- Email: [email protected]
Microstructure and physical–chemical properties of chicken collagen
Accepted Manuscript Title: Microstructure and Physical–Chemical Properties of Chicken Collagen Author: Anja Maria Oechsle Dila Akg¨un Franziska Krause Christiane Maier Monika Gibis Reinhard Kohlus Jochen Weiss PII: DOI: Reference:
S2213-3291(16)30002-8 http://dx.doi.org/doi:10.1016/j.foostr.2016.02.001 FOOSTR 44
To appear in: Received date: Revised date: Accepted date:
22-7-2015 22-1-2016 15-2-2016
Please cite this article as: Oechsle, Anja Maria., Akg¨un, Dila., Krause, Franziska., Maier, Christiane., Gibis, Monika., Kohlus, Reinhard., & Weiss, Jochen., Microstructure and PhysicalndashChemical Properties of Chicken Collagen.Food Structure http://dx.doi.org/10.1016/j.foostr.2016.02.001 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
MICROSTRUCTURE AND PHYSICAL–CHEMICAL PROPERTIES OF CHICKEN COLLAGEN Anja Maria Oechsle1, Dila Akgün1, Franziska Krause1, Christiane Maier1,
Monika
Gibis1, Reinhard Kohlus2, Jochen Weiss1,* 1
Department of Food Physics and Meat Science, Institute of Food Science and
Biotechnology, University of Hohenheim, Garbenstrasse 21/25, 70599 Stuttgart, Germany 2
Department of Food Process Engineering and Food Powders, Institute of Food Science
and Biotechnology, University of Hohenheim, Garbenstrasse 25, 70599 Stuttgart, Germany
Submitted to Food Structure in July, 2015 Running Title: Microstructure and Physical–Chemical Properties of Chicken Collagen To
whom correspondence should be addressed: [email protected], Tel.: +49
711 459 24415, Fax: +49 711 459 24446
1
Graphical abstract
Highlights
Chicken skin collagen resembles the fibrous bovine telopeptide-poor collagen most.
Chicken bone collagen reveals firm fragments and a low intrinsic viscosity proposing insufficient extraction.
The shear thinning behavior increases with the degree of crosslinks within the collagen.
The intrinsic viscosity is linked inversely to the collagen complex viscosity.
2
ABSTRACT The global consumption of sausages has increased immensely over the last few years, while bovine collagen has become scarce. Thus, collagen from alternative sources is being considered for application in the food industry. Therefore, chicken skin and bone collagen were characterized and compared with the bovine telopeptide-poor collagen aiming at the feasibility of producing pure, co-extruded chicken sausages. Hence, the chemical composition, microstructure and rheological properties of the different collagen samples were examined and SDS-PAGE, mass spectroscopy, and ζ-potential analysis were conducted. Weak bands in the SDS-PAGE gel indicated only partial maceration of chicken bone collagen, whereas chicken skin and telopeptide-poor collagen revealed distinct bands indicating collagen type I and III. This was also verified by mass spectroscopy. Large fragments were visible in optical microscopy for chicken bone collagen, whereas chicken skin collagen revealed a delicate network. Moreover, the highest dynamic consistency index, at 14146 Pa s n*, was determined for chicken bone collagen, followed by chicken skin and telopeptide-poor collagen at 606 and 320 Pa s n*, respectively. By contrast, the intrinsic viscosity was the highest for telopeptide-poor collagen (3.16 - 3.26 L/g), whereas chicken bone collagen exhibited the lowest value at 0.13 L/g, suggesting poor swelling behavior. Moreover, telopeptide-poor collagen featured the highest dynamic power law factor, suggesting the least crosslinks, serving a negative standard rather than an actual chicken collagen analog. Finally, chicken skin collagen displayed the most suitable source of collagen for the coextrusion process compared to the well established bovine hide split collagen. Keywords: Chicken skin collagen; chicken bone collagen; bovine telopeptide-poor collagen; intrinsic viscosity
3
1
INTRODUCTION
