Introduction

Keratins from animal fibres are fibrous protein, characteristically abundant in cysteine residues (7–20 % of the total amino acid residues) that can be obtained from renewable source [1]. In fact, every year, a great amount of short fibres and crude fibres is abandoned in the rangeland during the shearing, representing a significant cause of environmental pollution [2]. Since they are chemically identical to longer fibres, which are used for textile applications, a potential natural resource is not being harvested or utilised. To study the recycling process of the aforementioned, keratin waste is, therefore, of paramount importance to reduce the carbon footprint of this material, offering to it alternative scenarios to incineration or land filling. Most studies concerning the valorisation of these wastes are based on the extraction and purification of keratin from poor quality animal fibres by chemical cleavage of the disulphide bonds, which are responsible for the high stability of the protein [3, 4]. Composite manufacturing has emerged as an efficient strategy to upgrade the structural and functional properties of synthetic biopolymers for regenerative medicine applications. A logical consequence is the development of new hybrid biomaterials with enhanced properties obtained through the introduction of organic and inorganic nanofillers into biodegradable polymers [5, 6]. The design and preparation of multi-component polymer systems fully renewable represents a viable strategy to develop innovative multifunctional biomaterials with specific properties able to be used in tissue engineering.

Poly (l-lactide) (PLLA) is an important biodegradable biopolymer, approved by Food and Drug Administration (FDA), which can be obtained from renewable agriculture materials and widely investigated in several areas, such as tissue engineering, drug delivery, packaging, etc. [7–9]. However, despite its biodegradability and biocompatibility, the application of PLLA in tissue engineering is limited by its somewhat weak cell-recognisable signals [10, 11]. In order to improve the cell affinity of PLLA, proteins as collagen [12], fibroin [13] and keratin from various sources [4, 14–16] were used as a coating on the polymer surface or as a filler in the polymer bulk structure.

Keratin has been used as a biomaterial for a variety of biomedical applications, since, being a natural protein, has the ability to improve cell affinity of scaffolds for tissue engineering [15, 16] keratin is, therefore, widely used for various biomedical purposes such as wound dressing [17] and tissue engineering scaffolds [18], owing to its biocompatibility and biodegradability. For example, Yamauchi et al. extracted the keratin from wool to fabricate sponge scaffolds, whose suitability for long term cell cultivation was evaluated [19]. The overall objective of this work is to study the use of keratins extracted from Merino wool (KM) and Brown Alpaca fibres (KA) as filler for poly (lactic acid) biocomposites obtained from solutions of the polymer in chloroform. In particular, the study is aimed at investigating the efficiency of sulphitolysis as proteins extraction method on two kinds of keratins and the effects of the selected organic solvent useful for PLLA composite development, by concentrating on the morphology, molecular weights, supermolecular structure and thermal behaviour of keratin from KM and KA.

Materials and methods

Materials

Australian KM and KA were used as source of keratins. PLLA with a molecular weight (M n) of 120000 g mol−1 and a polydispersity index (M w/M n) of 1.27 was supplied by Purac Biochem. All the reagents were purchased from Sigma-Aldrich.

Keratin extraction from Merino wool and Brown Alpaca fibres

Keratins were extracted from KM and KA using the sulphitolysis reaction according to the previously reported method [20]. The dried fibre samples (5 g) were cleaned by Soxhlet with petroleum ether to remove fatty matter, washed with distilled water, and conditioned at 20 °C, 65 % RH for 24 h. The clean and conditioned wool fibres were cut into snippets and treated with 100 mL of a solution containing urea (8 M) and sodium metabisulfite (0.5 M), adjusted to pH 6.5 with NaOH (5 M), under shaking for 2 h at 65 °C. The mixture was filtered through a 5-μm pore-size filter and dialysed against distilled water in a cellulose tube (molecular cut off 12–14 kDa) for 3 days, changing the outer solution four times a day. The keratin aqueous solutions obtained after dialysis were freeze-dried in order to recover keratin powders. Keratin morphologies, after the freeze-drying process, were investigated by means of a field emission scanning electron microscopy (FESEM, Supra 25-Zeiss). Small amounts of keratin powders were put on a conductive adhesive, gold sputtered and visualised.

