low molecular mass compounds. A stacking gel (5 % T) was polymerized on top of the separation gel, the IEF strip put on it and fixed in place by 1 % agarose. Separation was performed in a Hoefer SE600 vertical electrophoresis chamber (Hoefer Scientific Instruments, San Francisco, CA, USA) and proteins stained with an MS compatible silver stain 8 . Individual spots were excised and destained as described in 9 . Further sample preparation was performed according to Method B with adaptations for trypsin digestion at pH 8.5.
For all washing steps 100 mM ABC was used. Tryptic digest was conducted with 0.2 μg of trypsin (Trypsin Gold, Mass Spectrometry Grade, Promega, USA) in 50 mM ABC and 5 mM calcium dichloride for 8 h at 37°C. Peptide separation and protein identification was performed with nanoHPLC-ESI-IT-MS as described in Method B adapted for trypsin and using a nanoHPLC-ESI-QTOF-MS (see Method A).

MUP identification
In our study, protein identification was based on unique peptides of MUPs. In contrast to other proteomics studies, we could not set a minimum number of two identified peptides per protein because MUPs are so highly homologous and often differ by a single unique peptide only.
Identification of MUP proteins was considered significant if one proteotypic peptide was identified. A decoy database was created from reversed database entries and the FDR was calculated from performing a search against this database. A protein confidence threshold of 0.05 was used and the FDR was 1 %.

Database construction
For database searches on an in-house Mascot server 2.3.01 (Matrix Science, UK), a database of isoform-specific unique peptides was created based on UniprotKB/Swiss-Prot, UniprotKB/TrEMBL, NCBInr and a MUP cDNA database 10 . This approach utilized all available MUP sequence information by including both reviewed (independently confirmed MUP sequences) and unreviewed (MUP sequences from computationally generated annotation and/or not independently confirmed) MUP entries. Allowing variable modifications, we were able to choose a peptide for identification. Proteins were identified based on peptide sequences computed from MS/MS spectra and identification was not influenced by any PTMs as the software takes possible PTMs into account.

Supplemental Results
To identify individual MUP proteins in our samples, we conducted several additional analyses using Spot number

Spot number
Supplemental