Comparative proteomic analysis of human lung telocytes with fibroblasts

Telocytes (TCs) were recently described as interstitial cells with very long prolongations named telopodes (Tps; http://www.telocytes.com). Establishing the TC proteome is a priority to show that TCs are a distinct type of cells. Therefore, we examined the molecular aspects of lung TCs by comparison with fibroblasts (FBs). Proteins extracted from primary cultures of these cells were analysed by automated 2-dimensional nano-electrospray ionization liquid chromatography tandem mass spectrometry (2D Nano-ESI LC-MS/MS). Differentially expressed proteins were screened by two-sample t-test (P < 0.05) and fold change (>2), based on the bioinformatics analysis. We identified hundreds of proteins up- or down-regulated, respectively, in TCs as compared with FBs. TC proteins with known identities are localized in the cytoskeleton (87%) and plasma membrane (13%), while FB up-regulated proteins are in the cytoskeleton (75%) and destined to extracellular matrix (25%). These identified proteins were classified into different categories based on their molecular functions and biological processes. While the proteins identified in TCs are mainly involved in catalytic activity (43%) and as structural molecular activity (25%), the proteins in FBs are involved in catalytic activity (24%) and in structural molecular activity, particularly synthesis of collagen and other extracellular matrix components (25%). Anyway, our data show that TCs are completely different from FBs. In conclusion, we report here the first extensive identification of proteins from TCs using a quantitative proteomics approach. Protein expression profile shows many up-regulated proteins e.g. myosin-14, periplakin, suggesting that TCs might play specific roles in mechanical sensing and mechanochemical conversion task, tissue homoeostasis and remodelling/renewal. Furthermore, up-regulated proteins matching those found in extracellular vesicles emphasize TCs roles in intercellular signalling and stem cell niche modulation. The novel proteins identified in TCs will be an important resource for further proteomic research and it will possibly allow biomarker identification for TCs. It also creates the premises for understanding the pathogenesis of some lung diseases involving TCs.

However, the protein expression profile for this type of cells has not been reported yet. In a previous study, we reported the gene expression profile of mice lung TCs compared to fibroblasts (FBs), and the results showed that more than~1000 genes were found upor down-regulated respectively [42]. Significantly improved technology in quantitative MS-based proteomics permits the measurement of relative protein loads in cell culture or tissue samples with unique precision [43,44].
To prevent further confusion between TCs and other interstitial (stromal) cells, particularly with FBs and the so-called FBs-like cells, we carried out a comprehensive study using iTRAQ-coupled 2D LC-MS/MS analysis to identify and quantify the proteins. This study is the first that allows major insight into proteome differences between these cells in human lung and which identify the proteins that are specifically expressed in TCs.

Cell lines and tissue sampling
Human lung samples were obtained from the patients undergoing surgery for lung cancer. The normal tissue was defined as being located at a distance of at least 15 cm from the tumour tissue and verified by light microscopy. The application of human tissue for research was approved by the Ethical Evaluation Committee of Zhongshan Hospital, Fudan University, Shanghai, China. Human lung fibroblast cell lines were obtained from Chinese Academy of Science (Cat. no. GNHu28; Shanghai, China).