Optimisation of the keratin regeneration process in organic solvent

Keratin extracted from KM and KA fibres were dispersed in two different organic solvents, and the dispersion methods and parameters were studied in order to optimise the following keratin distribution in PLLA matrix. Chloroform (CHCl3) and tetrahydrofuran (THF) were selected as organic solvents for keratin regeneration because these two specific solvents can be considered suitable for PLLA casting and preparation of keratin containing PLLA composites. An ultrasound tip homogenizer (Vibracell 75043, 750 W, Bioblock Scientific) at 40 % of amplitude for 2 min, 15 min and 30 min, in ice bath was used as dispersion equipment, and the effect of the treatment times was evaluated. Moreover, an ultrasound bath treatment (Ultrasonicbath-mod.AC-5, EMMEGI) for 2 h was considered as an alternative method for keratin regeneration in the two selected organic solvents and the effect compared with other treatments and parameters.

Keratin characterisations

Mean diameter of KM and KA were measured by Optical-based Fibre Diameter Analyser.

The yield of keratin extraction from wool and Alpaca is evaluated using following Eq. 1

$$ Y\left( \% \right) = \frac{KP}{{W_{d} }}x100 $$
(1)

where KP (mg) is the weight of the freeze-dried keratin powders extracted from the two fibre samples, and W d (mg) is the weight of the initial fibre samples dried at 105 °C for 2 h.

Visual observation and the sedimentation times of the solutions obtained with different treatment methods and parameters were evaluated, while the morphology of the protein re-dispersed in THF and CHCl3 with a tip sonication for 30 min was evaluated by means of a field emission scanning electron microscopy. Few droplets of the solutions were cast on a silicon wafer, gold sputtered and visualised.

The molecular weight distributions of keratin powders extracted from fibre samples, both untreated and treated with CHCl3 were determined by sodium-dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE), according to Laemmli’s method [21]. Freeze-dried powders were dissolved in Tris/HCl (550 mM, pH = 8.6), dithiothreitol (DTT, 140 mM), ethylene diamine tetra acetic acid (EDTA, 5 mM) and urea (8 M) overnight under nitrogen atmosphere. After that, the prepared solutions were properly diluted with a sample buffer containing NuPAGE LDS Sample Buffer and NuPAGE Sample Reducing Agents recommended by Invitrogen protocol, in order to deliver a 30 μg sample to the gel. The SDS-PAGE was performed using a Xcell SureLock Mini-Cell (Invitrogen), on 4–12 % polyacrylamide pre-cast gel (Invitrogen) using NuPAGE MES SDS Running Buffer, suitable for proteins with molecular weights from 188 to 3 KDa, referring to myosin, bovine serum albumin, glutamic dehydrogenase, alcoholdehydrogenase, carbonic anhydrase, myoglobin, lysozyme, aprotinin and insulin B chain as the molecular weight markers (Seeblue Pre-stained Standard, Invitrogen).

FT-IR spectra were acquired using a Thermo Nicolet Nexus Spectrometer, by placing the powders dried at 105 °C for 1 h, on a diamond crystal mounted on the ATR cell. The keratin treated with CHCl3 was investigated in transmission mode. Few droplets of the solutions were cast on a silicon wafer, and then analysed. The spectra, recorded in the 4000–650 cm−1 range using 100 scans and a resolution of 4 cm−1, were baseline corrected and smoothed with a nine-point Savitzky–Golay function. The region of Amide I and Amide II bands (1750–1450 cm−1) was resolved in Gaussian shape bands that were used to identify the secondary structures assumed by the protein. The fitting procedure was done with the ORIGIN 8.1® (OriginLab Corporation, MA, USA) defining the peak numbers by second order derivative analysis of the spectra. The area of each fitting peak was integrated and normalised for the Amide I and Amide II total area, in order to evaluate the percentage of α-helix, β-sheet and disordered secondary structures. The fitting results were further evaluated by examining the residual from the difference between the fitted curve and the original curve and accepted when the R2 was higher than 0.9900. Differential scanning calorimeter (TA Instrument, Q200) measurements were performed in the temperature range from −50 to 400 °C, at 10 °C min−1, in nitrogen atmosphere, both for untreated and chloroform-treated keratin powders. Thermogravimetric measurements (TGA) of KM and KA, untreated and treated with CHCl3, were performed using a Seiko Exstar 6300. Heating scans from 30 to 900 °C at 10 °C min−1 in nitrogen atmosphere were performed for each sample.