Collagen represents a family of proteins that serve as major structural components of connective tissues in vertebrates (Lee, Singla, & Lee, 2001). It has unique physical properties, including uniformity, tensile strength, flexibility, biocompatibility, and biodegradability. Therefore, it is used for a wide range of applications, e.g. as scaffold in tissue engineering, for implantations or wound dressing in surgical operations, as a capsule matrix material or binder in pharmaceutical applications, and for the production of gels and films in food (Osburn, 2002). A promising use of collagen in the food industry is the manufacture of coextruded sausage casings, due to the rapidly rising cost of natural, intestinal-derived casings (Barbut, 2010). Moreover, continuous processes allow the hygienic production of sausages in large quantities with consistent end product quality attributes (Irmscher et al., 2013; Irmscher et al., 2015; Osburn, 2002). Currently, bovine hide split collagen is used in the co-extrusion process (Bueker, Bueker, & Grolig, 2009). However, the global consumption of meat products, in particular sausages, is rising and supplies of bovine collagen are getting low. Moreover, the need of kosher and halal products is increasing constantly. As a result, chicken collagen could supply the need, since it might be easily gained as a by-product from skins, cartilages, bones, and feet from the poultry manufacturing industry. In addition to good availability, fibril-forming collagens are predestined for the production of edible sausage casings, as the stable network enables shrinking and stretching to accommodate the contraction and expansion of meat batter during processing (Osburn, 2002). Bovine and chicken skins predominantly contain the fibril-forming collagen types I and III (Abedin & Riemschneider, 1984; Bräumer, 1974; Gelse, Pöschl, & Aigner, 2003; Ramshaw, 1986). On a molecular basis, fibril-forming collagens feature an uninterrupted helical region with alternating polar and nonpolar domains leading to lateral alignment of molecules in a quarter staggered array (Reiser, McCormick, & Rucker, 1992). Collagen type I is a heterodimer composed of two identical α1-chains and one α2-chain (Bräumer, 1974; Gelse et al., 2003; Ramshaw, 1986), whereas collagen type III is a homotrimer of three α1(III)-chains and usually occurs in the same fibril with type I collagen (Kadler, Holmes, Trotter, & Chapman, 1996). Collagen stability and structure is based on hydrogen bonds between polar residues of 4-hydroxyprolin and 5-hydroxylysin, the formation of 4
hydration networks, and electrostatic interactions (Gelse et al., 2003). The latter ones emerge between ionizable side groups present in 15 – 20 % of all amino acid residues either in the X or in the Y position of the Gly-X-Y triplets (Chan, Ramshaw, Kirkpatrick, Beck, & Brodsky, 1997). Chicken collagen is less crosslinked due to the degree of covalent crosslinks increasing with advancing age due to lysyl oxidase-initiated crosslinks (Bailey & Shimokomaki, 1971). Thus, chickens are slaughtered after approximately 6 – 7 weeks (Maurer, 2003), while the average age of beef is 18 – 36 months (Joseph, 2003). The performance of telopeptide-poor collagen as a model system for less crosslinked collagen types, such as chicken collagen, was investigated in former studies of the author. We investigated the influence of acids, Hofmeister salts, and co-gelling proteins on native and telopeptidepoor bovine collagen rheology and microstructure (Oechsle, Häupler, Gibis, Kohlus, & Weiss, 2015; Oechsle, Landenberger, et al., 2015; Oechsle, Wittmann, Gibis, Kohlus, & Weiss, 2014). Thus, the aim of this study was to characterize chicken bone and skin collagen, which was provided by the industry partners due to their low-cost availability, and to evaluate their suitability for the extrusion process compared to a telopeptide-poor collagen. Therefore, results obtained from chemical analysis, microscopy, ζ-potential, mass spectrometry, and sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDSPAGE) analyses were compared. In this context, we hypothesized that chicken collagen behaves according to the telopeptide-poor collagen in terms of rheological and microstructural characteristics. Furthermore, we postulated that chicken skin collagen fibers stay intact, whereas the demineralization of bone leads to a break down of the collagen fibers, leading to weaker gels than telopeptide-poor collagen.
