RESEARCH PAPER
A novel bioactive peptide: assessing its activity over murine neural stem cells and its potential for neural tissue engineering Q1
Andrea Caprini1,2, Diego Silva1,2, Ivan Zanoni2, Carla Cunha1,2,4, Carolina Volonte`2,5, Angelo Vescovi1,2,3 and Fabrizio Gelain1,3, 1
Center for Nanomedicine and Tissue Engineering, A.O. Ospedale Niguarda Ca’ Granda, Milan 20162, Italy Biotechnology and Biosciences Department, University of Milano-Bicocca, Milan 20126, Italy 3 IRCCS Casa Sollievo della Sofferenza Opera di San Pio da Pietrelcina, San Giovanni Rotondo 71013, Italy 2
The design of biomimetic scaffolds suitable for cell-based therapies is a fundamental step for the regeneration of the damaged nervous system; indeed growing interest is focusing on the discovery of peptide sequences to modulate the fate of transplanted cells and, in particular, the differentiation outcome of multipotent neural stem cells. By applying the Phage Display technique to murine neural stem cells we isolated a peptide, KLPGWSG, present in proteins involved in both stem cell maintenance and differentiation. We show that KLPGWSG binds molecules expressed on the cell surface of murine adult neural stem cells, thus may potentially be involved in stem cell fate determination. Indeed we demonstrated that this peptide in solution enhances per se cell differentiation toward the neuronal phenotype. Hence, we synthesized two LDLK-12-based self-assembling peptides functionalized with KLPGWSG peptide (KLP and Ac-KLP) and characterized them via atomic force microscopy, rheometry and circular dichroism, obtaining nanostructured hydrogels supporting murine neural stem cells differentiation in vitro. Interestingly, we demonstrated that, when scaffold stiffness is comparable to that of the brain in vivo, the Ac-KLP SAP-based scaffold enhances the neuronal differentiation of neural stem cells. These evidences place both KLPGWSG and the functionalized self-assembling peptide Ac-KLP as promising candidates for, respectively, biomimetic studies and stem cell therapies for nervous regeneration.
Introduction Regenerative medicine is a scientific discipline aimed at developing new strategies for the repair or replacement of damaged tissues and organs. The structural and functional complexities characterizing each organ unequivocally lead to the need of multidisciplinary approaches for successful regenerative therapies. Tissue
Corresponding author: Gelain, F. (
[email protected]), (
[email protected]) 4 Current address: INEB-Instituto de Engenharia Biome´dica, Rua do Campo Alegre, 823, 4150-180 Porto, Portugal. 5 Current address: Department of Physics, University of Milano,Via Celoria 16, 20133 Milan, Italy.
engineering responds to that need bringing together biology, chemistry, physics and material science principles [1]. One of the major aims of tissue engineering is the creation of materials suitable for cell-based regenerative medical strategies. To achieve this ambitious goal, a set of key physico-chemical and biological peculiarities must be considered in the material design. In particular, biomaterials should be feasibly synthesized to high purities, pathogen-free and ‘prone’ to functionalizations with bioactive molecules. At the same time they should support the growth, survival and/or differentiation of specific cell types both in vitro and in vivo. Moreover, they should mimic the nanostructured architecture of native extracellular matrices, thus allowing both
www.elsevier.com/locate/nbt 1 Please cite this article in press as: Caprini, A. et al., A novel bioactive peptide: assessing its activity over murine neural stem cells and its potential for neural tissue engineering, New Biotechnol. (2013), http://dx.doi.org/10.1016/j.nbt.2013.03.005
1871-6784/$ - see front matter ß 2013 Published by Elsevier B.V. http://dx.doi.org/10.1016/j.nbt.2013.03.005
Research Paper
New Biotechnology Volume 00, Number 00 April 2013
NBT 596 1–11 RESEARCH PAPER
Research Paper
cell remodeling and a proper diffusion of biomolecules [1]. Because the visco-elastic properties of materials are known to modulate cell differentiation [2], they must be tunable according to the viscoelastic properties of the tissue to be regenerated. Finally, materials should be biodegradable and should not elicit significant adverse immune responses. Among several possible candidates, self-assembling peptides (SAPs) display a satisfactory balance between the mentioned peculiarities [3]. Indeed they have been demonstrated to support the survival, growth and differentiation of many cell types, especially neural cells, both in vitro and in vivo [1], thus making SAPs particularly appealing for strategies aimed at neural tissues regeneration. In this optic, in the recent years we demonstrated that SAPs support the growth and differentiation of adult neural stem cells (NSC) [4–6]. NSC are undifferentiated multipotent stem cells found in the subventricular and subgranular zones (SVZ and SGZ , respectively) of the mammalian adult brain [7]; identifiable in vivo by the expression of specific proteins, such as Nestin, GFAP [8] and Sox2 [9], NSC represent a good cell type for neuroregenerative purposes not only because of their ability to differentiate into neurons, astrocytes and oligodendrocytes in vitro [10] but also because they do not show tumorigenic potentials upon transplantation in different animal models of neurodegenerative diseases [11]. Moreover, the neural differentiation of NSC can be modulated by specific peptide sequences present in extracellular matrices, in soluble molecules and/or in membrane-bound cell molecules [12]. One of the most appealing techniques to identify short peptides able to bind proteins expressed in living cells is the Phage Display [13,14]; indeed a Phage Display-based approach has been successfully applied to many primary cell types and cell lines including primary motor neurons [15], neural progenitors [16], adult stem cells [17] and cancer cell lines [18,19]. In this work we exploited a multidisciplinary approach to create SAP-based scaffolds suitable for the stem cell-based regeneration of the damaged nervous system. By targeting undifferentiated murine NSC via Phage Display technique we identified a short peptide (KLPGWSG) present in proteins involved in the modulation of stem cells proliferation and differentiation. This peptide, tagged with a FITC fluorochrome, was shown to bind NSC and, when added to NSC differentiation medium or linked to a SAP scaffold, it significantly shifted their differentiation toward the neuronal phenotype. Indeed, self-assembling hydrogels functionalized with the mentioned sequence displayed a promising pro-neuronal differentiative potential suitable for future NSCbased therapies in vivo. These data show the potential of multidisciplinary studies for the development of novel regenerative medical approaches.