Energy Dispersive X-ray (EDX) analysis and FT-IR investigation of the residual material after calcination of the keratins at 600 °C for 3 h were also performed, in order to further elucidate the chemical structure of the degraded keratins.

Preparation of PLLA/keratin biocomposite films: development and characterisation

Biocomposite films were prepared by solvent casting method in chloroform. Keratins (both KM and KA) were dispersed in CHCl3 using the tip sonicator (Vibracell Sonics mod. 750 W) for 30 min in ice bath. PLLA was mixed with keratin dispersions, by magnetic stirring for 5 h and, after it was completely dissolved, the mixture was cast onto a Teflon® substrate and air dried at room temperature for 24 h, and for a further 48 h in vacuum. Films of 80 mm in diameter and 0.2-mm thick were obtained. The polymer/solvent ratio was chosen as 10 % (wt/v). PLLA/keratin samples containing 1 and 5 wt% as respect to PLLA initial weight and designed as PLLA/1KM or PLLA/1KA and PLLA/5KM or PLLA/5KA, respectively, were prepared. Neat PLLA film was also prepared by solvent casting for comparison. The microstructure of PLLA/keratin biocomposite films was investigated by scanning electron microscope. The sample surfaces were sputtered with gold and then analysed. Optical images were also performed by an inverted microscope, Nikon Epiphot 300. The mechanical behaviour of neat PLLA and PLLA biocomposites was evaluated by tensile tests, performed on rectangular probes (50 × 10 mm2) on the basis of UNI EN ISO 527-5 with a crosshead speed of 1 mm min−1, a load cell of 500 N and an initial gauge length of 25 mm. The specimens were dried in a vacuum oven at 40 °C for 72 h, then cooled in a desiccator and immediately tested. The measurements were carried out at room temperature and at least five samples for each formulation were tested.

Results

Keratin extraction yield

In order to extract keratin from animal fibres, the typical disulphide covalent bonds of the protein must be chemically broken through reduction or oxidation reactions. However, because nearly all oxidising reagents are toxic and difficult to apply at industrial level, in general sulphitolysis reaction (Eq. 2) is used as extraction method [22]. During sulphitolysis, the disulphide covalent bonds are broken into cystein ([RS −]) and cystein-S-sulphonated ([RSSO] (2−)3 ) residues, as shown in Eq. 2.

$$ RSSR + SO_{3}^{2 - } \leftrightarrow RS^{ - } + RSSO_{3}^{2 - } $$
(2)

Brown Alpaca fibres (Fig. 1a) have a mean diameter larger than KM (Fig. 1d) (Table 1) and unlike KM, they are characterised by the presence of egg-shaped pigment granules (melanines) of about 0.5 μm diameter, mainly distributed within the cortical cells [23].

Fig. 1
figure 1

Images of Alpaca (a) and Merino wool (d) fibres observed at the scanning electron microscope; Alpaca (b) and Merino wool (e) freeze-dried powders; Alpaca (c) and Merino wool (f) powders observed by FESEM

Full size image
Table 1 Fibres properties and keratin extraction yields
Full size table

As reported in Table 1, even though the keratin extraction yield is lower when coarser fibres are used, the very low extraction yield of keratin from Alpaca compared to Merino wool was also attributed to the presence of closely packed pigment granules, which reduce the efficiency of the filtration steps during the extraction procedure, causing a significant loss of material.

Keratin structure and effect of CHCl3

Figure 1 shows the freeze-dried powders of keratins extracted from Alpaca (Fig. 1b) and Merino wool (Fig. 1e) and their microstructure observed by FESEM (Fig. 1c, f). Both the keratins showed a flask-like structure with some external filaments, more evident in the case of KM. As reported by Cardamone et al., a flake structure was identified in lyophilised powder of keratin [24]. Moreover, the visual observation of KM and KA underlined that unlike white keratin extracted from Merino wool fibres, keratin extracted from Alpaca fibres has a light brown colour, which is probably due to the presence of pigment granules [25, 26].