2 2.1
MATERIALS AND METHODS Materials
The raw material of the bovine telopeptide-poor collagen was kindly provided by the Kalle (Wiesbaden, Germany) and Protein Consulting (Singhofen, Germany) prepared the telopeptide-poor collagen by splitting off telopeptides and intermolecular crosslinks from 5
native collagen to obtain single collagen triple helices (Bueker et al., 2009). Chicken collagen was extracted using a procedure similar to that used for bovine collagen using chicken bone and skin as raw materials (Protein Consulting, Singhofen, Germany). For each samples one batch was provided by the industry partners that comprised the extraction product of several extraction procedures. 2.2
Chemical characterization
The standard analysis methods according to German law were used for the collagen characterization, according to § 64 Lebensmittel- und Futtermittelgesetzbuch (BVL, 2011). Dry matter content was determined according to L 06.00-3 (BVL, 2011). The ash (minerals) determination was conducted according to L 06.00-4 (BVL, 2011). The protein content in collagen samples was quantified applying the Kjeldahl method L 06.00-7, whereas a combination of two methods was used (L 07.00-41 and L 07.00-57) for nonprotein nitrogenous substances in the samples (BVL, 2011). The amount of collagen degradation products as well as hydroxyproline and connective tissue content in the collagen samples were determined according to L 06.00-8 (BVL, 2011). The lipid content was assessed by the method L 06.00-6 (BVL, 2011). 2.3
Sample preparation
The collagen samples were diluted with 0.001 M phosphoric acid to a concentration of 6 g/100 g based on their protein content and mixed with an Artisan 5KSM 150 equipped with a flex edge beater 5KFE5T (KitchenAid, St. Joseph, Minnesota, USA) for 5 min at the highest setting. Samples were subsequently adjusted to pH 3 with phosphoric acid (Th. Geyer AG, Renningen, Germany) prior to storage over night to enable the hydration of the collagen molecules. Phosphoric acid at pH 3 was found in a previous study to yield the lowest critical overlap concentration by promoting collagen interactions (Oechsle et al., 2014). Before measurement, the 6 g/100 g suspensions were diluted to concentration needed for the analysis as indicated in sections 2.4 - 2.7, homogenized for 5 min (300 rpm) using a Stomacher Circulator 400 from Seward (West Sussex, UK), and the pH was adjusted if necessary.
6
2.4
SDS-PAGE
Sodium dodecyl sulfate polyacrylamide gel electrophoresis was performed according to the method of Laemmli (1970). Solubilized collagen samples were mixed in a ratio of 1:1 (v/v) with the sample buffer containing 0.5 M tris(hydroxymethyl)aminomethane-HCl, pH 6.8 (Applichem, Darmstadt, Germany), 20 g/L SDS, 250 mL/L glycerol (VWR, Fontenay-sous-Bois, France), 50 mL/L β-mercaptoethanol (Sigma Aldrich, Munich, Germany), and 0.1 g/L Bromphenol Blue (Merck, Darmstadt, Germany). Consequently, the sample solution was denatured at 95 °C for 4 min. An aliquot of 10 µl of 1 g/L protein suspension was loaded onto Precast mini-Protean TGX gel (Bio-Rad Laboratories, Inc., Richmond, VA, USA) and subjected to electrophoresis at 200 V for 35 min using a MiniProtean Tetra System (Bio-Rad Laboratories Inc., Richmond, VA, USA). Subsequently, the gel was stained with Coomassie Brilliant Blue R-250 (Bio-Rad Laboratories Inc., Richmond, VA, USA) for 30 min, and afterwards, destained with 150 mL/L methanol and 100 mL/L acetic acid solution until the bands were clear. Roti-Mark prestained T852 Marker (Carl Roth, Karlsruhe, Germany) was applied to determine the molecular weight of the proteins. The contrast of the image was enhanced by imageJ 1.48v software (Wayne Rasbund, National Institutes of Health, Bethesda, MD, USA). 2.5