Results To develop nanostructured scaffolds suitable for NSC-based regenerative purposes we applied a multidisciplinary approach. In particular, molecular and cell biology have been applied to identify peptides and SAPs having biological effects on NSC and to evaluate the magnitude of these effects: similarly chemistry and physics allowed us to, respectively, synthesize ad hoc nanostructured scaffolds supporting NSC differentiation and to characterize their visco-elastic properties. 2
New Biotechnology Volume 00, Number 00 April 2013
TABLE 1
Peptide sequences of 20 randomly picked phage plaques after four rounds of biopanning Membrane-bound phage clones
Internalized phage clones
Sequence
Frequency
%
Sequence
Frequency
%
KLPGWSG
9
45
KLPGWSG
0
0
HAIYPRH
1
5
HAIYPRH
4
20
GETRAPL
2
10
GETRAPL
2
10
ALTPWAF
2
10
ALTPWAF
0
0
GKPMPPM
2
10
GKPMPPM
0
0
SILPYPY
0
0
SILPYPY
2
10
EQRGAPR
1
5
EQRGAPR
0
0
YSIPKSS
1
5
YSIPKSS
0
0
TDSLRLL
1
5
TDSLRLL
0
0
SMYGSYN
0
0
SMYGSYN
1
5
KLPISSK
0
0
KLPISSK
1
5
KDPGWSG
0
0
KDPGWSG
1
5
STASYTR
0
0
STASYTR
1
5
IQSPHFF
0
0
IQSPHFF
1
5
YLTMTPT
0
0
YLTMTPT
1
5
GPIIPRN
0
0
GPIIPRN
1
5
Phage Display: identification of peptide sequences binding to NSC We applied the Phage Display technique to undifferentiated NSC to find peptide sequences interacting with them. After a negative selection against the plastic cell culture dishes and four rounds of panning we found consensus for a panel of membrane-bound and internalized peptides (Table 1). Among them, the 87.5% resulted exclusively detected in one of the two conditions, with around 43% of this total represented by membrane-bound and 57% by internalized phage. KLPGWSG and HAIYPRH heptapeptides showed the highest consensus among, respectively, the exclusively membrane-bound (45%) and internalized (20%) sequences. The highest recurrence of KLPGWSG, its selective presence in membrane-bound phage and the crucial role played by cell membrane molecules as direct interfaces between biomaterials and cells prompted us to focus our attention on the KLPGWSG peptide.
KLPGWSG is found in proteins involved in stem cells proliferation and differentiation BLAST analysis of KLPGWSG was performed on murine proteins. As expected by the limited number of amino acids composing the identified peptide, a large amount of proteins were recovered from the analysis. The nature of the cells subjected to the Phage Display and the exclusive localization of the KLPGWSG sequence as membrane-bound prompted us to focus on membrane proteins involved in stem cells proliferation and differentiation. Between them, major consensuses for KLPGWSG were found on three proteins known to affect stem cells behaviors: Notch1 [20,21], Dll4 [22,23] and MEGF10 (Table 2) [24–30]. Interestingly, the consensus sequences fall mainly inside protein modules playing known roles in cellular events such as adhesion and receptor– ligand interactions called Epidermal Growth Factor (EGF)-like
www.elsevier.com/locate/nbt Please cite this article in press as: Caprini, A. et al., A novel bioactive peptide: assessing its activity over murine neural stem cells and its potential for neural tissue engineering, New Biotechnol. (2013), http://dx.doi.org/10.1016/j.nbt.2013.03.005
NBT 596 1–11 RESEARCH PAPER
TABLE 2
BLAST analysis of KLPGWSG: potentially relevant sequence homologies in mouse proteins Proteins
BLAST accession number
Alignments
MEGF10
EDL09867
128 303 474 673
Notch1
AAM28905.1
131 – PGWSG 135 743 – PGWSG 747 1644 – LPG-SG 1649
Dll4
AAF76428.1
210 – LPGW-G 215
– – – –
PGW-G 132 PGWSG 307 LPGWSG 479 PGW-G 677
domains (data not shown) [31]. Despite the intrinsic limitations of this BLAST analysis and the consequently logic need for more direct evidences concerning the nature of the KLPGWSG partners, these data suggest that a possible effect of the mentioned peptide on NSC behavior can be considered realistic.