In order to establish the respective effect of two organic solvents, useful for the subsequent composite production, and the influence of the dispersion methods on the original properties of the proteins, visual observation and studies of the sedimentation times of the solutions in THF and CHCl3, obtained with different treatment methods and parameters were evaluated.

Figure 2 shows the results of the visual observations and sedimentation times (Panel A) and the microstructures of KM and KA, re-dispersed in organic solvents and deposited on a silicon wafer (Panel B). The ultrasonic tip treatment of the KA in chloroform for 2 min was not sufficient to guarantee the dispersion of the keratin that remains agglomerated in the organic solvent, while a homogenous and stable solution was obtained with 30 min of sonication (Fig. 2, Panel A). Moreover, comparable stability in organic solvent was obtained by means of an ultrasound bath treatment conducted for 2 h (Fig. 2, Panel A) and similar results were also obtained for KM in chloroform (images not shown). Concerning the selection of the organic solvent, Fig. 2, Panel B underlined the capability of the chloroform to dissolve the keratin powder, while a considerable amount of agglomerates and flask-like structures of the original keratin morphology remain when THF was used as medium. The results suggested that the best conditions were obtained with CHCl3 and using the tip sonicator for a time of 30 min in ice bath. The selected solvent and the methods applied guaranteed the dispersion of both keratins in chloroform and the stability of the solutions for over a week. The keratins regenerated in chloroform were further characterised in order to establish the effect of organic solvent and treatments on the protein structure.

Fig. 2
figure 2

Panel A visual observations of sedimentation times after dispersion in ultrasonic bath (a) and by means of ultrasonic tip (b and c) at different times for KA; Panel B Microstructures of KM and KA deposited on a silicon wafer after 30 min of ultrasound tip dispersion in chloroform (a, b) and tetrahydrofuran (c, d)

Full size image

Effect of CHCl3 on molecular weight distribution and secondary structure of keratins

Figure 3 depicts molecular mass distribution of keratin extracted from Merino wool and Brown Alpaca fibres, both untreated and treated with CHCl3. The electrophoresis patterns of both keratins KM (Fig. 3, lane 2) and KA (Fig. 3, lane 4) show two major bands between 62 and 43 kDa and minor bands between 28 and 9 kDa. The high-molecular-weight fraction is referred to low-sulphur content keratin proteins deriving from intermediate filaments while the low-molecular-weight fraction is referred to high-sulphur content keratin associated proteins deriving from the interfilament matrix (KAPs) [26, 27]. The SDS-PAGE analyses suggest that the treatment with CHCl3 solvent did not degrade the KM and KA keratins; in fact, the treated proteins showed the same molecular weight distribution of the untreated one (Fig. 3, lanes 3 and 5).

Fig. 3
figure 3

SDS-PAGE gel of marker (lane 1), untreated KM (lane 2), untreated KA (lane 4), KM treated with CHCl3 (lane 3), KA treated with CHCl3 (lane 5)

Full size image

Molecular secondary structure of keratin can be revealed by FT-IR spectroscopy. Infrared adsorption spectra of KM and KA untreated and treated with chloroform, reported in Fig. 4, show characteristic absorption bands assigned mainly to the amide bonds (–CONH–). The broad band at about 3200 cm−1 (amide A band) can be correlated to the N–H and O–H stretching vibrations, the amide I band at approximately 1600-1700 cm−1 is connected mainly with the C=O stretching vibrations, while the amide II band, which falls at approximately 1550 cm−1, is related to N–H bending and C–N stretching vibrations. Finally, the amide III band, which occurs in the 1200-1300 cm−1 range, results from in phase combination of C–N stretching and in N–H in plane bending, with some contribution of C–C stretching and C=O bending vibrations [28]. Figure 4 shows two further adsorption peaks at 1195 and 1021 cm−1, which are related to the asymmetric and symmetric S–O stretching vibrations of the cysteine-S-sulphonated residues, respectively [29].

