New insight into the role of MMP14 in metabolic balance

The mouse model

C57BL/6 back-crossed mice carrying the LacZ gene under the control of the Mmp14 promoter (Mmp14^tm2Ski; Mmp14 (+∕ −, lacZ) heterozygotes) were used to obtain homozygotes and siblings (Mori et al., 2013; Yana et al., 2007). Inguinal mammary glands were isolated from each mouse at the age of day 7 to day 15. Genotypes were checked by PCR of tail DNA. Procedures were followed in strict adherence to animal use protocol (Animal Use Protocol No. 17301) and guidelines established by Lawrence Berkeley National Laboratory’s Animal Welfare and Research Committee (AWRC).

Materials for MS analysis

Invitrosol™ was purchased from Invitrogen (Carlsbad, CA, USA). Trypsin (modified, sequencing grade) was obtained from Promega, WI. Other laboratory reagents were purchased from Sigma-Aldrich (St. Louis, MO, USA) and Thermo Fisher Scientific (Waltham, MA, USA) unless noted otherwise.

Sample preparation for MS analysis

Isolated tissues were homogenized in the mass spectrometry-compatible lysis buffer (50 mM Ammonium Bicarbonate, 4 M Urea, 1X invitrosol, protease and phosphatase inhibitors) using the Precellys 24 tissue homogenizer. After homogenization, tissue lysates were cleared by centrifugation. The protein concentration was determined using an EZQ protein assay (Invitrogen, Carlsbad, CA, USA). For quantitative global protein analysis, we used mixed tissue lysates (brain, liver, lung, and kidney) from metabolically labeled mice as reference standards for quantification. C57BL/6 mice were labeled metabolically using stable isotope-labeled (¹⁵N) amino acids (SILAM, Silantes, Germany). The isotopic enrichment of proteins in the labeled tissue was determined by LC-MS/MS to ensure greater than 97% of incorporation of ¹⁵N amino acids. Equal protein amounts were mixed with unlabeled WT or Mmp14KO mammary gland anlagen for relative quantification.

Trypsin digestion for MS analysis

Unlabeled lysates (20 µg) from WT or Mmp14KO mammary gland anlagen were mixed with an equal amount of ¹⁵N labeled mixed tissue lysates and diluted in 50 mM Ammonium Bicarbonate for trypsin digestion. Trypsin was added to each sample at a ratio of 1:30 enzyme/protein along with 2 mM CaCl₂ and incubated for 16 h at 37°C. Following digestion, all reactions were acidified with 90% formic acid (2% final) to stop proteolysis. Samples were centrifuged subsequently for 30 min at 14,000 rpm to remove any insoluble material. The soluble peptide mixtures were collected for LC-MS/MS analysis.

Multidimensional chromatography and tandem mass spectrometry

Peptide mixtures were pressure-loaded onto a 250 µm inner diameter (i.d.) fused-silica capillary packed first with 3 cm of 5 µm strong cation exchange material (Partisphere SCX, Whatman), followed by 3 cm of 10 µm C18 reverse phase (RP) particles (Aqua, Phenomenex, Torrance, CA, USA). Loaded and washed microcapillaries were connected via a 2 µm filtered union (UpChurch Scientific) to a 100 µm i.d. column, which had been pulled to a 5 µm i.d. tip using a P-2000 CO₂ laser puller (Sutter Instruments), then packed with 13 cm of 3 µm C18 reverse phase (RP) particles (Aqua, Phenomenex, CA, USA) and equilibrated in 5% acetonitrile, 0.1% formic acid (Buffer A). This split-column was then installed in-line with a NanoLC Eskigent HPLC pump. The flow rate of channel 2 was set at 300 nl/min for the organic gradient. The flow rate of channel 1 was set to 0.5 µl/min for the salt pulse. Fully automated 11-step chromatography runs were carried out. Three different elution buffers were used: 5% acetonitrile, 0.1% formic acid (Buffer A); 98% acetonitrile, 0.1% formic acid (Buffer B); and 0.5 M ammonium acetate, 5% acetonitrile, 0.1% formic acid (Buffer C). In such sequences of chromatographic events, peptides are sequentially eluted from the SCX resin to the RP resin by increasing salt steps (increase in Buffer C concentration), followed by organic gradients (increase in Buffer B concentration). The last chromatography step consists in a high salt wash with 100% Buffer C followed by acetonitrile gradient. The application of a 2.5 kV distal voltage electrosprayed the eluting peptides directly into a LTQ-Orbitrap XL mass spectrometer equipped with a nano-LC electrospray ionization source (Thermo Finnigan). Full MS spectra were recorded on the peptides over a 400 to 2,000 m∕z range by the Orbitrap, followed by five tandem mass (MS/MS) events sequentially generated by LTQ in a data-dependent manner on the first, second, third, and fourth most intense ions selected from the full MS spectrum (at 35% collision energy). Mass spectrometer scan functions and HPLC solvent gradients were controlled by the Xcalibur data system (Thermo Finnigan, San Jose, CA).