KLPGWSG affinity for NSC To understand if the KLPGWSG peptide effectively binds to NSC we performed FACS analysis using antibodies recognizing NSC and the peptide functionalized with the fluorochrome FITC. Inside the bulk population obtained by dissociating neurospheres we first detected 100% of cells expressing Nestin and 98.7% co-expressing also GFAP (Fig. 1a). The 98.6% of GFAP expressing cells coexpressed Sox2 as well (Fig. 1b), indicating that the 96.4% of the cell bulk were positive for the three markers. Among them, 95% of cells were also recognized by the KLPGWSG-FITC peptide (Fig. 1c), while FITC alone does not specifically recognize any cell (data not shown), thus finally indicating that the 91.6% of the double positive population bind the KLPGWSG peptide. We confirmed these data in three independent experiments. These results strongly indicate a binding of our peptide to NSC.
The biological effect of KLPGWSG on NSC proliferation Taken together, the BLAST findings and the general understanding that short EGF-based synthetic peptides can mimic some of the EGF functions [32] suggested us a possible effect of the KLPGWSG
peptide on NSC stemness maintenance and/or neural differentiation. To understand if the KLPGWSG peptide can mimic the proliferative effect of EGF, NSC were exposed for seven days to different peptide concentrations (ranging from picograms to milligrams) in culture media without EGF, bFGF or both (Fig. 2). At Day 7 the cell viability test MTS was performed. We also quantified NSC proliferation without the KLPGWSG peptide in the bFGF or EGFdepleted culture media immediately after plating (columns called Untreated Day 0 in (a) and (b)), after seven days of culture (columns called Untreated Day 7 in (a) and (b)) and again after seven days but under normal culture conditions (columns called EGF + bFGF Day 7 in both (a) and (b)). As expected we found almost no cell proliferation in the presence of bFGF alone, a slight proliferation rate in cells cultured with EGF alone (column Untreated Day 0 versus column Untreated Day 7 in (a) and (b), respectively) and almost no living cells without any growth factor (not shown). As expected, in all cases cell proliferation resulted significantly lower than the one found in cells cultured with both bFGF and EGF (columns EGF + bFGF Day 7 versus columns Untreated Day 7) [33]. Cells treated with the KLPGWSG peptide in the absence of bFGF or EGF displayed a proliferation rate comparable to that found in the untreated counterparts (columns underlined with the black line called + KLPGWSG Day 7 versus columns Untreated Day 7). In the absence of both the growth factors we found that the KLPGWSG peptide is not sufficient per se to sustain cell survival and proliferation (data not shown). These data demonstrate that the KLPGWSG peptide does not replace the biological effect of EGF on NSC.
KLPGWSG enhances NSC neuronal differentiation We then tested if the KLPGWSG peptide affects per se NSC neural differentiation. We differentiated NSC for seven days on Cultrex or Laminin-coated dishes treated with the KLPGWSG peptide (Fig. 3). In negative controls KLPGWSG was not administered. On both substrates we found a significant increase in the number of Beta III Tubulin+ neurons with a concomitant decrease in the number of GFAP+ astrocytes: a more pronounced enhancement of neuronal differentiation was detected on Laminin. No significant differences were found in both substrates in the number of GalC/O4+
FIGURE 1
FACS analysis of NSC incubated with FITC-tagged KLPGWSG peptide. 98.7% of cells co-expressed Nestin and GFAP (a), with 100% of cells expressing Nestin; inside a 98.6% of Sox2/GFAP co-expressing cells (b) a 95% of cells is recognized by the KLPGWSG-FITC peptide (c). www.elsevier.com/locate/nbt 3 Please cite this article in press as: Caprini, A. et al., A novel bioactive peptide: assessing its activity over murine neural stem cells and its potential for neural tissue engineering, New Biotechnol. (2013), http://dx.doi.org/10.1016/j.nbt.2013.03.005
Research Paper
New Biotechnology Volume 00, Number 00 April 2013
NBT 596 1–11 RESEARCH PAPER
New Biotechnology Volume 00, Number 00 April 2013
Research Paper FIGURE 2
Proliferation of NSC versus presence of soluble KLPGWSG peptide. NSC were cultured in medium containing bFGF or EGF alone (columns in a and b underlined with the black line called bFGF alone or EGF alone, respectively) for one hour (columns called Untreated Day 0) or seven days (from columns called Untreated Day 7 to columns called 1 mg/ml) in the presence (columns underlined with the black line called + KLPGWSG Day 7) or not of the KLPGWSG peptide in the medium. As reference, the proliferation rate of NSC cultured under normal conditions is shown (columns called EGF + bFGF Day 7). Proliferation rate is expressed as Optical Density minus blank (medium without cells). All the columns in each graph display a statistical significance with the respective EGF + bFGF Day 7 column (not shown; P < 0.05).
and Nestin+ cells. These data indicate that the KLPGWSG peptide positively modulates the neuronal differentiation of NSC, thus giving us also the emphasis for the creation of functionalized SAPbased scaffolds potentially enhancing neuronal differentiation and survival.