Fig. 4
figure 4

FT-IR spectra of KM and KA untreated and treated with CHCl3

Full size image

The amide I adsorption band is related to the carbonyl stretching vibration and, being particularly sensitive to supermolecular structures, is commonly used to determine the secondary structures of proteins [22, 30, 31]. In order to study the effect of chloroform in the proteins secondary structures, the amide I and amide II regions of each samples were resolved in seven Gaussian shape bands (Fig. 5) among which the band centred between 1657 and 1650 cm−1 (red curve) is assigned to the α-helix conformation, that centred between 1621 and 1631 cm−1 (blue curve) is related to β-sheet structure, while that falling between 1670-1697 cm−1 (green curve) comprises the β-turn and random coil secondary structures [25, 28]. In Fig. 6, the quantitative percentage of the amide I components are compared. It can be observed that treatment with CHCl3 causes a slight reduction of β-sheet structure in KM, while no significant changes were observed for α-helix and β-turn/Coil structures. Instead, a different behaviour was observed for KA, since, after the treatment with CHCl3, the reduction of β-sheet structure is matched with a significant increase of α-helix structure, while no changes were observed for β-turn/Coil structures.

Fig. 5
figure 5

FT-IR peak resolution of Amide I region of KM and KA untreated and treated with CHCl3

Full size image
Fig. 6
figure 6

KM and KA composition (in terms of α and β structures) for, untreated and treated keratins with chloroform

Full size image

Thermal behaviour of keratins regenerated from CHCl3

Figure 7a shows the DSC curves of KM and KA, both untreated and treated with chloroform and in Table 2 peak temperatures and related enthalpy values are reported. All DSC curves present three thermal events: a broad endotherm peak below 120 °C due to the water evaporation, endothermic peaks in the 200–260 °C due to the denaturation/degradation of crystalline regions at α-helices of intermediate filaments and uneven peaks at temperatures higher than 280 °C due to the degradation of keratin chains that comprise β-keratins and high-sulphur keratins [32]. It appears that for both samples KA and KM, the temperature peaks associated to water evaporation (T w) and the related enthalpy values (ΔH w) of chloroform-treated proteins are lower than that of untreated ones; this means that the amino acid side chains exposed to the chloroform on the outside of the protein chains could decrease the hydrophilicity of keratin. As regards the thermal denaturation/degradation of α-helix crystalline regions, it is possible to observe two different behaviours of KM and KA keratins. Starting from KM keratin, the treatment with CHCl3 results in a decrease of the enthalpy associated to α-helix unfolding (ΔH α), a reduced shift of the related peak (T α) to a lower temperature (from 222 °C up to 218 °C) and a broader endothermic peak. Being the ΔH α correlated with the content of α-helix crystalline regions, this means that the treatment with CHCl3 reduces the amount of α-helix, in agreement with the previous results obtained by FT-IR analysis. Moreover, the decrease of the T α value suggests the presence of less thermally stable α-helices after CHCl3 treatment whereas the broader peak indicates that CHCl3 treatment reduces the α-crystallites perfection [33]. In contrast, the KA keratins show an endothermic peak at 215 °C followed by a second maximum at 227 °C due to α-helices at different thermal stability. In the KA keratin treated with CHCl3 the two peaks are superimposed and the area under the peaks increases. Therefore, in agreement with the FT-IR analysis, the treatment with CHCl3 increases the amount of α-helix crystallites with different thermal stability.

Fig. 7
figure 7

DSC curves of KM and KA untreated and treated with CHCl3 (a); Derivative thermogravimetric curves (DTG) of KM and KA untreated and treated with CHCl3 (b); FT-IR spectra of calcinated untreated keratins (c) and EDX results of residual material from calcination d

