Tuftsin signals through its receptor neuropilin-1 via the transforming growth factor beta pathway


  • Jillian C. Nissen,

    1. Program in Molecular and Cellular Pharmacology, Department of Pharmacological Sciences, Stony Brook University, Stony Brook, New York, USA
    Search for more papers by this author
  • David L. Selwood,

    1. The Wolfson Institute for Biomedical Research, UCL Gower Street, London, UK
    Search for more papers by this author
  • Stella E. Tsirka

    Corresponding author
    1. Program in Molecular and Cellular Pharmacology, Department of Pharmacological Sciences, Stony Brook University, Stony Brook, New York, USA
    • Address correspondence and reprint requests to Stella E. Tsirka, Department of Pharmacological Sciences, Stony Brook University, Stony Brook, NY 11794-8651, USA. E-mail: styliani-anna.tsirka@stonybrook.edu

    Search for more papers by this author


Tuftsin (Thr-Lys-Pro-Arg) is a natural immunomodulating peptide found to stimulate phagocytosis in macrophages/microglia. Tuftsin binds to the receptor neuropilin-1 (Nrp1) on the surface of cells. Nrp1 is a single-pass transmembrane protein, but its intracellular C-terminal domain is too small to signal independently. Instead, it associates with a variety of coreceptors. Despite its long history, the pathway through which tuftsin signals has not been described. To investigate this question, we employed various inhibitors to Nrp1's coreceptors to determine which route is responsible for tuftsin signaling. We use the inhibitor EG00229, which prevents tuftsin binding to Nrp1 on the surface of microglia and reverses the anti-inflammatory M2 shift induced by tuftsin. Furthermore, we demonstrate that blockade of transforming growth factor beta (TGFβ) signaling via TβR1 disrupts the M2 shift similar to EG00229. We report that tuftsin promotes Smad3 phosphorylation and reduces Akt phosphorylation. Taken together, our data show that tuftsin signals through Nrp1 and the canonical TGFβ signaling pathway.


Despite the 40-year history of the tetrapeptide tuftsin (TKPR), a macrophage and microglial activator, its mechanism of action has not been defined. Here, we report that the tuftsin-mediated anti-inflammatory M2 shift in microglia is caused specifically by tuftsin binding to the receptor neuropilin-1 (Nrp1) and signaling through TGFβ receptor-1, a coreceptor of Nrp1. We further show that tuftsin signals via the canonical TGFβ pathway and promotes TGFβ release from target cells.

Abbreviations used



hepatocyte growth factor


neuronal conditioned media






transforming growth factor beta


TGFβ receptor 1


VEGF receptor


vascular endothelial growth factor

Tuftsin is a small, naturally occurring tetrapeptide with the sequence threonine-lysine-proline-arginine. It was originally described at its discovery in 1970 as a phagocytosis-stimulating factor derived from the proteolytic degradation of IgG (Najjar and Nishioka 1970; Nishioka et al. 1973). In cells of monocytic origin such as macrophages, microglia, and neutrophils, tuftsin promotes phagocytic activity. It can also stimulate migration and chemotaxis, as well as promote the antigen presentation of these cell types which ultimately aids in adaptive immune function through T cell populations (Siemion and Kluczyk 1999). In many animal disease models, such as sepsis (Wardowska et al. 2009) and multiple sclerosis (Bhasin et al. 2007), tuftsin treatment has been associated with anti-inflammatory effects, particularly among microglial cell populations (Wu et al. 2012).