ζ-potential
The ζ-potential of 0.1 g/100 g collagen suspensions were measured between pH 3 and 7 using a particle electrophoresis instrument (Nano, ZS, Malvern Instruments, Malvern, UK) to determine the isoelectric point. The titration was conducted automatically with a multipurpose titrator (MPT-2, Malvern Instruments, UK). The pH was adjusted by 0.1 M hydrochloric acid (Carl Roth, Karlsruhe, Germany) and 0.1 M sodium hydroxide (Carl Roth, Karlsruhe, Germany). 2.6
Rheology
Experiments were conducted using a modular compact rheometer Physica MCR 502 (Anton Paar, Ostfildern, Germany) at 5 °C. Flow curves of dilute suspensions were conducted using a single gap cylinder geometry CC27 with 28.92 mm cup diameter and 26.66 mm bob diameter (Anton Paar, Ostfildern, Germany). The dynamic viscosity of the solvent ηs and 0.1 – 0.4 g/L collagen suspensions η were characterized to calculate the 7
reduced viscosity ηred and inherent viscosity ηinh in order to obtain the intrinsic viscosity [η] according to Huggins (1942) and Kraemer (1938) with Equation (1) and (2), respectively: 𝜂 − 𝜂𝑠 ) 𝑐→0 𝑐→0 𝑐 𝜂𝑠 𝜂 𝑙𝑛 (𝜂 ) 𝑠 [𝜂] = 𝑙𝑖𝑚(𝜂𝑖𝑛ℎ ) = 𝑙𝑖𝑚 ( ) 𝑐→0 𝑐→0 𝑐
(1)
[𝜂] = 𝑙𝑖𝑚(𝜂𝑟𝑒𝑑 ) = 𝑙𝑖𝑚 (
(2)
Moreover, 4 g/100 g collagen gels were analyzed by oscillating measurements using a plate-plate geometry PP25 with 25 mm diameter (Anton Paar, Ostfildern, Germany). Strain sweeps were performed at 1 Hz in order to determine the linear viscoelastic range. Consequently, frequency sweeps were performed in the linear viscoelastic range applying 1 % strain from 1 to 100 rad/s. Moreover, the complex viscosity η* was determined as a function of the angular frequency ω applying Equation (3) (Macosko, 1994): 2
2 1/2
𝐺′′ 𝐺′ 𝜂∗ = [( ) + ( ) ] 𝜔 𝜔
(3)
The power law relation of the complex viscosity η* as a function of the angular frequency ω according to Equation (4) was assessed in order to fit the dynamic consistency index k* and the dynamic power law factor n* (Keogh & O’Kennedy, 1998). 𝜂∗ = 𝑘 ∗ 𝜔 𝑛 2.7
∗
−1
(4)
Microscopy
Optical microscopy was conducted to visualize the microstructure of 4 g/100 g collagen gels. Unstained gel samples were compressed in between the microscope slide and the cover glass in order to minimize the collagen layer thickness and to visualize the collagen microstructure. Microscopy images were taken with an an Axio Scope optical microscope (A1, Carl Zeiss Microimaging GmbH, Göttingen, Germany) equipped with a digital camera (AxioCam ICc3, Jena, Germany). A magnification of 5x was used. The samples were dipped into liquid nitrogen and freeze-dried (Lyovac GT 2, Finn Aqua Santasalo-Sohlberg, Hürth, Germany and Lyovac GT 4, LH Leybold, Köln, Germany) for approximately 48 h for the scanning electron microscopy (SEM). The samples were 8
sputtered with gold and palladium (20:80) for about 8 min with a sputtering system (SCD 040, Balzers, Bingen, Germany). The images were taken with a scanning electron microscope from Zeiss (DSM 940, Zeiss, Hamburg, Germany) under vacuum at 5000 kV. A magnification of 500x was applied. 2.8
Statistical analysis
We prepared and measured the samples for the measurements at least twice using at least duplicate samples for each experiment. The number of measurements i and number of repetitions of one dilution n are indicated in the figure captions. Means, standard deviations, and the significance of determination were calculated from themeasurements using Excel (Microsoft, Redmond, WA, USA) and SigmaPlot 12.3 (Systat Software, Inc., San Jose, CA, USA). 3 3.1
RESULTS AND DISCUSSION Collagen type and composition