Full size image
Table 2 Thermal parameters from DSC measurements
Full size table

The decomposition step was also studied by thermogravimetric analysis. Derivative thermogravimetric curves for pristine and regenerated from chloroform keratins, extracted from KM and KA, are reported in Fig. 7b. The comparison of KM with KA evidenced that keratins extracted from Alpaca have reduced thermal stability with respect of keratins from KM, but both fibres have a common thermal behaviour when degraded in nitrogen atmosphere. The initial weight loss at 100–150 °C is due to moisture loss, while the major weight losses took place in the temperature range from 200 to 300 °C. The α-helix denaturation occurs in the temperature range of 200–250 °C [34, 35], while the second weight losses that took place in the temperature range from 250 to 400 °C is due to the decomposition step of β-sheet structures and high-sulphur content keratins [36]. These steps correspond to the weight loss caused by the decomposition of the protein fibre structure [37], and coincide with the temperature range over which hydrogen bond peptide helical structure ruptures and the ordered regions of keratins undergo degradation. Several chemical reactions occur in this region where protein compounds are decomposed to lighter products and volatile compounds such as H2S, CO2, H2O, HCN are released [38].

The peak found only in KA, before and after regeneration, at about 750 °C is probably due to the pigmented nature of the keratin. As evidenced by TGA analysis, the interaction with chloroform affected the maximum rate of both α-helices degradation and of the decomposition step. In particular, the α-helices degradation rate was 0.0023 and 0.0037 (%/ °C), respectively, for KM and KA before regeneration, while the value for the same systems increased to 0.029 and 0.0050 (%/ °C) after treatment with the organic solvent. Therefore, in agreement with DSC analysis, this behaviour suggests that the α-crystallites formed by casting of keratin/chloroform systems are thermally less stable. An opposite trend was otherwise revealed over the last decomposition step. In particular, the second maximum degradation rate was 0.0048 and 0.0056 (%/ °C), respectively, for KM and KA before regeneration, while the value for the same systems decreased to 0.0044 and 0.0049 (%/ °C) after treatment. Moreover, the maximum temperature of thermal decomposition for Merino wool and Alpaca keratins regenerated from chloroform was slightly lower compared with the natural keratins, reflecting the comparably lower thermal stability of the former. The reason might be related to the presence of a lower amount of crystalline β-sheet structure and, consequently, attenuated intermolecular interaction between the protein chains. In order to better understand the structure of the thermally degraded samples, the keratins were calcinated at 600 °C for 3 h and FT-IR analysis of the residual material was performed. As can be seen in Fig. 7c, in the spectra of both calcinated Merino and Alpaca keratins, the amide bands disappear indicating that the proteins and amino acids have been totally degraded. The degradation products show intense absorption peak at about 1130 cm−1 probably assigned to C–O stretching vibrations of ethers groups, a weaker absorption band around 1440 cm−1 due to the asymmetric and symmetric bending vibration of CH2 and CH3 groups and a broad band between 3250 and 3700 cm−1 related to the N–H and O–H stretching vibrations. In Fig. 7d, by EDX examination, the absence of heavy metal elements was proved, while the residual material presence consist of carbonate, oxygen, sodium and sulphonium elements related to the pigmented nature of the KA keratin, was demonstrated [39]. According to this, the presence of carbonates in the spectrum of the residual-calcinated KA corroborates the attribution of the experimental peak observed between 650 and 800 °C, to metal carbonates (such as Na- or Ca-based carbonates), in that, these undergo thermal decomposition in that temperature range in inert atmosphere [40]. Moreover, according to [41], wool contains significant quantities of calcium, potassium, sodium, zinc, copper, manganese, iron and selenium [42], and some of these elements were also found in the spectrum of residual material of both calcinated KA and KM.