Database search and interpretation of MS/MS datasets

Tandem mass spectra were extracted from raw files, and a binary classifier—previously trained on a manually validated data set—was used to remove the low quality MS/MS spectra. The remaining spectra were searched against a mouse protein database downloaded as FASTA-formatted sequences from UniProt (database released on June, 2012) (Besson, Soustelle & Birman, 2000). To calculate confidence levels and false positive rates, we used a decoy database containing the reverse sequences of annotated protein sequences appended to the target database (Elias & Gygi, 2007), and the SEQUEST algorithm (Yates et al., 1995) to find the best matching sequences from the combined database.

SEQUEST searches were done using the Integrated Proteomics Pipeline (IP2, Integrated Proteomics Inc., CA) on Intel Xeon X5450 X/3.0 PROC processor clusters running under the Linux operating system. The peptide mass search tolerance was set to 50 ppm. No differential modifications were considered. At least half tryptic status was imposed on the database search, so the search space included all candidate peptides whose theoretical mass fell within the 50 ppm mass tolerance window. The validity of peptide/spectrum matches was assessed in DTASelect2 (Tabb, McDonald & Yates, 2002) using SEQUEST-defined parameters, the cross-correlation score (XCorr) and normalized difference in cross-correlation scores (DeltaCN). The search results were grouped by charge state (+1, +2, and +3) and tryptic status (fully tryptic, half-tryptic, and non-tryptic), resulting in 9 distinct sub-groups. In each one of the sub-groups, the distribution of XCorr and DeltaCN values for (a) direct and (b) decoy database hits was obtained, and the two subsets were separated by quadratic discriminant analysis. Outlier points in the two distributions (for example, matches with very low Xcorr but very high DeltaCN were discarded. Full separation of the direct and decoy subsets is not generally possible; therefore, the discriminant score was set such that a false positive rate of 1% was determined based on the number of accepted decoy database peptides. This procedure was independently performed on each data subset, resulting in a false positive rate independent of tryptic status or charge state.

In addition, a minimum sequence length of seven amino acid residues was required, and each protein on the final list was supported by at least two independent peptide identifications unless specified. These additional requirements—especially the latter—resulted in the elimination of most decoy database and false positive hits, as these tended to be overwhelmingly present as proteins identified by single peptide matches. After this last filtering step, the false identification rate was reduced to below 1%.

Quantitative global protein analysis

SEQUEST identified ¹⁴N and ¹⁵N labeled peptides based on their fragmentation spectra. CenSus, an algorithm-based quantification software (Park et al., 2008), was used to identify co-eluting ¹⁴N and ¹⁵N peptide peaks from the MS based on MS/MS identifications, generate ratios of co-eluting ¹⁴N and ¹⁵N peptides based on the measured ion intensities, and perform statistical analysis (R² correlation, ratio distribution of peptides, and etc.). Only co-eluting ¹⁴N and ¹⁵N peptides with R² scores greater than 0.5 were used for protein quantification. Expression changes at the protein level were calculated by averaging the ratio measurements (weighed average) of unlabeled and ¹⁵N labeled peptides for each protein. Differential protein expression in the WT and Mmp14KO tissues was calculated by dividing WT ratios (¹⁴N WT/¹⁵N standard) by KO ratios (¹⁴N Mmp14KO/¹⁵N standard).