Receptors for tuftsin have been described in the literature; using affinity chromatography, two binding activities were reported in peritoneal granulocytes corresponding to 250 and 500 kDa molecular masses consisting of two subunits of 62 and 52 kDa (Bump et al. 1986). More recently, neuropilin-1 (Nrp1) was identified as a receptor for tuftsin (von Wronski et al. 2006). Nrp1 is a single-pass transmembrane glycoprotein that plays important roles in development, immunity, and cancer. Nrp1 shares 44% homology with neuropilin-2, with which it shares many structural and biological properties; however, tuftsin has been shown to bind only to Nrp1 (von Wronski et al. 2006). During angiogenesis, Nrp1 binds vascular endothelial growth factor (VEGF) and promotes vessel formation (Soker et al. 1998; Gu et al. 2003), which is exploited by tumor cells to provide a blood supply to cancerous tissues (Klagsbrun et al. 2002). In neural development, Nrp1 binds class 3 semaphorins (Sema3), which provide inhibitory signals in axonal guidance (Kitsukawa et al. 1995; Gu et al. 2003). Nrp1 also plays an important role in the immune system by promoting long contacts between dendritic cells and immunosuppressive regulatory T cells (Treg) (Tordjman et al. 2002; Sarris et al. 2008), and can bind and activate transforming growth factor beta (TGFβ), an anti-inflammatory cytokine (Glinka and Prud'homme 2008). Recently, Tregs have been shown to be recruited to tumors by Nrp1 in a VEGF-dependent manner, which contributes to immune system evasion by tumor cells (Hansen 2013).

A notable feature of Nrp1's structure is that it has a short cytoplasmic domain that is only about 44 amino acids long, and contains no known signaling motif (Vander Kooi et al. 2007). Although there have been several studies indicating that this domain is capable of signaling independently, the mechanism has not been well defined (Evans et al. 2011; Fantin et al. 2011). However, Nrp1 can associate with various coreceptors via which downstream signaling can occur; PlexinA1 for Sema3, VEGF receptor (VEGFR) for VEGF, TGFβ receptor 1 (TβR1) for TGFβ, and c-Met for hepatocyte growth factor (HGF) (Prud'homme and Glinka 2012). Tuftsin competes with VEGF for binding on the Nrp1 molecule, as it shares sequence homology with the C-terminus of VEGF (von Wronski et al. 2006). The structure of VEGF allows its N-terminus to interact with VEGFR while bound to Nrp1 (Wiesmann et al. 1997), promoting coreceptor dimerization and signaling, which is not the case with tuftsin because of its small size. Although tuftsin has been used in various studies for over forty years, its mechanism of action and signaling pathway are still unknown.

In this study, we show that the Nrp1 inhibitor EG00229 prevents tuftsin binding to the cell surface, and effectively blocks the anti-inflammatory shift induced by tuftsin in microglial cells. We also demonstrate that an inhibitor of TβR1 function, but not of c-Met function, disrupts tuftsin's downstream effects similar to EG00229. Moreover, we show that tuftsin signals through the canonical TGFβ signaling pathway.

Materials and methods


C57BL/6 mice were bred in-house (SBU Animal Facility) under specific pathogen-free conditions set by the Division of Laboratory Animal Resources at Stony Brook University. The environment was controlled for temperature (21°C), and maintained under a 12-h light/dark cycle. Access to food and water was ad libitum. All procedures were approved by the IACUC committee at Stony Brook University in compliance with the ARRIVE guidelines. Mice of both sexes were used.

Mixed cortical and primary microglia cultures

Microglia were isolated as previously described (Rogove and Tsirka 1998). In short, newborn (d0-d2) pups of wild-type mice were used to isolate cortical cells. The brains were removed, and cortices were freed from meninges, hippocampi and basal ganglia digested in 0.25% Trypsin/EDTA (Sigma-Aldrich, St. Louis, MO, USA) at 37°C for 20 min. To obtain a single-cell suspension, the tissue was then triturated and filtered through a 40-μm cell strainer, and plated in mixed cortical medium (Dulbecco's modified Eagle's medium , 10% fetal bovine serum , 40-μg/mL Gentamycin). Tissue culture plates used for plating mixed cortical cultures were coated overnight at 4°C with 5 μg/mL poly-d-lysine (PDL; Sigma).

