Introduction

Thymosin beta-4 (Tβ4) is a highly conserved polypeptide originally identified as T cell maturation factor in the thymus gland that occurs ubiquitously in different cells and tissues of many organisms [1, 2]. Tβ4 binds to cytoplasmic G-actin preventing its polymerization to F-actin, thereby regulating cytoskeletal organization and cell motility [3, 4]. Extracellular Tβ4 promotes a variety of physiological functions such as wound healing, chemotaxis, angiogenesis, and down regulates inflammation [5–8]. Tβ4 apparently possesses antimicrobial activities, facilitates antigen presentation, and it is one of the major genes up-regulated following immune activation [9, 10]. Although Tβ4’s many diverse extracellular effects have been demonstrated, the regulation of its synthesis and secretion to the extracellular environment are not understood [7, 11].

In the course of screening for bioactive peptides in phagocytic cells, we identified Tβ4 as an abundant peptide present in chicken macrophages [12]. The macrophages constitute a major component of innate immunity and participate in many functions attributed to Tβ4 such as, wound healing, angiogenesis, and tissue remodeling [13–15]. The macrophages recognize various microbial pathogens and their products through a series of ‘pattern recognition receptors (PRRs)’ called toll-like receptors (TLRs) [16] which cause their activation leading to the synthesis and secretion of various cytokines, metabolites, and enzymes that in turn mediate their different biological effects [13, 17]. Therefore, the objective of this study was to find whether TLR activation can induce Tβ4 production by the macrophages which in turn may participate in post-inflammatory healing processes and other biological events attributed to this peptide.

Materials and methods

Reagents and chemicals

Dialyzed fetal bovine serum (FBS) and l-arginine and l-lysine deplete SILAC (stable isotope labelling of amino acids in cell culture) RPMI-1640 cell culture media were purchased from Pierce (Rockford, IL). 13C1 or ‘heavy isotope (H)’ labeled L-lysine, was obtained from Cambridge Isotope Laboratories (Andover, MA). Micro BCA protein assay kit (Pierce, Rockford, IL), RNAeasy mini and on-column DNA digestion kits (Qiagen Corp, Chatsworth, CA), Retroscript reverse transcriptase kit (Ambion, Austin, TX), SYBR green PCR master mix (Applied Biosystems, Austin, TX), and lactate dehydrogenase (LDH) cytotoxicity kit (Promega, Madison, WI) were purchased from the respective vendors. All other chemicals including l-lysine (L), l-arginine, non-enzymatic cell dissociation medium, and the antibiotic antimycotic solution, were obtained from Sigma–Aldrich (St. Louis, MO).

Toll-like receptor ligands

The peptidoglycan-polysaccharide polymers PG-PS 10S (PGN), 5 mg rhamnose equivalent/ml of 0.85% saline, a sonicated cell wall preparation of Streptococcus pyrogenes, was a gift from BD Bioscience (San Jose, CA). A synthetic lipoprotein Pam3CSK4 (palmitoyl- 3-cysteine-serine-lysine-4; PAM), Salmonella typhimurium flagellin (FGN) and, guanine analog loxoribine (LOX) were purchased from Invivogen (San Diego, CA). The CpG-oligodeoxynucleotide (CpG-ODN 2006), corresponding to the sequence TCGTCGTTTTGTCGTTTTGTCGTT [18], was synthesized by Invitrogen (Carlsbad, CA). Salmonella typhimurium lipopolysaccharide (LPS) and dsRNA analog poly I: C was purchased from Sigma–Aldrich (St. Louis, MO). All TLR ligands were prepared as stock solutions in sterile water.

SILAC media

l-lysine and l-arginine were dissolved in sterile PBS at a stock concentration of 10 and 50 g/l, respectively, and filtered through a 0.2 μm filter. A stock concentration of 13C1-lysine (10 g/l) was similarly prepared and filtered. The “Light (L)” SILAC medium for the control cells was prepared by adding 2 mL of stock solution of l-arginine and 4 ml of l-lysine to 500 ml of depleted RPMI 1640 medium. The “Heavy (H)” SILAC medium was similarly prepared by adding 2 ml of l-arginine and 4 ml of 13C1-lysine stock solutions to 500 ml of depleted RPMI 1640 medium. Both ‘L’ and ‘H’ SILAC media were filtered and supplemented with 10% dialyzed FBS, 20 mM glutamine, and 1× concentrations of antibiotic antimycotic solution.