The chemical composition of the collagen samples is listed in Table 1. Telopeptide-poor collagen was already utilized in previous studies (Oechsle, Häupler, et al., 2015; Oechsle et al., 2014) to signify the effects on microstructure and rheology for less crosslinked collagen types, such as chicken collagen. The telopeptide-poor collagen sample used in this study, however, had a different chemical composition. In particular, the mineral content differed considerably, which might be attributed to variations in beef type, age and breeding conditions, since the extraction method was standardized according to the supplier. Due to the different drying levels of the samples extracted, chicken bone collagen revealed the highest dry matter content with 41.29 g/100 g, followed by chicken skin collagen and telopeptide-poor collagen with 34.24 g/100 g and 9.88 g/100 g, respectively. Accordingly, the chicken bone collagen served the highest connective tissue content that corresponds to the collagen concentration with 34.70 g/100 g, followed by chicken skin collagen and bovine telopeptide-poor collagen with 20.89 g/100 g and 7.22 g/100 g, respectively. We expected the protein and connective tissue content to be identical in each of the three collagen samples. The deviations might be ascribed to inaccuracies of the photometrical measurement of the sensitive connective tissue determination. Moreover, the protein concentration is calculated from the nitrogen 9
concentration using the factor 5.55 according to BVL (2011). This factor is based on the occurrence of nitrogen in proteins with higher nitrogen content such as gelatin; however this might lead to inaccuracies, since we did not determine the actual nitrogen concentration in the collagen samples by an independent method. According to the connective tissue content the samples were diluted to the concentration desired for analysis. Figure 1 depicts the SDS-PAGE gel with bovine telopeptide-poor collagen featuring two distinctive bands at approximately 123 kDa indicating α1(I)- and α2(I)-chain monomers of type I collagen. Furthermore, the band of the slightly larger α1(III)-chain was visible above, indicating collagen type III. Mass spectrometry analysis (supplementary data) verified the presence of α1(I)- , α2(I)-, and α1(III)-chains with 133, 129 and 138 kDA, respectively. Similar to telopeptide-poor collagen, chicken skin collagen delivered α1(I)and α2(I)-bands in the SDS-PAGE gel. Mass spectrometry revealed type I and III collagen, although no α1(III)-band was evident in the SDS-PAGE gel. In this way, Abedin and Riemschneider (1984) demonstrated collagen type I and III in chicken skin collagen. The absence of α1(III)-bands within the SDS-PAGE gel can be explained by the fact that α1(III)-chains might get chemically reduced during extraction and appear at the position of α1(I)-bands in SDS-PAGE (Wu et al., 2011). Chicken bone collagen revealed solely type I collagen via mass spectroscopy analysis indicated by α1(I) and α2(I)-chains, as reported by Strawich and Glimcher (1983). The SDS-PAGE gel exhibited only weak bands. However, large fragments were visible to the naked eye in the chicken bone collagen sample, indicating insufficient extraction, leading to incomplete solubilization during SDS sample preparation. Supporting our findings, Glimcher and Katz (1965) reported that very little of the collagen of demineralized bone, even in young rapidly growing animals, could be extracted in solutions normally used to extract collagen with existing tertiary structure from unmineralized bone. On the other hand, they reported that a large fraction of the chicken bone collagen was found to be extracted as gelatin, signifying the complexity of the extraction procedure. Overall, the protein sizes determined via mass spectrometry did not correspond to the protein sizes perceived by SDS-PAGE. This might be due to fibrillar collagen molecules featuring a higher electrophoretic mobility within the SDS-PAGE gel and, therefore, they appear smaller 10