The medium was changed 3 days after plating. Microglial cells were harvested 10 days after plating. Briefly, lidocaine was added directly to the culture medium at a final concentration of 1 mM and the culture left at 21°C for 15 min. The medium containing the floating microglia was collected and centrifuged at 500 g for 5 min, following which the cell pellet resuspended in microglia medium (Dulbecco's modified Eagle's medium, 1% fetal bovine serum) and counted on a hemocytometer. Cultures are > 98% pure, as previously described (Rogove and Tsirka 1998; Siao and Tsirka 2002; Yao and Tsirka 2010).

Inhibitors used to treat microglia were diluted in dimethylsulfoxide , and were EG00229, a Nrp1 inhibitor (Jarvis et al. 2010); LY 364947, a TβR1 inhibitor; and PF 04217903, a c-Met inhibitor (Tocris, Bristol, UK). The concentrations were chosen based on previous published work utilizing these compounds (Shiou et al. 2006; Jarvis et al. 2010; Jia et al. 2010; Zou et al. 2012). Microglial cells were plated at a density of 5 × 105 cells per 35 mm plate, and NCM was utilized in a 1 : 1 ratio to microglial culture medium.

Primary neuronal cultures

Primary neuronal cultures were prepared from embryonic day 17–19 mice as previously described (Siddiq and Tsirka 2004). Briefly, mouse cortices were dissected and put in Hanks solution (HBSS), and triturated to form single-cell suspensions. The cells were plated at a density of 100 000 cells/cm2 in Neurobasal medium with B27 supplements, 25 μM glutamate, 0.5 mM l-glutamine, and 10 g/L gentamycin sulfate. Tissue culture plates used for plating neuronal cultures were coated overnight at 4°C with 5 μg/mL PDL. Cells were utilized for experiments after 7 days in vitro.

Neuronal conditioned media (NCM) was prepared from primary cortical neurons exposed overnight to 100 μM glutamate added directly to the culture medium to induce excitotoxic injury. Prior to treating microglia with NCM, the media was spun down to remove debris.

Tuftsin binding assay

To observe tuftsin binding to the cell surface in the presence or absence of Nrp1 inhibitor EG00229, primary microglia cells were exposed to varying concentrations of inhibitor at 37°C for 1 h, and then treated with 100 μg/mL of biotinylated tuftsin (Genscript, Piscataway, NJ, USA) for 30 min at 21°C to minimize internalization of the peptide. After thorough washing, tuftsin's presence on the cell surface was observed by fixation of the cells and staining with Cy3-conjugated streptavidin.


Cells used for immunofluorescence were fixed for 30 min at 21°C in 4% paraformaldehyde, but not permeabilized. After washing with phosphate-buffered saline (PBS), they were blocked in serum of the host of the secondary antibody (5% serum in PBS with 0.5% TritonX-100), and then incubated overnight at 4°C with rabbit anti-mouse Nrp1 (Abcam, Cambridge, MA, USA) at a 1 : 500 dilution in PBS to detect tuftsin's receptor. After washing with PBS, cells were incubated with fluorescence-conjugated FITC goat anti-rabbit secondary antibody and Streptavidin-conjugated Cy3 (to detect surface bound biotinylated tuftsin) for 1 h at 21°C to minimize potential internalization, washed three times with PBS, and mounted using Fluoromount-G with 4′,6-diamidino-2-phenylindole (Southern Biotech, Birmingham, AL, USA). These were then visualized with a Nikon Eclipse E600 microscope (Nikon Instruments, Melville, NY, USA).


Cells were lysed in 50 mM Tris-HCl (pH 7.4) containing 1% Nonidet P-40, 0.25% Na-deoxycholate, 150 mM NaCl, 1% sodium dodecyl sulfate, and protease inhibitors cocktail (Sigma-Aldrich). After incubation on ice for 10 min, the lysates were centrifuged at 16 100 g for 10 min to remove debris, and the supernatant was collected. The extracts were separated on a reducing 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis SDS-PAGE gel, and blotted to polyvinylidene difluoride membrane (Immobilon-P; Millipore, Billerica, MA, USA). Membranes were then incubated overnight at 4°C with anti-phospho-Smad3 or anti-phospho-Akt (1 : 1000; Cell Signaling, Danvers, MA, USA) primary antibodies. These were then incubated with horseradish peroxidase-conjugated anti-rabbit secondary for 1 h at 21°C, and visualized with ECL (Pierce, Rockford, IL, USA). After stripping, membranes were reprobed with anti-Smad3 or anti-Akt antibodies (1 : 1000; Cell signaling).