Primary cell culture and lysis of lung TCs
The methods of isolation and culture of lung TCs were previously described by Zheng et al. [22]. Briefly, lung tissue was cut into small pieces and harvested under sterile conditions and collected into sterile tubes containing DMEM (Gibco, Grand Island, NY, USA), supplemented with 100 UI/ml penicillin and 0.1 mg/ml streptomycin (Sigma Chemical, St. Louis, MO, USA), and the samples were brought to the cell culture room immediately. Samples were further rinsed with sterile DMEM and minced into fragments about 1 mm 3 , which were then incubated at 37°C for 4 hrs on an orbital shaker, with 1 mg/ml type II collagenase (Sigma-Aldrich, St. Louis, MO, USA) in PBS without Ca 2+ and Mg 2+ . Dispersed cells were separated from non-digested tissue by the filtration through a 40-lm diameter cell strainer (BD Falcon, Franklin, NJ, USA), harvested by centrifugation and resuspended in DMEM supplemented with 10% foetal calf serum (Gibco), 100 UI/ml penicillin and 0.1 mg/ml streptomycin. Cell density was counted in a haemocytometer and viability was assessed using the Trypan blue. Cells were distributed in 25 cm 2 culture flasks at a density of 1 9 10 5 cells/cm 2 and maintained at 37°C in a humidified atmosphere (5% CO 2 ) until becoming semiconfluent (usually 4 days after plating). Culture medium was changed every 48 hrs. Cultured cells were examined by phase contrast microscope, under an inverted Olympus phase contrast microscope (1 9 51). Cells (1 9 10 5 ) were placed in 10-cm dishes with 10 ml high glucose DMEM (Gibco) complete medium, including 10% foetal calf serum (Gibco), 100 UI/ml penicillin and 0.1 mg/ml streptomycin (Sigma Chemical) in a humidified atmosphere of 5% CO 2 at 37°C. Confluent cells were trypsinized at day 5 and day 10 respectively. Approximately 10 6 cells from day 5 or day 10 were resuspended in a solution of 9.5 moles/litre urea, 1% dithiothreitol (DTT), 40 mg/ml protease inhibitor cocktail, 0.2 mmoles/litre Na 2 VO 3 and 1 mmole/litre NaF. The mixture was incubated and stirred by end-over-end rotation at 4°C for 60 min. The resultant suspension was centrifuged at 40,000 9 g for 1 hr at 15°C. The supernatant was stored in small aliquots at À80°C, and the protein concentration was determined using a modified Bradford method.

Automated 2-D nano-ESI LC-MS/MS analysis of peptides
Proteins extracted from primary cultures of TCs and FBs were analysed by automated 2-dimensional nano-electrospray ionization liquid

Sample preparation
The samples were ground in liquid nitrogen. One millilitre of lysis buffer (8 M urea, 19 Protease Inhibitor Cocktail; Roche Ltd., Basel, Switzerland) was added to sample, followed by sonication on ice and centrifugation at 29,000 9 g. for 10 min. at 4°C. The supernatant was transferred to a fresh tube and stored at À80°C until needed.

iTRAQ labelling and protein digestion
For each sample, proteins were precipitated with ice-cold acetone and then were redissolved in the dissolution buffer (0.5 M triethylammonium bicarbonate, 0.1% SDS). Then proteins were quantified by the bicinchoninic acid protein assay, and 100 lg of protein was tryptically digested and the resultant peptide mixture was labelled using chemicals from the iTRAQ reagent kit (Applied Biosystems, Foster City, CA, USA). Disulphide bonds were reduced in 5 mM Tris-(2-carboxyethy) phosphine (TCEP) for 1 hr at 60°C, followed by blocking cysteine residues in 10 mM methyl methanethiosulfonate (MMTS) for 30 min. at room temperature, before digestion with sequence-grade modified trypsin (Promega, Madison, WI, USA). For labelling, each iTRAQ reagent was dissolved in 50 ll of isopropanol and added to the respective peptide mixture. Proteins were labelled with the iTRAQ tags as follows: Fibroblast (5 days) -114 isobaric tag, TCs (5 days) -116 isobaric tag, Fibroblast (10 days) -118 isobaric tag, TCs (10 days) -121 isobaric tag. The labelled samples were combined and dried in vacuo. A SepPac TM C18 cartridge (1 cm 3 /50 mg, Waters Corporation, Milford, MA, USA) was used to remove the salt buffer and then was dried in a vacuum concentrator for the next step.

High pH reverse phase separation
The peptide mixture was redissolved in the buffer A (buffer A: 20 mM ammonium formate in water, pH 10.0, adjusted with ammonium hydroxide), and then fractionated by high pH separation using a Aquity UPLC system (Waters Corporation) connected to a reverse phase column (XBridge C18 column, 2.1 9 150 mm, 3.5 lm, 300 A, Waters Corporation). High pH separation was performed with a linear gradient. Starting from 2% B to 40% B in 45 min. (B: 20 mM ammonium formate in 90% ACN, pH 10.0, adjusted with ammonium hydroxide). The column was re-equilibrated at initial conditions for 15 min. The column flow rate was maintained at 200 ll/min. and column temperature was maintained at room temperature [47]. Fourteen fractions were collected, and each fraction was dried in a vacuum concentrator for the next step.