than globular marker proteins, namely bovine myosin (245 kDa) and rec. E.coli βgalactosidase (123 kDa). For all three samples, β-bands and γ-bands were visible below and above 245 kDa representing dimers and trimers, respectively, as already observed by Oechsle, et al. (2015). 3.2
ζ-potential
Figure 2 demonstrates exemplarily a measurement of the ζ–potential as a function of the pH. Bovine telopeptide-poor collagen displayed an isoelectric point at pH 4.62 ± 0.07, followed by chicken bone collagen and chicken skin collagen at pH 5.09 ± 0.09 and 6.54 ± 0.32, respectively. The expected isoelectric point of bovine collagen was reported to be located at pH 4.7, which matched the data collected for bovine telopeptide-poor collagen (Highberger, 1939). However, both chicken collagen samples revealed higher isoelectric points, which might be ascribed to different animal species, leading to alterations in the amino acid side groups and net charge, since the extraction was based on the patent established for bovine collagen by Bueker, Bueker, and Grolig (2009). Moreover, only the ζ-potentials of the soluble fractions of the chicken bone collagen sample might have been detected due to its inhomogeneity. Consequently, the results should be considered with caution, since the solubilized collagen fraction might discriminate vigorously from the insoluble collagen bound to the bone matrix. Nonetheless, all three collagen samples exhibited a positive net charge at pH 3, signifying that all gels feature repulsive forces preventing collagen from precipitation and facilitating hydration. 3.3
Rheology
The reduced viscosity ηred and inherent viscosity ηinh are displayed in Figure 3 as a function of the collagen concentration with the y-intersect being the intrinsic viscosity [η] listed in Table 2. The intrinsic viscosity [η] is a measure of a polymer’s ability to increase the viscosity of a solvent (De Gennes, 1976). Telopeptide-poor collagen delivered the highest [η] with 3.16 – 3.26 L/g. Kahn and Witnauer (1966) determined an intrinsic viscosity of 2.7 L /g for calf collagen diluted in citrate buffer at pH 3.5. This is in rough accordance with the telopeptide-poor collagen. The slight variation might be due to different conditions such as pH and buffer, which also affect the conformation and the 11
interaction with the bulk media. By contrast, the chicken skin and bone collagen samples revealed 10- and 30-fold lower values of 0.28 – 0.31 L/g and 0.13 L/g, respectively. Hence, the telopeptide-poor collagen molecules might either exhibit the largest hydrodynamic volume or the lowest molecular weight in order to get a lower intrinsic viscosity, since these parameters are inversely correlated in Equation (5) (Hiemenz & Lodge, 2007). [𝜂] =
5 𝑁𝐴𝑉 𝑉ℎ 2𝑀
(5),
with the Avogadro constant NAV, the hydrodynamic volume Vh and the molecular weight M. As mentioned before, extraction of the chicken bone collagen might have been too faint, leading to an inhomogeneous sample with larger fragments, which, in turn, might not be able to swell and, therefore, not be able to affect the bulk water. As a consequence, unswollen collagen structures do not form a homogeneous gel and are therefore unable to yield homogenous thin films after extrusion. Supporting the hypothesis stated above, Glimcher and Katz (1965) reported that chicken bone collagen is not able to swell in mild acidic conditions. However, they also postulated that very few intra- or intermolecular covalent