RNA isolation and quantitative real-time PCR

To prepare RNA, primary microglia that were treated for 10 h were washed with PBS and lysed with RNA-Bee (Tel-Test, Friendswood, TX, USA), by the manufacturer's protocol. To obtain cDNA, one microgram of RNA was reverse transcribed on a Veriti Thermocycler (Applied Biosystems, Grand Island, NY, USA) using the High Capacity cDNA Reverse Transcription kit. Amplification was performed on a StepOnePlus real-time PCR machine using a SYBR Green kit (Applied Biosystems). Primer sequences are as follows: GAPDH forward, 5′-GCACAGTCAAGGCCGAGAAT-3′; GAPDH reverse, 5′-GCCTTCTCCATGGTGGTGGA-3′; IL10 forward, 5′-TGGCCACACTTGAGAGCTGC-3′; IL10 reverse, 5′-TTCAGGGATGAAGCGGCTGG-3′; TNFα forward, 5′-ATGAGCACAGAAAGCATGATC-3′; TNFα reverse, 5′-TACAGGCTTGTCACTCGAATT-3′. For gene expression analysis, the relative quantitation method was used (ΔΔCt) with GAPDH as an internal control.

Measurement of cytokine and NO levels

Primary microglial cells were treated with combinations of NCM, 100 μg/mL tuftsin, 10 μM TβR1 inhibitor, 5 nM c-Met inhibitor, or 30 μM EG00229 for 10 h.

To measure TGFβ levels, the Ready-Set-GO TGFβ cytokine ELISA kit from eBioscience (San Diego, CA, USA) was utilized. Briefly, 96-well plates were coated overnight with the appropriate dilution of capturing antibody in coating buffer (0.2 M sodium phosphate buffer pH 6.5). The plate was washed three times with PBS-T (PBS, 0.05% Tween-20) and blocked with Assay Diluent for 1 h at 21°C. To activate latent TGFβ1, samples were treated with 20 μL of 1 N HCl per 100 μL of sample for 10 min, and then neutralized with 1 N NaOH. After washing as before, 100 μL of sample or cytokines standard prepared in assay diluent were added followed by 2-h incubation at 21°C. After five washes with PBS-T, 100 μL of assay diluent containing biotin-conjugated detection antibody and Avidin-horseradish peroxidase reagent at the appropriate dilutions were added and incubated for 1 h at 21°C. Following seven washes with PBS-T, 100 μL of Substrate Solution was added to each well. After 30 min incubation in the dark, 50 μL of Stop Solution (2 N H2SO4) were added and absorbance at 450 nm was read within 30 min on a SpectraMax microplate reader using the Softmax Pro software (Molecular Devices, Sunnyvale, CA, USA) . Final readings were calculated with a dilution factor of 1.4 to account for acid activation/neutralization.

To observe nitric oxide levels, a nitrite assay was performed. Nitrite level measurement was based on the reaction of nitrites with 2,3-diaminonaphthalene (DAN) under acidic conditions, which results in 2,3-naphthyltriazole formation. Briefly, 100 μL of sample or standard (NaNO2) was loaded in a black 96-well plate with a clear bottom. 20 μL of DAN (0.05 mg/mL in 0.62 M HCl) was added to each well. After proceeding at 21°C for 20 min, the reaction was terminated by the addition of 100 μL 0.28 M NaOH. After an additional 10-min incubation at 21°C, fluorescence was measured on a Titertek Fluoroscan II fluorescence plate reader using a filter pair with excitation at 355 nm and emission at 460 nm.