Low pH nano-HPLC-MS/MS analysis
The peptides were resuspended with 80 ll solvent C (C: water with 0.1% formic acid; D: ACN with 0.1% formic acid), separated by nanoLC and analysed by on-line electrospray tandem mass spectrometry. The experiments were performed on a Nano Aquity UPLC system (Waters Corporation) connected to an LTQ Orbitrap XL mass spectrometer (Thermo Electron Corp., Bremen, Germany) equipped with an online nanoelectrospray ion source (Michrom Bioresources, Auburn, CA, USA). 20 ll peptide sample was loaded onto the Thermo Scientific Acclaim PepMap C18 column (100 lm 9 2 cm, 3 lm particle size), with a flow of 10 ll/min. for 5 min. and subsequently separated on the analytical column (Acclaim PepMap C18, 75 lm 9 15 cm) with a linear gradient, from 5% D to 45% D in 165 min. The column was re-equilibrated at initial conditions for 15 min. The column flow rate was maintained at 300 nl/min. and column temperature was maintained at 35°C. The electrospray voltage of 1.9 kV versus the inlet of the mass spectrometer was used.
LTQ Orbitrap XL mass spectrometer was operated in the data-dependent mode to switch automatically between MS and MS/MS acquisition. Survey full-scan MS spectra (m/z 400-1600) were acquired in the Orbitrap with a mass resolution of 30,000 at m/z 400, followed by five sequential HCD-MS/MS scans. The automatic gain control was set to 500,000 ions to prevent over-filling of the ion trap. The minimum MS signal for triggering MS/MS was set to 1000. In all cases, one microscan was recorded. MS/MS scans were acquired in the Orbitrap with a mass resolution of 7500. The dissociation mode was HCD (higher energy C-trap dissociation). Dynamic exclusion was used with two repeat counts, 10-sec. repeat duration, and the m/z values triggering MS/MS were put on an exclusion list for 120 sec. For MS/MS, precursor ions were activated using 40% normalized collision energy and an activation time of 30 ms.  Database searching, criteria and protein parameterization Protein identification and quantification for the iTRAQ experiment was performed with the ProteinPilot software version 4.0 (Applied Biosystems). The database was the Human UniProtKB/Swiss-Prot database (Release 2011_10_15, with 20248 sequences). The Paragon Algorithm in ProteinPilot software was used for peptide identification and isoformspecific quantification. The data search parameters were set up as follows: trypsin (KR) cleavage with two missed cleavage was consid-ered; fixed modification in cysteines by MMTS; iTRAQ modification in peptide N termini, methionine oxidation and iTRAQ modification in lysine residues were set as variable modification. To minimize false positive results, a strict cut-off for protein identification was applied with the unused ProtScore ≥1.3, which corresponds to a confidence limit of 95%, and at least two peptides with the 95% confidence were considered for protein quantification. The resulting data set was auto bias corrected to get rid of any variations imparted because of unequal mixing during combining different labelled samples. For iTRAQ quantification, the peptide for quantification was automatically selected by Pro Group algorithm (at least two peptides with 99% confidence) to calculate the reporter peak area, error factor and p-value. For the selection of differentially expressed proteins, we considered the following situation: (i) the proteins must contain at least two unique high-scoring peptides; (ii) the proteins must have a P < 0.05 and the proteins identified with mass tag changes ratio must be ≥1.3 or ≤0.75.
Differentially expressed proteins were screened by two-sample t-test (P < 0.05) and fold change (>2), based on the bioinformatics analysis.

Data set analysis
MS/MS fragmentation spectra were analysed using PEAKS search engine tool (PEAKS Studio 7; Bioinformatics Solutions Inc., Waterloo, ON, Canada).
The protein details obtained were analysed for the function, process, location, by PANTHER (Protein ANalysis THrough Evolutionary Relationships) classification system (http://www.pantherdb.org/) [48] which is based on a controlled dialect to describe a protein regarding its subcellular localization, molecular function or biological process.
Top over-expressed proteins were used to create radar (spider) plots Using Microsoft Excel.