SAPs characterization We synthesized and characterized two SAPs sharing the same LDLK-12 self-assembling core sequence: KLP (derived from the N-terminal functionalization of the LDLK-12 core with the KLPGWSG peptide) and Ac-KLP (similar to KLP but acetylated at the N-terminus) SAPs. We choose the LDLK-12 motif as selfassembling core sequence because of its spontaneous folding into b-sheets upon exposure to neutral pH [34] allowing the formation of hydrogels suitable for cell cultures and in vivo injection [35]. We generated the Ac-KLP because acetylation is known to improve nanofibers stability [36]. Moreover, to improve the exposure of the functional motifs, a spacer of three-glycines was interposed between the functional motifs and the LDLK-12 core [6]. Finally, N-terminal functionalization, instead of C-terminal functionalization of LDLK-12, was preferred to ‘mimic’ the exposure of KLPGWSG in Phage Display experiments [16]. Acetylated 4
LDLK-12 (LDLK-12) was synthesized as well and adopted as a control SAP without functionalization. The sequence of the three SAPs is shown in Table 3. All three SAPs formed linear nanofibers (Fig. 4). From morphometrical analyses we evinced that nanofibers of Ac-KLP and KLP displayed statistically significantly larger fibers compared to those of LDLK-12 (14.66 0.55 nm and 14.78 0.67 nm versus 11.22 0.79 nm, respectively) whereas nanofiber height was similar (1.76 0.07 nm and 1.7 0.07 nm versus 1.58 0.09 nm, respectively) (Fig. 4a). The increased width of functionalized peptides with respect to LDLK-12 is compatible with the addition of the flanking functional motifs at N-termini of the self-assembling core sequences [37]. Circular dichroism shows the presence of b-sheet structures in the nanofibers of the three SAPs (Fig. 4b), with higher content for Ac-KLP with respect to KLP. We next analyzed their propensity to form hydrogel through rheometry. The synthesized peptides were dissolved at different concentrations and self-assembled by adding PBS, then their storage and loss moduli (G0 and G00 , respectively) were determined. For the sake of simplicity, the G0 and G00 profiles of the Ac-KLP only is shown in Fig. 4c. For all the SAPs we found both an increase of the G0 and G00 moduli directly proportional to the SAPs concentration and a G0 value always higher than G00 : the latter feature
www.elsevier.com/locate/nbt Please cite this article in press as: Caprini, A. et al., A novel bioactive peptide: assessing its activity over murine neural stem cells and its potential for neural tissue engineering, New Biotechnol. (2013), http://dx.doi.org/10.1016/j.nbt.2013.03.005
NBT 596 1–11 RESEARCH PAPER
Research Paper
New Biotechnology Volume 00, Number 00 April 2013
FIGURE 3
Neural differentiation of NSC on Cultrex and Laminin in the presence of soluble KLPGWSG peptide. NSC were differentiated seven days on Cultrex- or Laminincoated dishes (left and right panel, respectively) in the presence (columns KLPGWSG 100 mg/ml in the graphs and left group of images in the two panels) or not (columns Untreated in the graphs and right group of images in the two panels) of the KLPGWSG peptide and then subjected to immunofluorescence to detect neurons (Beta III Tubulin staining), astrocytes (GFAP staining), oligodendrocytes (GalC and O4 staining) and immature neural precursor cells (Nestin staining). Nuclei were visualized with DAPI (in blue). The quantification of the mentioned phenotypes and the corresponding statistical significances are shown in the graphs (*P < 0.05, **P < 0.01). Scale bars: 100 mm.
indicates that in the tested SAPs the elastic component prevails to the viscous one, a classical peculiarity of solid hydrogels [38]. These analyses demonstrate not only that the addition of the chosen functional motifs at the N-terminus did not impair the
capability of the LDLK-12 core to form nanostructured hydrogels but also that the LDLK-12, the KLP and the Ac-KLP are able to form scaffolds with tunable visco-elastic characteristics. Then we were interested in understanding if these physico-chemical features
www.elsevier.com/locate/nbt 5 Please cite this article in press as: Caprini, A. et al., A novel bioactive peptide: assessing its activity over murine neural stem cells and its potential for neural tissue engineering, New Biotechnol. (2013), http://dx.doi.org/10.1016/j.nbt.2013.03.005
NBT 596 1–11 RESEARCH PAPER
New Biotechnology Volume 00, Number 00 April 2013
TABLE 3
Sequences of the synthesized SAPs and their respective acronyms Sequence
Abbreviation
Ac-LDLKLDLKLDLK-CONH2
LDLK-12
Ac-KLPGWSGGGGLDLKLDLKLDLK-CONH2
Ac-KLP
NH2-KLPGWSGGGGLDLKLDLKLDLK-CONH2
KLP
cooperate in affecting NSC differentiation; therefore we next extrapolated by rheology tests the SAP concentrations necessary for having G0 values around and exceeding 500 Pa (Fig. 4d), a value matching the in vivo brain stiffness [39]. Data are summarized in Table 4. Despite these data indicating differences in the fiber densities for the three materials at these stiffnesses (two parameters hardly manageable independently), we used these SAPs concentrations to determine if, and to what extent, stiffness and functionalization may impact neural differentiation of NSC.