crosslinks are present in chicken bone collagen, signifying strong noncovalent intermolecular forces playing the major role in its relative insolubility. Furthermore, the solubilized chicken bone collagen molecules might feature a smaller molecular weight that also contributes to a lower intrinsic viscosity (Katz, Francois, & Glimcher, 1969). Sakai, Ikeda, and Isemura (1967) determined a [η] of 1.25 g/L for soluble chicken leg tendon collagen which is higher than the [η] of both chicken collagen samples. The divergence might be ascribed to the fact that in addition to type I collagen, chicken tendon also contains type V and III collagen (Birk, Hahn, Linsenmayer, & Zycband, 1996), resulting in a higher swelling capacity, which, in turn, influences the intrinsic viscosity. Moreover, Sarbon, Badii, and Howell (2013) stated an intrinsic viscosity for chicken skin gelatin of 0.15 L/g, which is in agreement with the chicken bone collagen results, implicating that the bone collagen sample might be degraded. Supporting this hypothesis, Glimcher and Katz (1965) stated that 80 % of the collagen from chicken bone was extracted as gelatin featuring a smaller hydrodynamic volume, due to the random coil conformation, rather than the tertiary triple helical fiber structure of collagen. 12
Figure 4 demonstrates the complex viscosity η* that was measured in the linear viscoelastic region regarding the angular frequency of the collagen samples. Hereby, chicken bone collagen exhibited the highest η*, followed by chicken skin and telopeptide-poor collagen. The complex viscosity comprises a real and imaginary component that is associated with the storage and loss modulus, respectively, suggesting that the chicken bone collagen features the highest moduli. However, the results need to be considered carefully, since SDS-PAGE analysis and the intrinsic viscosity implemented collagen molecules being extracted insufficiently. Hence, large fragments might have misled the rheological analysis by increasing the torque. Therefore, η* appeared to be higher than the suspension actually was. Meeting expectations, the lowest η* was determined for telopeptide-poor collagen, which might be ascribed to the low molecular weight based on the absence of intermolecular crosslinks and telopeptide ends. This phenomenon was also observed in a previous study where the highly crosslinked native collagen delivered a lower [η] and a higher storage modulus than telopeptide-poor collagen (Oechsle et al., 2014). The dynamic consistency index k* and dynamic power law factor n*, as listed in Table 3, were deduced from the power law fit of Figure 4. As for the complex viscosity, the dynamic consistency index k* describes the consistency and was the highest for the chicken bone collagen, followed by chicken skin and telopeptide-poor collagen with 15146, 606, and 320 Pa s
n*
, respectively. The dynamic power law factor n* indicates a
shear-thickening (n* > 1), Newtonian (n* = 1) or shear-thinning (n* < 1) behavior (Rao, 2007). Chicken bone collagen revealed the highest shear-thinning behavior with n* = 0.11, followed by chicken skin collagen with n* = 0.13 and telopeptide-poor collagen with n* = 0.31. In a previous study (unpublished data), the highly crosslinked native collagen displayed a dynamic power law factor n* of 0.14. Consequently, a connection between increasing covalent crosslinks and decreasing shear-thinning behavior might exist for the collagen samples analyzed. As a result, chicken skin and bone collagen might contain more intra- and intermolecular crosslinks than telopeptide-poor collagen. Thus, telopeptide-poor collagen is actual a negative standard with the least crosslinks and complex viscosity.