For multiple comparisons within a group, statistical analysis was performed using one-way anova followed by a Bonferroni–Dunn test. For comparisons between groups, a two-tailed t-test was used, as indicated by the figure legends. For all figures, p < 0.05 was considered significant and is marked by *p < 0.01 and p < 0.001 was considered very significant and is marked by ** and ***, respectively. All results are represented as average with error bars indicating the standard error of the mean. In all experiments, n refers to the number of biological replicates used for each condition.


Neuropilin-1 inhibitor EG00229 prevents tuftsin binding to the cell surface

Tuftsin has been shown to bind to Nrp1 on endothelial cells (von Wronski et al. 2006). To examine if tuftsin is acting through Nrp1 on microglia, we used a small molecule inhibitor for VEGF binding to Nrp1 designated as EG00229. This inhibitor was designed around the final C-terminal residues of the VEGF molecule (DKPRR), as these were the residues that interacted with the Nrp1 binding pocket (Jarvis et al. 2010). As tuftsin shares a strong homology with the C-terminus of VEGF (TKPR), EG0029 can function effectively as an inhibitor of tuftsin binding to Nrp1 as well.

Microglia readily express Nrp1, as shown in Fig. 1. As there are no antibodies that can stain for tuftsin because of its small size, and conjugation with GFP could potentially disrupt its binding properties, we used biotinylated tuftsin that was then detected by streptavidin-conjugated Cy3 antibody. Primary microglia were treated with a combination of EG00229 and the biotinylated tuftsin. At all concentrations tested, the inhibitor seems not activate microglia as they remained in a ramified, resting state. Tuftsin binding to Nrp1 was significantly decreased by EG00229 (Fig. 1a), and appears to act in a dose-dependent manner (Fig. 1b). This result demonstrates that tuftsin's function on microglial cells is mediated by Nrp1 exclusively, and not by some alternative receptor (Bump et al. 1986), as its binding is prevented by the highly specific Nrp1 inhibitor.

Figure 1.

EG00229 reduces tuftsin binding to the surface of microglia in a dose-dependent manner. Microglial cells were pre-blocked with the indicated concentrations of EG00229, and incubated with biotinylated tuftsin at 21°C as described in Methods. Cells were then fixed and stained for Nrp1 (FITC) and with streptavidin-conjugated Cy3 to visualize biotinylated tuftsin on the cell surface (a). Fluorescence intensity of tuftsin staining was quantified in (b). n = 3; ***p < 0.001.

EG00229 blocks the anti-inflammatory shift in microglia induced by tuftsin

When activated, microglia can be polarized to either a pro- or anti-inflammatory subset known as M1 or M2, respectively. M1 microglia, which are neurodegenerative in a model of spinal cord injury, produce TNFα and nitric oxide, whereas neuroprotective M2 microglia release IL10 and TGFβ (Kigerl et al. 2009; Michelucci et al. 2009; Gordon and Martinez 2010). We previously reported that a ‘two-hit’ treatment with a combination of neuronal conditioned media (NCM), isolated from neurons treated overnight with 100 μM glutamate to induce excitotoxic injury, and tuftsin reduced the release of TNFα and promoted the release of IL10 in primary microglial cells, indicating an M2 shift in response to tuftsin treatment (Wu et al. 2012).

We wanted to examine whether EG00229 could prevent this tuftsin-mediated M2 microglial shift. We treated microglial cells for 10 h with NCM in the presence or absence of tuftsin and increasing concentrations of EG00229, choosing our inhibitor concentrations based on previous studies (Jarvis et al. 2010; Jia et al. 2010). We then harvested RNA and performed quantitative real-time PCR to observe microglial phenotype based on TNFα levels to indicate M1 polarization, and IL10 levels to indicate M2 polarization. While the combination of NCM and tuftsin reduced TNFα levels and increased IL10, as we have previously shown (Wu et al. 2012), EG00229 reversed these effects (Fig. 2a and b). While tuftsin and NCM alone significantly increase IL10 levels by about threefold, EG00229-treated cells at all concentrations showed no similar increase in IL10 levels, which remained comparable to control levels (Fig. 2b). Similarly, while cells treated with tuftsin and NCM resulted in a reduction in TNFα, the opposite was observed in groups treated with EG00229, which showed a slight increase in TNFα levels over control (Fig. 2a). Moreover, when the overall shift to an anti-inflammatory state in microglial cells was assessed, noted by the ratio of M2 to M1 gene expression, the EG00229 treatment resulted in reversion of these cells to a state similar to untreated controls (Fig. 2c). Thus, these experiments indicate that EG00229 can effectively prevent tuftsin's actions on microglial cells by blocking the M2 shift.