Research Paper FIGURE 4
Structural and mechanical characterization of the LDLK-12, Ac-KLP and KLP SAPs. Nanofibers of the three SAPs were visualized by atomic force microscopy (panel a); their width and height were indicated below each figure. Circular dichroism (panel b) shows the presence of b-sheet structures for all the three SAPs, while rheology (panel c) indicates that their elastic component prevails to the viscous one. Rheology was also used to determine for each SAP the G0 values corresponding to different concentrations (panel d), thus allowing the extrapolations of the concentrations needed to reach the G0 values of 100, 500 and 1000 Pa. 6
www.elsevier.com/locate/nbt Please cite this article in press as: Caprini, A. et al., A novel bioactive peptide: assessing its activity over murine neural stem cells and its potential for neural tissue engineering, New Biotechnol. (2013), http://dx.doi.org/10.1016/j.nbt.2013.03.005
NBT 596 1–11 New Biotechnology Volume 00, Number 00 April 2013
RESEARCH PAPER
TABLE 4
SAPs concentrations (w/v) corresponding to G0 values around 100, 500 and 1000 Pa. Values were extrapolated from the graph shown in Fig. 4 panel d SAP
100 Pa
500 Pa
1000 Pa
LDLK-12
0.3%
1.2%
2.9%
Ac-KLP
1.1%
3.4%
4.8%
KLP
1.3%
3.8%
5.5%
To assess the effect of the three SAP-based scaffolds on NSC differentiation we differentiated neural stem cells on LDLK-12, Ac-KLP or KLP, assembled at the concentrations indicated in Table 4 for seven days. From the daily observation of seeded cells during differentiation we found that all the scaffolds support a rapid cell attachment (data not shown), although we did not detect evident morphological differences among the materials and at different hydrogel stiffnesses (Supplementary Fig. S1). Additionally, major morphological signs of cell death or suffering were not visible, thus suggesting that they support also cell survival during differentiation. However, important differences were detected by analyzing the neural phenotypes evinced by immunostainings (Fig. 5). Indeed we found an inverse relationship between the SAPs stiffness and the percentage of neurons differentiated (Fig. 5a): differences were significant between the same SAPs at hydrogels stiffnesses of 100 Pa and 1000 Pa, and for LDLK-12 and Ac-KLP at 100–500 Pa and at 500–1000 Pa, respectively. An inverse trend was found for astrocytes, with a direct relationship between stiffness and astrocytes number (Fig. 5b). No differences were visible in the number of oligodendrocytes (Fig. 5c). Notably, if compared to LDLK-12, a clearly increased percentage of neurons was detected in both functionalized SAPs at 100 Pa and 500 Pa; however, a statistical significant difference was reached at 500 Pa only between Ac-KLP and the LDLK-12 counterpart. Immunofluorescence detection of the three neural phenotypes (Fig. 6) showed a very highly organized network of neurons on the Ac-KLP scaffold at both 100 Pa
Research Paper
Ac-KLP enhances NSC neuronal differentiation for stiffness comparable to the brain
FIGURE 5
NSC differentiated progeny on LDLK-12, Ac-KLP or KLP scaffolds (7 DIV). Neurons (a), astrocytes (b) and oligodendrocytes (c) were quantified at both 100, 500 and 1000 Pa. *P < 0.05, **P < 0.01.
and 500 Pa if compared to the other two SAPs at similar stiffness (green, Fig. 6a). However, no evident differences were seen among neuronal networks obtained at 1000 Pa. Morphology and organization of astrocytes (red, Fig. 6a) and oligodendrocytes (red, Fig. 6b) did not differ in all the experimental conditions. These data
FIGURE 6
Immunofluorescence detection of the neurons (panel a, Beta III Tubulin staining, green cells), astrocytes (panel a, GFAP staining, red cells) and oligodendrocytes (panel b, GalC and O4 staining, red cells) obtained differentiating NSC for seven days on LDLK-12, Ac-KLP or KLP scaffolds at stiffness values around 100, 500 and 1000 Pa. Nuclei were visualized with DAPI (in blue). Scale bar: 100 mM. www.elsevier.com/locate/nbt 7 Please cite this article in press as: Caprini, A. et al., A novel bioactive peptide: assessing its activity over murine neural stem cells and its potential for neural tissue engineering, New Biotechnol. (2013), http://dx.doi.org/10.1016/j.nbt.2013.03.005
NBT 596 1–11 RESEARCH PAPER
suggested that the presence of a NSC-related functional motif, its acetylation and an appropriately tuned scaffold stiffness synergically enhance the neuronal differentiation of NSC in the Ac-KLP scaffolds. Supplementary material related to this article found, in the online version, at http://dx.doi.org/10.1016/j.nbt.2013.03.005.