Tissue staining

H&E stained and unstained 5 µm tissue paraffin sections were generated by the UCSF Helen Diller Family Comprehensive Cancer Center Mouse Pathology Core. For glycogen staining, de-paraffinized tissue sections were stained with Alexa Fluor 594 conjugated GSA-II (Life Technologies) at 1 mg/ml in PBS (pH 7.4) containing 0.1 mM CaCl₂. DAPI was used to observe tissue architecture. Images were captured by laser scanning confocal microscopy (LSM710; Zeiss). The glycogen depositions in tissues were measured with IMARIS (Bitplane). H&E stained mammary gland tissues were used to measure the lipid amount in tissue. ImageJ was used to measure the area of lipid droplets and to count the stromal cell density in tissue sections. Wholemount mammary gland samples from Mmp14 (+∕ +) and Mmp14 (−∕ −) were used to render 3D tissue images (Mori et al., 2012) and lipid volume and cell density were measured with IMARIS.

Measurement of blood glucose and triglyceride

Blood glucose and triglyceride are measured after mice were euthanized. Glucose was detected with Bayer Contour blood glucose monitoring system (Bayer AG, Leverkusen, Germany). Triglyceride was measured with CardioChek portable blood test system (CardioChek, Indianapolis, IN, USA). Test was performed 3 times on each sample, and averaged numbers were used for statistical analyses (n = 5 for glucose, n = 4 for triglyceride).

Animal care, deuterium administration and tissue collection

Deuterium administration for isotopic labeling of tissue was based on a modified protocol (Louie et al., 2013). Briefly, mammary glands were dissected from WT and KO neonates that obtained nourishment from nursing on the mother mouse given free access to drinking water (8% D₂O). Animals were euthanized after 7 days, then mammary glands collected and immediately stored at –80°C. As an unlabeled control, a mammary gland was also collected from a neonate never exposed to D₂O.

Liquid Chromatography Electrospray Ionization-Mass Spectrometry (LC ESI-MS) and MS/MS of mammary glands

In preparation for mass spectrometry analysis, mammary glands were frozen at –80°C and lyophilized dry, then homogenized using a Mini-Beadbeater (BioSpec Products, Bartlesville, OK) for 5 s. A chloroform-based lipid extraction was performed using a modified BligWh-Dyer approach (Bligh & Dyer, 1959). Briefly, 2:1:0.8 MeOH:CH₃Cl:H₂O was added to the tissue and sonicated 30 min in a water bath. Water and CH₃Cl for a final solvent ratio of 1:1:0.9 MeOH:CH₃Cl:H₂O were then added and samples sonicated again for 30 min, centrifuged 6 min at 6,000 rpm, after which the bottom lipid-enriched chloroform phase was transferred to a new tube. Additional CH₃Cl was added followed by further sonication and centrifugation, with bottom chloroform phase combined with the previously collected extract. Chloroform extracts of lipids were then dried down in a SpeedVac (Thermo Savant SPD111) and stored at –20°C until further use.

LC-MS/MS was performed on mammary gland lipid extracts resuspended in 3:3:4 IPA:ACN:MeOH and filtered through a 0.22 µm PVDF membrane. Reverse phase chromatography was performed using an Agilent 1290 LC system and Agilent C18 column (5 µm 150 × 0.5 mm Zorbax SB-C18) at a flow rate of 30 µL/min using a 1–2 µL injection volume. The column was equilibrated with 80% buffer A (60:40 H₂O:ACN w/ 5 mM ammonium acetate) for 2 min, which was then diluted down to 65% with buffer B (90:10 IPA:ACN w/5 mM ammonium acetate) over 8 min, then down to 25% A over 22 min, followed by isocratic elution in 99% buffer B for 6 min. Lipids were identified using exact mass and retention time (Table S1).