Figure 2.

The tuftsin-mediated M2 shift in microglia is disrupted by EG00229. Quantitative RT-PCR was performed to analyze changes in gene expression of the M1 marker TNFα (a) and the M2 marker IL10 (b). Primary microglia were treated 100 μg/mL tuftsin or 100 μg/mL tuftsin and NCM. Some groups were additionally treated with increasing concentrations of EG00229 as shown. (c) The ratio of the fold change in IL10 (M2) to the fold change in TNFα (M1). n = 3, *p < 0.05.

Blockade of TβR1 prevents the tuftsin-induced anti-inflammatory shift in microglia

Nrp1 employs different coreceptors which signal following ligand binding (Prud'homme and Glinka 2012). We investigated which one of these coreceptors is involved in mediating tuftsin signaling. A likely candidate is TβR1, as its classic ligand TGFβ has been extensively associated with anti-inflammatory effects. Nrp1 can bind and activate the latent form of TGFβ, which is associated with immunosuppressive regulatory T cell function (Karpanen et al. 2006; Wei et al. 2007). It is also essential in the development of alternatively activated M2 microglia (Zhou et al. 2012).

To test if TβR1 is the coreceptor involved in tuftsin signaling, we used an inhibitor capable of blocking the kinase activity of TβR1 at 10 μM as previously described (Shiou et al. 2006). For comparison, we also used an inhibitor of c-Met kinase activity at 5 nM, in line with prior studies (Zou et al. 2012), which is an alternative coreceptor that Nrp1 could signal through (Prud'homme and Glinka 2012). Similar to the experiments in Fig. 2, microglia were treated for 10 h with combinations of tuftsin and NCM, in the presence or absence of c-Met inhibitor or TβR1 inhibitor (Shiou et al. 2006; Zou et al. 2012). After harvesting RNA, qPCR was performed to quantify the expression of M1 and M2 markers. The ratio of M2/M1 in c-Met inhibitor-treated samples was comparable to controls, with a threefold decrease in TNFα and threefold increase in IL10 in tuftsin and NCM-treated samples. However, in cells treated with TβR1 inhibitor a slight increase in TNFα levels and no change in IL10 levels were observed, as was the case for EG00229-treated microglia (Fig. 3a and b). Furthermore, while there was a significant anti-inflammatory switch in both control and c-Met inhibitor-treated cells when exposed to NCM and tuftsin, this was abolished in TβR1-treated samples (Fig. 3c). Taken together, these data indicate that tuftsin signals through the TGFβ signaling pathway via TβR1.

Figure 3.

Blockade of TGFβ receptor 1 (TβR1) prevents the tuftsin-mediated M2 shift in microglia. Quantitative RT-PCR was performed to analyze changes in gene expression of the M1 marker TNFα (a) and the M2 marker IL10 (b). Primary microglia were treated 100 μg/mL tuftsin or 100 μg/mL tuftsin and NCM. Some groups were additionally treated with 5 nM c-Met inhibitor or 10 μM TβR1 inhibitor. (c) The ratio of the fold change in IL10 (M2) to the fold change in TNFα (M1). n = 3–5, *p < 0.05; **p < 0.01; ns, not significant.