Discussion
Research Paper
A fundamental step in the field of regenerative medicine is the creation of scaffolds with ad hoc physico-chemical characteristics for the tissue to be regenerated. In our work we described the multidisciplinary approach allowing us to create a nanostructured SAP-based scaffold potentially suitable for the cell-based regeneration of neural tissues. By applying the Phage Display technique to NSC we identified a short peptide (KLPGWSG) binding one or more molecules expressed on their cell surface. Despite having already exploited the power of this approach in identifying functional motifs suitable for neuroregenerative purposes [16], a more detailed analysis of the biological effect of the identified motifs on NSC behavior was missing. Here we showed that, when linked to the FITC fluorochrome, the KLPGWSG peptide may be used as a NSC marker that, together with other reporter systems (such as biotin or radioactive isotopes) may potentially be used on fixed tissues, ex vivo samples or even in vivo. Nevertheless, we identified three biological characteristics of the peptide per se on the NSC biology: it does not substitute or increase the proliferative effect of EGF, it does not prevent cell death in the absence of both EGF and bFGF and it enhances neuronal differentiation on differentiative substrates. However, the molecular mechanisms linking these phenomena to our sequence remain unclear. Based on the BLAST analysis, the recovery of the KLPGWSG motif in proteins involved in stem cells differentiation fits with our results, especially considering the Notch/Dll system; regardless of that, no significant molecular hypotheses can be considered in the absence of direct data showing what are the molecules bound by the peptide. Indeed many molecular pathways are frequently crosstalking in response to a given stimulation [40], thus making difficult the identification of the ‘triggering steps’ starting only from the final effect. Therefore we did not quantify the expression of specific downstream target genes, with the idea to perform this analysis after having some more clear ideas about the KLPGWSG partners; in this optic we are now trying to set up a protocol for isolating the molecules bound by our peptide. For the same reasons we did not analyze even the phenotypical nature of the differentiated neurons. SVZ-derived NSC are known to generate GABAergic interneurons to be mainly integrated in the neural circuits of the olfactory bulbs [41,42]; it will be interesting to determine if the KLPGWSG peptide is also able per se to shift their default neuronal phenotype toward others such as the glutammatergic. We showed also that, when covalently linked to the LDLK-12 self-assembling core, both the KLPGWSG peptide and its acetylated counterpart do not interfere with the core self-assembling propensity, thus allowing their use as functional motifs for the creation of SAP-based scaffolds. Indeed both the KLP and the Ac-KLP SAPs form nanostructured scaffolds with tunable visco-elastic properties as well as their nonfunctionalized counterpart, the LDLK-12. When used as matrices for NSC cultures we found that they support cell attachment and survival in the range between 100 and 1000 Pa. However 8
New Biotechnology Volume 00, Number 00 April 2013
the elastic properties of matrices are known to modulate stem cell differentiation toward many lineages including the neural lineages [2], with an enhancement of neuronal differentiation inversely proportional to the matrix stiffness [43]. Accordingly, we demonstrated the same tendency also for NSC when differentiated in a pure SAP-made milieu. We noticed also two very interesting phenomena: (1) a higher amount of differentiated neurons when the KLPGWSG sequence is not simply dissolved in the culture medium, but immobilized on a LDLK-12 backbone and (2) a significant enhancement of neuronal differentiation on Ac-KLP compared to LDLK-12 at 500 Pa, thus suggesting that this result derives from the simultaneous cooperation of an ad hoc chemical composition, of the presence of a NSC-related functional motif and of a well-tuned stiffness. The enhancement of neuronal differentiation when the KLPGWSG sequence is immobilized on a SAP backbone can be interpreted hypothesizing that the immobilization of the sequence allows the clustering of specific membrane molecules, a phenomenon known to be essential in enhancing signal transduction [44]. Another factor potentially involved in the cellular response to our matrices is the density of the ligands presented to the cells; ligand density is known to affect cell behavior and this effect can be directly attributable to the expression levels on the cell of the molecules that bind the ligands presented by the biomaterials [45]. Therefore, it is possible that also the fibers density affects the obtained results. However, from a technical point of view in our system it is unrealistic to affect ligand density without affecting also the biomaterial stiffness. It will be interesting for future works to analyze also this point by adopting a different system to modulate ligand density without critically affecting the biomaterial visco-elastic properties. In any case these results extend our previous analyses on the effect of functionalized SAPs on NSC differentiation [16] by including also the effect of the stiffness and place Ac-KLP as good candidate for future in vivo investigations as scaffold for NSC-based regenerative purposes. In this light, investigations for understanding if and how it affects the neuronal phenotype determination become very appealing. However, it is equally important to underline that data concerning the NSC biological response to the mentioned SAPs in terms of survival and differentiation were obtained culturing cells on the top of the scaffolds, therefore in a bidimensional (or partially tridimensional) environment. The physiological properties of organs are not only determined by the presence of both specific cell types and extracellular matrix components, but also by specific tissue architectures. Therefore, materials designed for cell-based regenerative approaches of tissues displaying a tridimensional cytoarchitecture, as in the case of the nervous system, must support cell growth, survival and differentiation in the three dimensions. SAPs have been demonstrated able not only to support the three-dimensional cell culture of specific cell types such as chondrocytes [46] and NSC [4] but also to induce the differential expression of specific genes according to the culture dimensionality [47]. Consequently, the dimensionality of our SAPs must be considered for a full understanding of their potentials as biomedical scaffolds. Taken together these results underline not only the simultaneous importance of both the SAPs chemical composition and visco-elastic properties in modulating NSC behavior, but also the need of a multidisciplinary approach to find the best fit between them.
www.elsevier.com/locate/nbt Please cite this article in press as: Caprini, A. et al., A novel bioactive peptide: assessing its activity over murine neural stem cells and its potential for neural tissue engineering, New Biotechnol. (2013), http://dx.doi.org/10.1016/j.nbt.2013.03.005
NBT 596 1–11 New Biotechnology Volume 00, Number 00 April 2013
We here proposed a multidisciplinary approach aimed at creating new SAPs for NSC-based regenerative purposes and we demonstrated its potential by producing a new SAP able to enhance NSC neuronal differentiation in vitro and potentially suitable for in vivo experimentations. These results emphasize the power of the collaboration between apparently distant science fields for the common goal of tissue regeneration.