Macrophage culture and activation by TLR agonists for SILAC studies

Transformed chicken HTC macrophage [19] cultured in RPMI-1640 media containing 10% FBS and 1× concentration of antibiotic antimycotic solution were initially grown in normal RPMI 1640 medium. At ~80% confluence, the media was aspirated and the cells were detached using a non-enzymatic dissociation medium for 5 min, and washed with PBS three times by successive centrifugation at 380×g for 8 min each [12]. One half of the detached cells were reconstituted in ‘L’ RPMI medium and the other half in the ‘H’ SILAC RPMI medium and grown in separate flasks. Metabolic labeling of cells were done according to Ong and Mann [20] with six successive passages of cells in their respective media that theoretically labels the entire proteome of the cells with 90–100% efficiency. Both L and H labeled HTC cells were then detached and plated in triplicates at a concentration of 1 × 106 cells/ml/well in 12 well culture plates in their respective media and grown overnight at 37°C under 5% CO2. The plates were then centrifuged at 100 g and replaced with fresh media for respective experiments. Two time point experiments were performed consisting of 6 and 24 h stimulation with TLR ligands. In both cases, the H labeled cultures, were stimulated with the following TLR cognate agonists: TLR2, gram-positive bacterial peptidoglycan (PGN); TLR3, poly inosinic, cytidylic (poly I: C); TLR4, gram-negative LPS; TLR5, flagellin (FGN); TLR2/1, PAM3CSK4; TLR7, loxoribine (LOX), and TLR9 or avian equivalent TLR 21, CpG-ODN [16, 21]. The final concentrations of the agonists, PAM (1 μg/ml), PGN (1 μg rhamnose equivalent/ml), FGN (100 μg/ml), poly I: C (1 μg/ml), LOX (100 μM), CpG-ODN (5 μM), and LPS (1 μg/ml), were based on the values recommended by the suppliers or earlier literature [22, 23]. Sets of H labeled and all L labeled cultures were used as respective controls without the addition of TLR agonists (Fig. 1).

Fig. 1
figure 1

A flow chart of SILAC method determining the effects of TLR ligands on Tβ4 production

Full size image

Sample preparation for MALDI-MS

After 6 and 24 h of stimulation with TLR agonists, the supernatant from each well were removed and centrifuged at 250×g for 10 min and the cell free conditioned media were saved for further extraction. Respective pelleted cells were added to the cells in the wells and lysed in 200 μl of 0.1% n-octyl β-glucopyranoside (OβG) by repeated trituration, and centrifuged at 21,000×g for 10 min to obtain the clear cell lysate. Aliquots of the lysates were diluted for further experiments. The conditioned media were extracted by mixing with equal volumes of 100% methanol containing 2% acetic acid to denature and precipitate proteins over a 16 h period at 4°C, and centrifuged at 21,000×g for 10 min to obtain the supernatant for Tβ4 analysis. The cell lysate and conditioned media extract of untreated control (C) cultures consisted of L and H labeled extract (CL and CH) whereas the TLR agonist treated groups (T) consisted of only H labeled samples (TH), based on labeling with their respective light or heavy isotope containing lysine (Fig. 1). Protein concentrations of both conditioned media extract and cell lysates were determined using micro BCA protein assay and normalized to equal concentrations in case of variations between different samples. The triplicates of the H labeled control and treatment samples (CH and TH, cell lysate and conditioned media extract) were used individually but the L labeled control samples (cell lysate and conditioned media extract) were pooled and used for respective mixing experiments (Fig. 1).