PLLA/keratin biocomposite films

PLLA biocomposite films reinforced with different amount (1 and 5 wt%) of keratin from KM and KA were successfully developed by solvent casting method in chloroform, the selected solvent for keratin regeneration. PLLA films containing keratin from Alpaca fibres show a light brown aspect, due to the pigment in the keratin itself, while PLLA films containing keratin from KM have white/transparent colour. Figure 8 shows the FESEM images at different resolution of the PLLA, PLLA/1KA, PLLA/5KA, PLLA/1KM, PLLA/5KM biocomposite film upper surfaces. The PLLA upper surface, exposed to air during the solvent casting process, appears flat, while a different behaviour was detected for composite systems, due to the presence of keratin in the PLLA polymer matrix. PLLA/5KA, PLLA/1KM and PLLA/5KM show the presence of a superficial and circular porous structure with a pore diameter ranging from 1 to 6 μm for Alpaca, and from 2 to 10 μm for KM-based formulations. However, in the PLLA/1KA film a defined porous structure is not identified, and only an increase in surface roughness was evident for this system. The specific round-like surface topography is the result of two combined phenomena during the composite processing: the chloroform evaporation and the presence of keratin flasks. These structures, in fact, were not observed in the pristine PLLA film. The process of pore formation is controlled by the rapid volatile solvent evaporation. It is clear that the solvent vapour pressure has a critical influence on the pore formation, associated with the changes in the solubility parameters induced by the presence of regenerated keratin. To obtain a good distribution of the keratin in the PLLA is an important target to proceed to the manufacturing of new biocomposite films with modulated properties induced by the presence of different amount of keratin. For this reason, the internal microstructure of the biocomposite films was investigated by optical microscope. Figure 9a shows the keratin distribution in the PLLA matrix, and for the PLLA/1KA, PLLA/5KA, PLLA/1KM and PLLA/5KM, at different resolution (×10, ×20), Biocomposite films show the presence of random distribution of keratin in flask-like structure with some external filaments. In order to test the mechanical behaviour of produced films, results of tensile tests (Young’s modulus) for neat PLLA and keratin-based biocomposites are reported in Fig. 9b. As commented in [43] and [44, 45], chemical–physical, topographical and mechanical properties of biomaterials, such as the stiffness or elasticity of the polymeric matrix, have a clear effect on the induction of stem cell differentiation as well as on the preservation of their stem cell status. So, according to this evidence, elastic properties of the PLLA biocomposites containing the two keratins were evaluated. With exception of the 1KA-based system, higher elastic moduli were obtained with respect to neat PLA matrix with increasing content of both keratins. The addition of keratin resulted in an increase of tensile modulus from (600 ± 50) MPa for neat PLLA film up to (1200 ± 70) MPa for the PLLA/1KM system, confirming the modification of the mechanical performance of the film in presence of keratins. Preliminary results of stem cell viability on the different substrates monitored through the XTT salt solution (data not shown) underlined that all PLLA-based composites allow stem cell growth as in Tissue Culture Plastics; however, a deeper investigation will be carried out to determine how the elastic property of the film could influence the interactions of PLLA composites with cells.

Fig. 8
figure 8

FESEM images at different resolution of the PLLA, PLLA/1KA, PLLA/5KA, PLLA/1KM, PLLA/5KM biocomposite films upper surfaces

Full size image
Fig. 9
figure 9

Optical images of keratin distribution at different resolution (10×, 20×) (a) and values of Young’s modulus (b) for the PLLA film and PLLA/1KA, PLLA/5KA, PLLA/1KM and PLLA/5KM biocomposites

Full size image

Conclusions

In this work, keratin proteins were successfully extracted from KM and KA by sulphitolysis reaction. The extraction yield of keratins is lower for KA than KM and could be due to two (or more) factors, especially the higher mean diameter of KA compared to KM and the presence of the pigment granules posing a hinder to the filtration process. Comparing the effects of the two solvents used in this study, THF and CHCl3, the latter offered the better keratin regeneration. Even if CHCl3 does not degrade the two kind of keratins, it affects in different manner the supermolecular structures and thermal behaviour of the two proteins. In particular, while for KM keratin it decreases the amount of α-helix structures, for Alpaca keratin an increase of helices was observed. In contrast, a slight decrease of β-sheet structures after CHCl3 treatment was observed for both keratins and the treatment with CHCl3 decreases the thermal stability of α-crystallites. Biocomposite films based on PLLA polymer matrix and two different contents of KM and keratin fibres were successfully developed by solvent casting in chloroform, with a random distribution of keratin in flask-like structure. PLLA/5KA and PLLA/5KM show a specific pore-like surface microstructure, induced by the solvent evaporation. Previous published studies found that keratin forms can have excellent intrinsic biocompatibility, biodegradability and mechanical durability [45]. Both KM- and KA-based PLLA biocomposite films hold promise as alternative materials for their potential biomedical application, such as wound dressing and scaffolds for tissue engineering.