Methods Murine neural stem cells culture and differentiation NSC were isolated from the SVZ of 8-week-old CD-1 albino mice and cultured until passage 10 as previously described [33]. Cell differentiation was done on cell culture 96 multiwells (Corning) uncoated (negative control), Cultrex or Laminin coated (positive control) or filled with SAP-based scaffolds. For coating, wells were incubated overnight at 378C with Cultrex-BME1 (R&D) or Laminin (Roche) diluted at 20 mg/ml in NS-A serum-free medium without growth factors. Scaffolds (30 ml for each well diluted at the desired concentration in water) were assembled by adding serum-free medium supplemented with bFGF (bFGF medium) two hours before cell deposition. In all cases 1.5 104 cells/cm2 were seeded and differentiated two days in bFGF medium, followed by five days in serumfree medium supplemented with Leukemia Inhibitory Factor (LIF, Chemicon) (20 ng/ml) and Brain Derived Neurotrophic Factor (BDNF, Peprotech) (20 ng/ml). This last medium was changed every two days. For experiments involving cell stimulation with the KLPGWSG peptide during differentiation, the peptide was dissolved at the desired concentration in bFGF medium.
Phage Display on adult neural stem cells Phage Display was done using the Phage Display 7 phage library from New England Biolabs with minor modifications of the supplied protocol. To avoid the selection of Phage Displaying peptides with affinity for the plastic cell culture dishes, 1012 colony forming units (cfu) of phage were placed on 10 cm Petri dishes (Falcon) for one hour at 378C in pre-warmed Incubation Medium (DMEM supplemented with 0.1% BSA). Unbound phages were recovered and amplified in Escherichia coli K12 ER2738. 1012 cfu of the negatively selected phage library were diluted in 10 ml of Incubation Medium and pipetted on 3 106 NSC seeded on Petri dishes (Falcon). The selection of membrane-bound phage was obtained by incubating cells with phage for one hour at 48C, and that of the internalized was obtained by incubating cells with phage one hour at 378C. After incubation, cells were washed five times with Washing Buffer (PBS supplemented with 1% BSA and 0.05% Tween-20) to remove unbound phage. Membrane-bound phage was recovered by stripping cells with 5 ml of glycine buffer (0.2 M glicine–HCl, pH 2.2, supplemented with 1 mg/ml BSA) for 5 min at room temperature. Internalized phages were recovered lysing cells with 9 ml of lysing buffer (0.1% TEA in PBS) for 5 min at RT. In both cases the solutions containing the recovered phages were neutralized with 1 ml of 1 M Tris–HCl buffer (pH 8.0) and the eluted phages were amplified and titered in E. coli K12 ER2738. Membrane-bound and internalized phages were amplified, positively selected and titered separately for four rounds of selection. After four rounds of biopanning, phage plaques were randomly picked out from the plates and sequenced by Macrogen Inc.
For homology identification, the obtained primary peptide sequences were compared to the NCBI database using the BLAST program.
Cell proliferation assay The KLPGWSG peptide was dissolved in medium without growth factors or supplemented with EGF or bFGF or both according to the experimental plan. 1.5 104 cells/cm2 were plated on uncoated 96 multiwells and exposed to the media mentioned above for one hour (Day 0) or seven days (Day 7). Cell proliferation was quantified using the CellTiter 961 Aqueous Proliferation Assay (Promega) following manufacturer’s instructions. Results (biological replicates n = 3) were obtained using a Vmax microplate reader (Molecular Devices) at 490 nm wavelength. Values were reported as means of the optical densities standard error of the mean.
NSC differentiation tests Phase contrast images of living cells were obtained after seven days of differentiation with Zeiss Axioplan 2 microscope. Immunofluorescence was done on NSC differentiated for seven days, fixed for 15 min at 48C with freshly prepared 4% paraformaldheyde solution. For Beta III Tubulin, GFAP and Nestin detection cells were permeabilized for 10 min at 48C with PBS/0.1% Triton X-100, blocked one hour with PBS/0.1% Triton X-100/10% Normal Goat Serum (NGS) and incubated overnight with the correspondent primary antibodies diluted in PBS/0.1% Triton X-100/1% NGS. After extensive washes with PBS/0.1% Triton X-100, cells were incubated one hour with the appropriate secondary antibodies diluted in PBS/0.1% Triton X-100/1% NGS. For GalC and O4 detection the same protocol was applied but omitting the permeabilization step and Triton X-100 addition to the solutions. The following antibodies were used: rabbit against GFAP (1:500, DakoCytomation), mouse against Beta III Tubulin (1:500, Covance), mouse against GalC (1:200, Chemicon) and O4 (1:200, Chemicon), Goat anti Mouse Alexa 488 (1:1000, Molecular Probes), Goat anti Mouse Cy3 (1:1000, Jackson Immunoresearch), Goat anti Rabbit Cy3 (1:1000, Jackson Immunoresearch). Nuclei were visualized with DAPI staining. Samples were mounted with FluorSave reagent (Calbiochem) and images were acquired with Zeiss ApoTome System microscope. The percentage of neural phenotypes was calculated dividing the number of antigens-expressing cells for the number of total cells. These numbers were obtained by counting around 300 cells for each of nine randomly chosen fields over a total of three independent experiments. Results were expressed as percentage means standard error of the mean.