Aliquots of each extract was mixed with an equal volume of one molar 2,5-dihydroxybenzoic acid (DHB) in 90% methanol containing 0.1% formic acid and 2 μl of each was spotted onto a Bruker MTP 384 stainless steel MALDI target for preliminary screening. MALDI-TOF spectra was acquired over the m/z range 1–10 kDa in the positive ion reflector mode using a Bruker Reflex III MALDI-TOF mass spectrometer (Bruker Daltonik GMBH, Bremen, Germany) [12, 24]. The data were processed using Bruker Flex Analysis 2.4/3.0 software. Initial screening of the cell lysate and conditioned media extract samples of CL, CH, and TH was done to detect the presence of Tβ4 (m/z 4963, protonated molecule in CL and m/z 4,972, protonated molecule in both CH and TH corresponding to 9 deuterium atoms present in 9 heavy labeled lysine residues. Following preliminary identification, pooled control CL and individual CH or TH were mixed 1:1 to determine the intensity ratios of L and H Tβ4 associated spectra in the cell lysate or their respective conditioned media extracts. The intensity ratios (CH/CL) for controls and (TH/CL) for all treatment groups were determined by triplicate spotting of each sample resulting in nine observations per treatment group. The change in the intracellular Tβ4 content was calculated by dividing the intensity ratio TH/CL over CH/CL. The mean and the standard error of the mean (SEM) was calculated for controls and each treatment group using measurements obtained from individual sets of experiments calculated in triplicate at both 6 and 24 h periods (Fig. 1). The results were evaluated using Students t test and a P value ≤ 0.05 was considered significant. The quantitative assays with conditioned media could not be performed, since most conditioned media extract did not show any detectable Tβ4 even after concentrating the respective samples. However, in groups which showed cellular depletion, the released Tβ4 peptides were detectable in the corresponding conditioned media and identified qualitatively.

Activation of macrophages by TLR agonists for gene expression studies

In separate set of experiments, the HTC cells were seeded in triplicates at 1 × 106 cells/ml/well in RPMI 1640 media containing regular L-lysine as above and treated with TLR agonists for 6 h to determine the macrophage activation-induced changes in the gene expression and 24 h to collect supernatant for nitrite and lactate dehydrogenase (LDH) determinations. Total RNA was extracted from control and TLR agonist treated macrophage cultures using RNAeasy mini kit. Any contaminating DNA was subjected to on column digestion with RNase-free DNase 1 according to the manufacturer’s instructions. The cDNA was synthesized from RNA using Retroscript reverse transciptase kit. The gene-specific primers for Tβ4, IL-6, IL-1β, iNOS, and the β-actin were designed with the Primer 3 software (http://frodo.wi.mit.edu/) (Table 1). The expression of genes for IL-6, IL-1β, and iNOS was used as positive indicators of macrophage activation, and β-actin as the reference. Quantitative polymerase chain reaction (qPCR) was done using two individual samples in each treatment group to determine the expression of target genes relative to β-actin. Real-time qPCR was performed following optimization of primer concentration and an ABI Prism 7700 Sequence Detector (Applied Biosystems, Austin, TX). Each mRNA sequence was amplified in duplicate in 25 μl reaction mixtures containing SYBR green PCR master mix (Applied Biosystems, Austin, TX); cDNA corresponding to 1 μg of reverse transcribed RNA, and 200 nM (optimized concentration) of forward and reverse primers. Both a no template and no reverse transcriptase control were included for each amplification reaction, and the homogeneity of the amplified products was confirmed routinely by melting curve analysis. The results were analyzed by the standard curve method and normalized to β-actin gene used as an endogenous control. The fold changes in the expression of target genes (Tβ4, IL-6, IL1β, and iNOS) in TLR agonist stimulated macrophages were determined relative to non stimulated control cultures.

Table 1 Primer sequences for target genes used in qPCR [12, 19]
Full size table

Nitrite and LDH measurement

The nitrite content of aliquots of conditioned media from different treatment groups was measured using Griess reagent [19]. The percent changes in nitrite content were calculated relative to the untreated control. The LDH activity of the same conditioned media from both 6 and 24 h cultures were measured used to assess cell damage or death [25] using a LDH cytotoxicity assay kit according to the manufacturer’s protocol. The LDH activity was expressed as arbitrary fluorescence units (AU).

Statistics

All the quantitative results were evaluated using Students t test. A P value ≤ 0.05 was considered significant.