Tuftsin signals through the canonical TGFβ signaling pathway

We extended these observations by further investigating the downstream targets of TβR1. TGFβ is characterized by signaling through two separate pathways, the canonical and non-canonical pathways, which have disparate effects. The canonical signaling pathway is associated with Smad2/3 phosphorylation, which then forms a complex with Smad4 that translocates to the nucleus, binds DNA, and regulates transcription. This series of events has been linked with inhibition of immune responses and cell proliferation (Rahimi and Leof 2007; Prud'homme and Glinka 2012). On the other hand, non-canonical TGFβ signaling occurs through various molecules such as phosphatidylinositide 3-kinase, Akt, or Erk (Zhang 2009). These two pathways have been shown to have complex cross talk, for the canonical and non-canonical pathways can antagonize each other. For example, hyperactivation of Akt can reduce canonical signaling, particularly through inhibition of Smad3 (Zhang 2009; Tian et al. 2011).

To investigate whether tuftsin signals through the canonical or non-canonical TGFβ pathway, we performed immunoblots to examine Smad3 and Akt phosphorylation levels. Microglial cells treated with combinations of NCM and tuftsin for 10 hours in the presence or absence of TβR1 inhibitor or c-Met inhibitor were re-stimulated at hour 9 to allow for phosphorylation changes to be more readily observable. Control samples showed a significant increase in Smad3 phosphorylation in response to tuftsin and NCM, indicating that tuftsin signals through the canonical pathway. This increase was also observed in samples treated with c-Met inhibitor, but was abolished in those treated with TβR1 inhibitor (Fig. 4a and b). Conversely, microglia treated with NCM and tuftsin revealed significantly decreased levels of Akt phosphorylation, indicating a decrease in non-canonical TGFβ signaling similar to that in samples treated with c-Met inhibitor. However, when the activity of TβR1 was blocked, a decrease in Akt phosphorylation was not observed in the microglial samples (Fig. 4c and d). Samples treated with 30 μM EG00229 behaved similar to those exposed to TβR1 inhibitor, for both p-Smad3 and p-Akt (data not shown). These experiments suggest that tuftsin signals via the canonical TGFβ signaling pathway.

Figure 4.

Tuftsin signals through the canonical transforming growth factor beta (TGFβ) signaling pathway, which is prevented by TGFβ receptor 1 (TβR1) inhibition. Western blots were performed on primary microglia extracts that were exposed to NCM, 100 μg/mL tuftsin, 5 nM c-Met inhibitor or 10 μM TβR1 inhibitor for 10 h. 1 h prior to harvesting, cells were restimulated with their designated treatment to better observe phosphorylation changes. Blots were probed for Smad3 and phospho-Smad3 (a), and fold change relative to NCM treatment alone was quantified in (b), and adjusted to reflect changes in total Smad3 levels. Blots were probed for Akt and phospho-Akt in (c), and fold change relative to NCM alone is quantified in (c), and adjusted to reflect changes in total Akt levels. n = 3, *p < 0.05; **p < 0.01; ns, not significant.

Tuftsin promotes TGFβ release from cells, which is prevented by TβR1 blockade

To further confirm polarization to an M2 subset as a result of tuftsin treatment, as well as the effect of c-Met, TβR1, and EG00229 inhibitors, we performed an ELISA assay for TGFβ. We show that tuftsin and NCM promote a significant release of TGFβ from microglial cells, similar to those treated with 5 nM of c-Met inhibitor. However, this does not occur when cells are exposed to TβR1 inhibitor, or EG00229 (Fig. 5a). To observe the release of reactive nitrogen species from cells as a means of determining a shift away from the M1 phenotype in cells, we quantified nitrite release by DAN reagent. However, there was no significant change between all treatments (Fig. 5b). Thus, blockade of tuftsin binding to Nrp1 by EG00229 or disruption of TβR1 signaling prevents the tuftsin-mediated M2 shift in microglia.

Figure 5.

Blockade of TGFβ receptor 1 (TβR1) prevents tuftsin-mediated transforming growth factor beta (TGFβ) up-regulation. An ELISA assay probing for TGFβ levels (a) or a nitrite assay using DAN reagent (b) was performed on media isolated from primary microglial cells treated 100 μg/mL tuftsin or 100 μg/mL tuftsin and NCM for 10 h. Some groups were additionally treated with 5 nM c-Met inhibitor, 10 μM TβR1 inhibitor, or 30 μM EG00229, also for 10 h. n = 3, **p < 0.01; ***p < 0.001, ns, not significant.