Peptide synthesis and purification The SAPs were synthesized with a standard fluorenylmethyloxycarbonyl (Fmoc) solid phase technique using the automated synthesizer Liberty (CEM) as previously described [16]. More detailed information are reported in the Supplementary materials and methods section. The sequences of the synthesized SAPs are reported in Table 4. FITC labeling of KPLGWSG was performed as previously described [48]. Briefly, 1 equiv. of peptides dissolved in borate buffer pH 9.0 (Fluka) and 1.2 equiv. of FITC (Sigma-Aldrich) in DMSO (Sigma-Aldrich) were mixed and stirred for 16 hours at room temperature. The correct molecular weight of FITC-labeled
www.elsevier.com/locate/nbt 9 Please cite this article in press as: Caprini, A. et al., A novel bioactive peptide: assessing its activity over murine neural stem cells and its potential for neural tissue engineering, New Biotechnol. (2013), http://dx.doi.org/10.1016/j.nbt.2013.03.005
Research Paper
Conclusions
RESEARCH PAPER
NBT 596 1–11 RESEARCH PAPER
peptide was confirmed via mass spectrometry and then purified via HPLC. Purification was performed with semi preparative C18 column and with a 40 min gradient (5–90% acetonitrile in 0.1% TFA).
FACS analysis
Research Paper
All the steps were performed at 48C. Cell suspensions from dissociated neurospheres were fixed in freshly prepared 4% paraformaldheyde solution for 15 min, permeabilized with PBS/0.05% Triton-X 100 for 10 min, blocked with PBS/10% NGS for one hour and incubated two hours with primary antibodies and/or KLPGWSGFITC peptide diluted in PBS/10% NGS. After extensive washes with PBS, cells were incubated one hour with secondary antibodies (diluted in PBS/10% NGS) and then analyzed using FACS Calibur (Becton Dickinson). The following antibodies were used: rabbit against GFAP (1:500, DakoCytomation), mouse against Nestin (1:100, Chemicon), mouse against Sox2 (1:200, R&D), Goat anti Mouse APC (1:1000, Becton Dickinson), Goat anti Rabbit PE (1:1000, Jackson Immunoresearch). The KLPGWSG-FITC peptide was dissolved in PBS/50% DMSO and used at a final concentration of 8 mM. Negative controls were obtained by incubating cells only with secondary antibodies and/or with FITC alone.
Scaffold characterization The detailed description of both the rheological and atomic force microscopy analyses is reported in the Supplementary methods section. Rheological analyses of SAPs were carried out with a controlled stress rheometer (AR-2000ex, TA instruments) with an acrylic cone-plate geometry (diameter, 20 mm; truncation, 34 mm; angle, 1%). All tests were performed at a controlled temperature of 258C and in a solvent chamber. SAPs solutions were prepared dissolving each peptide in GIBCO water at a concentration of 1%, 2% and 3% (w/v); functionalized peptides were also tested at 6% (w/v). Atomic force microscopy (AFM; Veeco diMultiMode V) tests on SAPs solutions were prepared dissolving peptides in GIBCO water one day before the analysis. For all SAPs
New Biotechnology Volume 00, Number 00 April 2013
initial sample concentration was 1% (w/v) and immediately before imaging SAP solution were diluted to a final concentration of 0.001% (w/v) and deposited on freshly cleaved mica. Nanofiber width measures were deconvolved using the following formula: Dx = {2[h(2rt h)}1/2 (Dx is the width broadening effect, h is the measured nanofiber height and rt is the tip radius) [37]. CD spectra were collected in continuous scanning mode between 190 and 260 nm; the speed of the acquisition was 10 nm/min. Samples were measured at room temperature on an Aviv 62DS spectrometer. All spectra were averaged over three accumulations on a 1 mm quartz cuvette. The peptides were initially dissolved in distilled water at a concentration of 1% (w/v); after 20 min of sonication the solutions were diluted to a final concentration of 0.02% (w/v) and then analyzed. CD spectra of the tested SAPs were blanked with distilled water spectra.
Statistical analyses All data were processed using GraphPad Prism 5 software. Results of the MTS assays, AFM measurements and scaffold-based differentiation tests were analyzed with one-way ANOVA analysis followed by Tukey’s comparison test. Results of the NSC differentiation in the presence of the soluble KLPGWSG peptide were analyzed by paired t-test. In all cases a P value