Results

TLR activation-induced Tβ4 release

MALDI-MS profiles of 1:1 mixtures of L and H cell extracts or similarly treated conditioned media extracts are shown in Figs. 2, 3, and 4. The figures represent the region of the spectra corresponding to the occurrence of Tβ4. Protonated ion m/z 4963 corresponds to Tβ4 peptide (12) while m/z 4972 corresponds to 9 residues of 13C1 labeled H lysine present in Tβ4. Signal intensities of L and H Tβ4 were monitored at 6 h and 24 h time points. Figure 2 (top panel) shows the MALDI-MS spectrum of a 1:1 mixture of L and H labeled control cell lysates whereas bottom panels represent spectra from 1:1 mixture of control (L) and respective TLR agonist stimulated H cell lysates at 6 h. The signal intensity of L and H labeled Tβ4 remained approximately the same in all groups in cell lysates at 6 h where as in the corresponding conditioned media extract it was below the detection limit (data not shown). At the 24 h time point, the cell-associated Tβ4 showed significant decreases in PGN, PAM, and LPS treated H cell lysates while CpG and FGN treatments showed moderate decreases with poly I: C and LOX showing no changes (Fig. 3). The significant decreases in the intensities of H Tβ4 detected in PGN, PAM, and LPS groups, was consistent with the detection of Tβ4 in their respective conditioned media both as intact peptides and their respective oxidized forms (Tβ4 sulfoxide L: m/z 4979, H: m/z 4988; Fig. 4). The oxidized form appeared to be dominant H labeled Tβ4 in the conditioned media (Fig. 4). The relative changes in cellular Tβ4 in macrophages treated with different TLR agonists at 6 and 24 h are shown in Fig. 5. While the cellular Tβ4 showed no change by any treatment at 6 h, at 24 h time point the PGN, PAM, and LPS caused significant decreases in cellular Tβ4 levels (P ≤ 0.001) with CpG and FGN showing a moderate albeit statistically significant decreases (P ≤ 0.05).

Fig. 2
figure 2

MALDI MS profiles of 1:1 mixture of cell lysate extracts containing light (L) and heavy (H) lysine labeled Tβ4 of representative samples after 6 h stimulation with TLR agonists

Full size image
Fig. 3
figure 3

MALDI-MS profiles of 1:1 cell lysate mixtures of 24 h stimulated samples containing light (L) and heavy (H) lysine labeled Tβ4

Full size image
Fig. 4
figure 4

MALDI-MS profiles of 1:1 mixtures of supernatant extracts from 24 h control, PGN, PAM, and LPS treated cultures, showing light and heavy lysine labeled Tβ4 (L: m/z 4963 and H: m/z 4972) and their corresponding oxidized forms of Tβ4 (Tβ4 sulfoxide L: m/z 4979, H: m/z 4988)

Full size image
Fig. 5
figure 5

Relative changes in the cellular Tβ4 levels induced by different TLR agonists at 6 and 24 h. Data represent the mean value ± SEM (n = 9). ** Indicates P ≤ 0.0001 and * is P ≤ 0.005

Full size image

Gene expression

qPCR results of the expression of Tβ4, IL-1β, IL-6, and iNOS mRNA relative to β-actin is shown in Fig. 6. Tβ4 mRNA expression did not change significantly relative to untreated control group. In contrast, PAM, PGN, LPS, CpG, and FGN, all induced a significant up-regulation of IL-1β, IL-6, and iNOS. The FGN produced a lesser effect on comparison with other agonists whereas both poly I:C and LOX agonists showed minimal effects that were not significantly different from control (Fig. 6).

Fig. 6
figure 6

qPCR gene expression results after 6 h stimulation of HTC cells by different TLR agonists. The results represent an average of duplicate experiments

Full size image

Functional activation of macrophages by TLR agonists

Cultures treated with TLR agonists, PGN, LPS, PAM, FGN, and CpG, showed changes in nitrite content of the conditioned media at both time points with much higher accumulation by 24 h relative to their controls. Except for poly I; C and LOX, rest other treatment groups showed a significant accumulation of nitrite at 24 h time points (Fig. 7).

Fig. 7
figure 7

TLR agonists induced changes in nitrite production by the HTC cells at 6 and 24 h. * Indicates P ≤ 0.0001

Full size image

LDH changes

There were no differences in LDH activities in the conditioned media between control and any TLR agonist treated cultures at 6 h time point but at 24 h, the LDH activity was significantly increased in all treatment groups except poly I: C and LOX. The conditioned media in all including that of control groups had higher LDH activities at 24 h compared with 6 h time point (Fig. 8).

Fig. 8
figure 8

LDH activity of cell conditioned media at 6 and 24 h. ** Indicates P ≤ 0.0001 and * is P ≤ 0.005