Determination of lipid isotopic enrichment

For each identified lipid, the ion intensity of the first and second isotopes, M0 and M1, respectively, were used to calculate relative isotopic enrichment. For an unlabeled lipid, the M1/M0 ratio is the expected isotopic ratio based upon the distribution of naturally occurring isotopes. For an isotopically enriched lipid that has been labeled with deuterium, the M1/M0 ratio increases, with higher ratios corresponding to higher enrichment levels as deuterium is incorporated from D₂O during new lipid synthesis. Here, “relative isotopic enrichment” is calculated as enrichment relative to the unlabeled lipid given by the equation: [(M1/M0)_D2O −(M1/M0)_noD2O]/(M1/M0)_noD2O.

Sample preparation for transmission electron microscopy

Dissected tissues were placed into 2% paraformaldehyde with 0.1% glutaraldehyde for at least 1 h prior to high pressure freezing (HPF). 1 mm biopsy cores were used to extract targeted regions of the mammary gland and intestine. These tissue pieces were placed in 1 mm wide by 200 mm deep aluminum freezing hats and, prior to freezing, were surrounded with 20% Bovine Serum Albumin as needed, here used as a cryo-protectant. Samples were then cryo-immobilized using a BAL-TEC HPM-010 high-pressure freezer (BAL-TEC, Inc., Carlsbad, CA) and freeze-substituted in 1% osmium tetroxide and 0.1% uranyl acetate in acetone with 5% ddH2O, as previously described (McDonald & Muller-Reichert 2002; McDonald & Webb 2011). Upon completion of freeze-substitution, samples were progressively infiltrated with an Epon-Araldite resin. Polymerization in Epon-Araldite resin was performed by flat-embedding between two glass slides to allow for precise localization of features of interest (Muller-Reichert et al., 2003). Samples were sectioned into 70–100 nm thin and 500 nm thick sections using a Leica UC6 Ultramicrotome (Leica Microsystems, Wetzlar, Germany). Sections were then collected onto formvar-coated, rhodium-enforced copper 2 mm slot grids. The grids were post-stained with 2% uranyl acetate followed by Reynold’s lead citrate. The sections were imaged using a FEI Technai 12 TEM (FEI, Eindhoven, The Netherlands), and images were recorded using a Gatan CCD with Digital Micrograph software (Gatan Inc., Pleasanton, CA). Serial EM software was used to collect wide-field montages for overview TEM images of mammary duct or intestine cross sections. ImageJ software and Adobe Photoshop CS4 were used for further image processing.

Observation of autophagosomes

To identify, and observe the dynamics of, autophagosomes, cDNA of EGFP tagged LC3 (GFP-LC3) (Lee et al., 2008) (Addgene) was ligated into pLenti-EF1α-Puro (Mori et al., 2013) and used for generating lentiviral particles containing this cDNA in order to transduce mouse mammary epithelial cell line (EpH4 cells). Lentiviruses were packaged within 293FT cells using FuGene6 (Roche, Basel, Switzerland). Transfected 293FT cells were cultured for 24 h in DMEM high glucose (Life Technologies), and the medium was replaced with a fresh lot for 48 h. Recombinant lentiviral particles were concentrated from the filtered culture medium (0.45 mm filter) using Lenti-X concentrator (Clontech). To transduce GFP-LC3 in EpH4 cells treated with control- or Mmp14-shRNA (Mori et al., 2013), 1.0 × 10⁵ cells were plated in 6-well plate, infected with the lentiviral particles in growth medium (DMEM/F12 supplemented with 2% fetal bovine serum (FBS), 5 µg/ml insulin and 50 µg/ml gentamycin) supplemented with 4 µg/ml polybrene, and selected with puromycin at 3 µg/ml in growth medium. GFP-LC3 transduced EpH4 cells were plated on LabTek II 8-well chamber cover glass (Nunc) at 20,000 cells per well for 48 h in growth medium. Time-lapse sequences were acquired using laser scanning confocal microscopy (LSM710, Zeiss) equipped with an incubation chamber to maintain constant temperature (37°C) and uniform levels of humidity and CO₂ (5%). To induce autophagy in cells, culture media was replaced with Hank’s balanced salt solution (HBSS). The imaging analysis was performed 24 h post nutrient starvation. Acquired images were analysed with IMARIS (Bitplane) to quantify vesicle number per cell.