Results
This study utilizes a novel proteomic approach based on isobaric tags for relative and absolute quantification (iTRAQ) using nano liquid chromatography tandem mass spectrometry analysis to identify spe-  cifically over expressed proteins in TCs comparative to FBs. Figure 1 shows the sequence of connected procedural steps used in this protocol.
We identified a total of 1609 proteins by iTRAQ technology using nano LC-MS/MS analysis. The temporal proteome was evaluated at 5 and 10 days, respectively, and the selected lists of differentially upregulated proteins identified in TCs and FBs, in cell culture, are shown in Table 1 (5th day) and Table 2 (10th day).
TCs versus FBs, 5th day in cell culture Identification of the differentially expressed protein We found that in TCs, as compared to FBs, there are 39 up-regulated proteins, especially Myosin-14 (15.72-fold). We subsequently examined the FBs by comparison with TCs and found that there are 25 up-regulated proteins, especially collagen alfa 3(VI) chain (4.71fold), secernin-1 (3.73-fold), fascin (3.25-fold) as detailed in Tables S1 and S2.

Bioinformatic analyses of the identified proteins
The distributions of differentially abundant proteins in putative functional categories are shown in Figures 2-7 according to the PANTHER. The various molecular functions which were found to be mapped with the participation of the identified proteins in TCs are involved mainly in catalytic activity (43%) and as structural molecule activity (25%; Fig. 2) compared with FBs in which they are chiefly involved in binding (32%), catalytic activity (28%) and as structural molecule activity (28%; Fig. 5).
There is a big difference regarding system development functions between TCs and FBs: muscle organ development 50%, nervous system and skeletal muscle development 25% each in TCs (Fig. 3D) and 50% nervous system development and skeletal muscle and muscle organ development 25% each for FBs (Fig. 6D).
The cellular localization of TCs proteome demonstrated proteins from intermediate filaments (56%), actin cytoscheleton (19%), mitochondria (13%), microtubule (6%) and plasma membrane (6%; Fig. 4C). In FBs, 75% of the proteins belong to the cytoskeleton and 25% are destined to the extracellular matrix (Fig. 7C). Figure 8 represents the heat map based on the results of protein quantification and demonstrates that proteins are differentially expressed between TCs and FBs.
The cellular localization of TCs proteome demonstrated proteins from cytoscheleton (87%) and plasma membrane (13%; Fig. 11C). In FBs, 76% of the proteins belong to the cytoskeleton, 18% to the extracellular matrix and 6% to the plasma membrane (Fig. 14C).
The heat map showing the expression of differentially expressed protein between TCs and FBs is showed in Figure 15 and demonstrated that the differences between this two cell types are still maintained in cell culture after 10 days.

Discussion
The domain of proteomics enables comparative analysis of existing proteins in different cells [49]. In addition, it allows monitoring the changes in protein expression during the dynamic transformations that cells undergo during their differentiation [50]. For this reason, we performed a proteomic analysis of TCs compared to FBs, at different time-points (the 5th and 10th day in primary cell culture), to verify if these cells are different from one another and if there are any changes in their differentiation (phenotype) over time. After the identification of targeted proteins, we characterized and associated them with protein families using PANTHER database and analysed them from the functional point of view.
We identified a panel of 1609 proteins from human lung TCs and FBs cell cultures samples which were quantified and the highest fold change expressed proteins between TCs and FBs were analysed and classified using the protein database. According to our results, there were top 39 ranked proteins over expressed in TCs compared to FBs at 5th day in cell culture and 24 up-regulated proteins in TCs by comparison to FBs at 10th day in cell culture and this supports the case for TCs as distinctive cells.