Full size image

Discussion

Thymosin β4 is a ubiquitously occurring peptide in eukaryotic cells which exerts a variety of extracellular effects including wound healing, angiogenesis, and tissue remodeling [6, 7, 26, 27]. Despite its diverse extracellular actions, the mechanism of its release into the extracellular fluid remains unclear. The Tβ4 also lacks a signal sequence which is essential for most secretory proteins and peptides [11]. However, the presence of Tβ4 in blister and wound fluids as well as in the conditioned media derived from bone marrow endothelial and myocardial cell cultures has suggested its secretion into the extracellular environment [28–30]. Our rationale to study the Tβ4 release was its abundance in macrophages [12, 31, 32] and these cells, when subjected to immunomodulation by TLR-agonists show increase in their secretory activities [33]. SILAC is a very sensitive and quantitative method to measure changes in protein and peptide expressions using the corresponding MALDI MS signal ratio of the light (L) and heavy (H) labeled peptide ions [34, 35]. Addition of 1 Da for each lysine residue due to 13C1 in H labeled peptide, compared to the natural abundance 12C labeled L, showed the expected overall 9 Da shift in mass consistent with the 9 lysine residues in mature chicken Tβ4. The ratios of the MALDI signals of these ions were measured to determine the expression differences. Our results showed no discernible changes in either cellular or conditioned-media associated Tβ4 at 6 h in any treatment group. In contrast, a significant decrease in cellular Tβ4 content was observed at 24 h in conditioned media of cultures treated with PGN, LPS, PAM, or CpG. Interestingly, not only L and H Tβ4 but their corresponding oxidized forms (Tβ4 sulfoxide) resulting from oxidation of single methionine residue showing a molecular mass increase of 16 Da were also detected in 24 h conditioned media [12, 36]. The H Tβ4 sulfoxide appeared to be the dominant species. The Tβ4 sulfoxide has been suggested as an anti-inflammatory mediator produced by the monocytes and macrophages by the action of glucocorticoids [36–38]. But, the mechanisms favoring the oxidation of Tβ4 to its sulfoxide form in conditioned media in the absence of glucocorticoids is not clear. It is likely that the phagocytic HTC cells which show respiratory burst when challenged with stimulants such as LPS [19] contribute to the formation of Tβ4 sulfoxide by extracellular oxidation of Tβ4.

The lack of any substantial effects of LOX, FGN, and poly I:C on HTC cells suggest that either these cells lack their cognate receptors or the concentrations were not sufficient enough to induce their activation. Our choice of TLR agonists and their effective concentrations were primarily based on literature reports. Based on the 24 h results, it is nonetheless, evident that at least, some TLR agonists are able to induce Tβ4 release into the culture media implying that similar mechanisms may be operative in vivo during inflammation and tissue remodeling events. Hence, we asked if Tβ4 was released by secretion induced by those agonists, then it should also be replenished in the cells and would be evident by the changes in its gene expression. The gene activation being an early event, we monitored its expressions along with the changes in the expressions of IL-1β, IL-6, and iNOS as positive indicators of macrophage activation, at 6 h. The changes in the nitrite levels in cell culture supernatants at both time points were also measured. The results showed that none of the TLR agonists modulated the expression of the Tβ4 gene implying that its expression may not be controlled by the factors that activate macrophages. The changes in the expression of IL-1β, IL-6, and iNOS genes and the release of nitrite into the conditioned media affected by PGN, LPS, PAM, and CpG, is consistent with HTC cell activation [19]. The mRNA expression is conventionally regarded to reflect the changes in their respective proteins or peptides. Therefore, the absence of any detectable change in Tβ4 gene expression coupled with the observed intracellular and extracellular changes in the levels of this peptide was intriguing. Alternatively, the possibility remained that the cell damage could cause the release of Tβ4 into the conditioned media. Lactate dehydrogenase (LDH) is a cytoplasmic enzyme, the leakage of which into the extracellular medium has been used as an indicator of cell damage and the loss of cell viability [25]. To assess whether some of the TLR agonists mediate the release of Tβ4 into the culture media by causing cell damage, we measured the LDH changes in cell culture supernatants. There were significant increases in LDH levels of the culture media treated with LPS, PGN, PAM, and CpG. These were the same treatment groups which showed the depletion of cellular Tβ4 at 24 h and its detection in the conditioned media extracts. Therefore, it appears that macrophage death induced by the action of these TLR agonists may be a possible explanation for the release of Tβ4 into the extracellular environment. This assumption is consistent with the fact that both gram-positive and gram-negative microbial products such as PGN and LPS do induce macrophage cell damage and death [39–41].

In conclusion, these results demonstrate that one of the mechanisms through which the immunomodulatory and inflammatory agents mediate the release of Tβ4 into extracellular environment is by inducing the damage and death of activated cells that subsequently helps repair processes. We believe these results provide a mechanism for Tβ4 release into wound environment and its role in post inflammatory healing.