Mmp14KO neonates ingest, digest and assimilate nutritive principles from their mother’s milk

Mmp14KO neonates have earlier been reported to have a craniofacial skeletal defects (Holmbeck et al., 1999). Therefore, we investigated whether they were able to ingest milk from their mother for nutrition and development. Dissection and gross anatomical examination of the alimentary canals of Mmp14KO mice revealed no defects such as atresia, fistulas, or adhesions, (although ileal coils were fewer than- and contained more flatus with respect to- WT) (Fig. 1A). Next, in order to test absorption and assimilation, lactating mothers were fed with D₂O (heavy water), and mammary glands were isolated from their infants. We observed deuterium-containing lipids in mammary glands from both WT and KO neonatal mice, confirming that Mmp14KO infant mice were able to drink milk and utilize its nutritive principle for the metabolism of their organs (Fig. 1B).

Proteomic profile of mammary glands of Mmp14KO reveals defective energy metabolism

To dissect the total effects of MMP14 deletion, we concentrated on isolating the mammary gland anlagen from 1 to 2 weeks-old Mmp14KO-and WT siblings (Fig. 2A). The relative protein levels in the two cohorts were quantified by two-dimensional LC-MS/MS with stable-isotope amino acids labeled reference tissue lysate as described (Koller et al., 2013; Sinnamon et al., 2012). We identified 1,346 proteins from WT and Mmp14KO anlagen. Protein expression ratios of WT and the KO tissues (logarithmic plot) were normally distributed and centered around 1:1 ratio (Ln(KO/WT) = 0) (Fig. 2B). Among the 1346 proteins identified, 142 proteins were significantly higher (Ln(KO/WT) > 0.5) and 325 proteins were significantly lower (Ln(KO/WT) < − 0.5) in the KO tissue (Fig. 2B). Gene ontological analysis of the altered proteins indicated that the deletion of MMP14 correlated with significant changes of several metabolic pathways (Fig. 3A). The finding that a subset of enzymes involved in glucose and lipid metabolism are altered in the Mmp14KO tissue compared to the WT (Figs. 3B and 3C) was unexpected since the literature essentially has concentrated on MMP14’s catalytic actions on ECM molecules.

Mmp14KO mice show decreased tissue glycogen and lipid levels and lower plasma glucose and triglyceride levels

To prove the functional significance of the observed changes in enzyme levels, we performed assays for glycogen and lipid levels in mammary tissues. Glycogen deposition was detected in mammary gland sections from 2 week-old WT and Mmp14KO mice with fluorescently labeled Griffonia simplicifolia agglutinin-II (GSA-II), which specifically binds to glycogen (Ebisu & Goldstein, 1978; Hennigar, Schulte & Spicer, 1986). Whereas the mammary gland from WT and Mmp14 (+∕ −) showed intense staining for glycogen (Fig. 4Aa; Fig. S1), sections from Mmp14KO mice showed scant deposition of glycogen (Fig. 4Ab). We quantified the glycogen levels in each different compartment of the mammary gland: epithelia (luminal (LEP) and myoepithelial (MEP)), peri-ductal stroma (PDS), capillary vessels (CV), fat cells (FC) and lymph nodes (LN) (Fig. Ac). In KO animals, all mammary gland tissues showed significantly lower glycogen deposition compared to corresponding tissue compartments in the WT animal. Given that liver is known to be a major storage tissue for glycogen, we assayed for glycogen levels also in this organ. Mmp14KO liver similarly had reduced levels of glycogen (Fig. S2), indicating that Mmp14KO mice have an overall glycogen storage defect.