In this study, we provide several pieces of evidence that support the hypothesis that tuftsin signals exclusively through its receptor Nrp1 via the canonical TGFβ signaling pathway. Our previous work showed that tuftsin was capable of promoting an anti-inflammatory, M2 shift in microglia (Wu et al. 2012). In this report, we show that blockade of tuftsin binding to Nrp1 prevents this shift, and that inhibition of TβR1 similarly blocks microglial polarization to an anti-inflammatory phenotype. As the canonical TGFβ pathway is associated with immunosuppression (Rahimi and Leof 2007), our current findings support our previously published work in the experimental allergic encephalomyelitis model of multiple sclerosis where tuftsin enhanced an M2 shift and promoted the expression of immunosuppressive Tregs (Wu et al. 2012).

It is notable that neither tuftsin treatment nor NCM treatment alone promotes an M2 shift in microglial cells (Wu et al. 2012). This is consistent with the idea that macrophages and microglia are often activated in a ‘two-hit’ process. This model describes that a weaker, non-specific initial signal produces a priming response in target cells, which, when faced with a second activation signal, are rapidly activated and produce a robust reaction (Hains et al. 2010; O'Leary et al. 2011). In our system, NCM would function as the initial signal, and tuftsin would then produce a strong anti-inflammatory response in ‘primed’ microglial cells.

As EG00229 was originally generated as an inhibitor of VEGF binding to Nrp1 (Jarvis et al. 2010), it is possible that tuftsin is in fact signaling through the VEGF pathway. There are data both supporting and refuting this theory. It has been shown that treatment of cells with TGFβ can promote the induction of VEGF (Shao et al. 2009), and that VEGF treatment promotes TGFβ production in macrophages (Luo et al. 2012). However, VEGF binds to both Nrp1 and its coreceptor VEGFR concomitantly, which aids in the recruitment of VEGFR for signaling (Wiesmann et al. 1997). This would not occur for tuftsin, as it is comprised of only four amino acids and binds deep within the VEGF pocket on the Nrp1 molecule (Vander Kooi et al. 2007), preventing its interaction with any other receptors. The fact that treatment of microglia with EG00229 resulted in small increase in TNFα over control cells may possibly be because of the fact that EG00229 does not only block tuftsin binding but also prevents VEGF binding as well (Jarvis et al. 2010), so it can result in different behavior in microglia rather than just loss of tuftsin interaction alone.

Another immune cell type that readily expresses Nrp1 are regulatory T cells, Tregs (Weiss et al. 2012). Tregs cells play an essential role in suppressing ongoing inflammatory immune responses (Kohm et al. 2002; McGeachy et al. 2005). These cells differentiate in response to, as well as bind and activate, TGFβ (Lohr et al. 2006; Glinka and Prud'homme 2008). It would thus be pertinent to investigate whether tuftsin also signals through the canonical TGFβ signaling pathway in these cells as well.

As tuftsin has been used extensively over its 40-year history, the lack of knowledge of its mechanism of action has been a notable void in the information regarding this useful molecule. Our work is novel as although it has been shown that tuftsin binds to Nrp1 (von Wronski et al. 2006), Nrp1 utilizes a wide variety of coreceptors that function in disparate pathways for signaling (Glinka and Prud'homme 2008). Downstream signaling as a result of tuftsin binding to cells is an area that has not been investigated prior to our study. Here, we present data that identify the canonical TGFβ signaling pathway as the means by which tuftsin exerts its anti-inflammatory effects.

Conflict of interest

None of the co-authors have a conflict of interest to declare.


We thank members of the Tsirka-lab for helpful discussions and suggestions, Francesca Gist Nakagawa for discussions and reagents about the TGFβ signaling pathway. This work was partially funded through a 3MT-IGERT fellowship to JCN and PP1815 Pilot NMSS grant and NIH R01NS42168 to SET.