Putative roles of differentially expressed proteins in TCs
In TCs, by comparison with FBs, the differentially identified proteins were mainly located in the cytoplasmic compartment and involved in cell signalling, energy and metabolic pathways, while a significant part of the FBs proteins were destined to the extracellular matrix which is in concordance with its well known function of producing extracellular matrix components (including collagen) [51]. Myosin-14, which was found to be up-regulated in TCs, is a conventional non-muscle myosin encoded by the MYH14 gene in human chromosome 19q13.3 [52]. We found that non-muscle myosin-14 seems to be involved in processes such as sensory perception, intercellular signalling and morphogenesis, according to the PANTHER classification system. Among previous assumptions on TCs functions it has been included that of mechanoreceptors [53], capable of detecting and translate stretch information to the nucleus [12,16]. The presence of cytoskeleton proteins and especially myosin-14 in TCs confirms this hypothesis. Taking into account that myosin-14 is known to be involved in sensory perception [54], we can propose TCs as candidates for a mechanical sensing and mechanochemical conversion task. In addition, the presence of TC primary cilia was reported in vasculature [55] and trachea [56], functioning as mechano-or chemosensors [57], probably involved in the initiation of cellular global positioning to initiate tissue renewal after damage [58].
Telocyte proteome also revealed the presence of periplakin. Periplakin, a protein which in humans is encoded by the PPL gene, links cytoskeleton elements together and connects them to junctional complexes [59]. Up till now, electron microscopy studies revealed that TCs establish homocellular (TC-TC junctions) and heterocellular junctions (TC-other cell type) [25,26]. Mechanical junctions are essential to the proliferation, migration and transformation of various cell types [60,61] and therefore we might suggest that TCs are involved in these mentioned processes. Also, intercellular signalling is pivotal in switching cells between different fates such as growth, differentiation and motility, and it could be influenced by tensional force generation within the cytoskeleton [62,63]. We suggest that TCs might participate in mechanical sensing and mechanochemical conversion task and also in tissue repair/remodelling/renewal, as shown before [38,39,64,65].
Telocytes proteomic analysis revealed the up-regulation of proteins with oxidoreductase activity, mostly located within mitocondria. It had been proved under electron microscope that TCs have calcium uptake/release units defined by a close relation between caveolae, endoplasmic reticulum and mithocondria [66], located in the podoms of telopodes [12,16,26]. It is well known that mitochondria and endoplasmic reticulum establish bilateral physiological interactions responsible (among numerous functions) for modulating the calcium signalling function, process which involves redox and redoxsensitive enzymes [67]. ER-mitochondria crosstalk is essential for eukaryotic cells preventing the onset of diseases by disrupted metabolism [68,69]. Moreover, Haines et al. suggested that the intensity of oxidative stress influences cell tissue composition towards desired cell type ratio, functioning as a cell death 'rheostat' [70]. Therefore, we consider that TCs are involved in the maintenance of cell and tissue homoeostasis.
Mammalian cells release extracellular vesicles produced by two mechanisms: (a) secretion from the endosomal membrane compartment after the fusion of multivesicular bodies with a plasma membrane and (b) shedding directly from plasma membrane [71]. Telocytes are no exception to this, the presence of exosomes and ectosomes being recently reported [7,16,26]. Several proteins up-regulated in TCs were found among the top 100 vesicular proteins that are present most frequently in mammalian extracellular vesicles proteome [72]. Among them we can exemplify with proteins such as: mitochondrial thioredoxin-dependent peroxide reductase, protein disulphide-isomerase A3, myosin-14, myosin-10, filamin-B, sodium/potassium-transporting AT-Pase subunit a-1 and keratin, type II cytoskeletal 1. We can assume that the release of extracellular proteins contributes to the extracellular environment homoeostasis, possibly influencing stem cell niches, leading to cell differentiation which is in congruence with very recent studies pointing out that TCs could function as an extensive intercellular information transmission system [73].

Conclusion
The present study represents the first proteomic analysis on TCs and will provide useful insights on the possible functions of these cells and involvement in lung (and not only) pathology. The data herein reported show that TCs are completely different from FBs, not only by their ultrastructural configuration [1,23], gene profile [42], immunophenotype [74], but also from the protein expression point of view. The data presented now are supporting for our previous assumptions regarding TCs functions in tissue morphogenesis, development and repair/remodelling, extracellular environment homoeostasis and intercellular signalling influencing stem cell niche fate. Table S1. Summary of up-regulated TCs proteins (more than twofold) versus FBs 5th day. Table S2. Summary of up-regulated FBs proteins (more than two-fold) versus TCs 5th day. Table S3. Summary of up-regulated TCs proteins (more than two-fold) versus FBs 10th day. Table S4. Summary of up-regulated FBs proteins (more than two-fold) versus TCs 10th day.