Mass Spec analysis predicted that the levels of lipid would be lower in Mmp14KO mammary tissues based on the reduced expression levels of enzymes involved in lipid synthesis (Fig. 3B). The hematoxylin and eosin (H&E) staining of Mmp14KO mammary gland tissue sections showed a smaller size of lipid droplets compared to WT (Figs. 4Ba and 4Bb). The lipid levels in KO tissue sections were significantly reduced (∼4 fold) relative to their control counterparts (Fig. Bc). Quantification of cell density within the stromal compartment as determined by counting nuclei in tissue sections showed Mmp14KO stroma having a ∼5 fold more cells compared to the WT (Fig. 4Bd, Fig. S3). Thus, Mmp14KO mammary glands have insufficient energy storage. One possible means of compensation might be to obtain glucose from the blood, but we found that blood glucose levels were significantly lower in KO compared to WT mice (WT: 214.1 ± 10.73 mg/dL, KO: 88.10 ± 11.49 mg/dL, Fig. 4Ca). We tested if higher blood triglyceride levels could be observed as a result of feedback mechanism that maintains homeostasis, as is observed in diabetes mellitus. Surprisingly again, unlike diabetic blood profiles, KO mice had significantly lower blood triglyceride levels compared to the WT (Fig. 4Cb) leaving us with the conclusion that the demise of the KO newborns would most likely be due to malnutrition.

The mammary epithelium of Mmp14KO show increased autophagy

To explain the above findings, we hypothesized that Mmp14KO mammary epithelial cells may undergo autophagy, a well-conserved cellular process that is deployed to recycle and degrade cell membranes, organelles and cytoplasmic complexes, as an alternative route to obtain energy (Rabinowitz & White, 2010; Singh & Cuervo, 2011). To confirm, we used electron microscopy to observe the differences in subcellular structures between WT and KO mammary anlagen. We observed dramatic increase in membrane-bound vesicles in the mammary gland epithelium of Mmp14KO suggestive of extensive autophagy (Figs. 5A, 5B and Fig. S5). Because autophagy is characterized by the presence of double- and multiple- membrane autophagic organelles, we counted the number of early and late endosomes, lysosomes, autophagophores, autophagosomes and autolysosomes (Fig. 5C and Fig. S6). Mammary anlagen from KO mice displayed significant increase in total endocytic (x) and autophagic (y) organelles per epithelial cell (x + y: 5.0/cell) compared to the WT counterpart (x + y: 2.2/cell) (Figs. 5D and 5E). Whereas 84% of the KO organelles were engaged in autophagy, WT mice showed only 31% autophagic organelles (Figs. 5D and 5E). These results confirm that MMP14 activity is necessary to prevent, tissues from resorting to increased autophagy in order to obtain energy.

MMP14 protects mammary cells from nutrient-deprived autophagy

To determine whether increased autophagy is directly related to loss of MMP14, we transduced a EGFP (GFP-LC3)- tagged autophagosome marker (Lee et al., 2008) into the mouse mammary epithelial cell line, EpH4, with or without Mmp14 silencing (Alcaraz et al., 2011; Mori et al., 2013). We observed formation of autophagic organelles induced by nutrient starvation in Hank’s balanced salt solution (HBSS) (Figs. 6A–6C). Silencing Mmp14 in EpH4 cells significantly enhanced autophagosome formation as measured by the numbers of GFP-LC3 positive vesicles per cell (Fig. 6C). MMP inhibitor-treated EpH4/GFP-LC3 cells also showed significantly higher number of autophagosome (Figs. 6D–6F) suggesting involvement of MMP proteolytic activity in this process. These results suggest that the autophagic phenotype in the mammary gland of KO mice could be caused by the absence of MMP14 implicating this MMP regulation of autophagy.