13
Microstructure
3.4
The optical microscopic images of all three collagen samples display fibrous structures (Figure 5). Bovine telopeptide-poor collagen exhibited the most distinct fibrous network, followed by chicken skin collagen, and chicken bone collagen revealed the most inhomogeneous structure with large and very small fragments. This supports the hypothesis stated above claiming bone collagen molecules are related to a low intrinsic and high complex viscosity due to large fragments, falsifying the apparent viscosity. A fibrous network was also visible for chicken skin collagen, however, with thinner collagen fibers. Thus, a correlation between a more pronounced fibrous structure and decreasing shear thinning behavior might exist. This can be explained by the fact that the absence of crosslinks allows the collagen molecules to assemble in large flexible fibers that can rearrange reversely while deformation takes place. Thus, the chicken skin collagen, which was proposed to feature more covalent bonds, revealed the most delicate network. This was also evident in the SEM images (Figure 6). On the contrary, telopeptide-poor and chicken bone collagen demonstrated a coarser matrix with larger pores. This might be explained by the mobility of the non-crosslinked telopeptide-poor collagen molecules and the extracted bone collagen molecules to form larger structures and larger pores. Key insights
3.5
In summary, we can highlight a number of key insights obtained from this study, as illustrated in Figure 7:
(A) Telopeptide-poor collagen featured the highest dynamic power law exponent, suggesting that the least covalent crosslinks were present in this sample, thus, warranting it as a negative standard.
(B) Chicken skin collagen exhibited a fibrous microstructure and complex viscosity resembling telopeptide-poor collagen most. However, chicken skin collagen revealed thinner fibers and a lower ability to swell, as indicated by its lower intrinsic viscosity. 14
(C) Chicken bone collagen extraction was not sufficient, because large and firm fragments led to an inhomogeneous particulate suspension with both high complex viscosity and a dynamic consistency index, while the intrinsic viscosity was found to be low, suggesting poor swelling capacity.
4
CONCLUSION
This study displays a first approach to evaluate the application of various chicken collagen samples within the co-extrusion process. Conclusion can be drawn from the structure–function–process relation established in previous studies (Oechsle, Häupler, et al., 2015; Oechsle et al., 2016; Oechsle, Landenberger, et al., 2015; Oechsle et al., 2014), where a highly crosslinked native collagen and a weakly crosslinked telopeptide-poor collagen was extruded. There, gels having an insufficient viscosity due to a low entanglement of insufficiently long collagen fibers were found to not be coextrudable. Based on these results, chicken bone collagen may not be not suitable for extrusion, since the extracted fibers were short. By contrast, chicken skin yielded more entangled collagen fibers and gels having higher viscosities possibly providing the needed properties for extrusion and subsequent fixation. Moreover, film forming abilities may be further modulated by addition of salts or proteins. Thus, collagen obtained from chicken skin might serve the need for the production of pure co-extruded chicken sausages. Nevertheless, further research on chicken skin collagen is mandatory and the actual application needs to be tested to produce high-end products.
5
ACKNOWLEDGEMENTS
The authors would like to thank Gert and Marion Bueker (Protein Consulting) and Robert Wilfer and Katja Mader (Kalle GmbH) for providing the collagen samples, Barbara Maier for her assistance with the SEM imaging, and Stefan B. Irmscher (University of Hohenheim) for his support in all fields. Moreover, we acknowledge Jens Pfannstiel and the Proteomics Core Facility of the Life Science Center at the Universitiy of Hohenheim (Germany) for their support in mass spectrometry analysis. This research project was supported by the German Ministry of Economics and Energy (via AiF) and the FEI (Forschungskreis der Ernährungsindustrie e.V., Bonn): Project AiF 17478 N.
15
6
REFERENCES
Abedin, M. Z., & Riemschneider, R. (1984). Chicken skin collagen-molecular diversity and susceptibility to neutral proteinases. Pharmaceutical Industry, 46, 532-535. Bailey, A. J., & Shimokomaki, M. S. (1971). Age related changes in the reducible crosslinks of collagen. FEBS Letters, 16(2), 86-88. Barbut, S. (2010). Microstructure of natural, extruded and co-extruded collagen casings before and after heating. Italian Journal of Food Science, 22(2), 126-133. Birk, D. E., Hahn, R. A., Linsenmayer, C. Y., & Zycband, E. I. (1996). Characterization of collagen fibril segments from chicken embryo cornea, dermis and tendon. Matrix Biol, 15(2), 111-118. Bräumer, V. K. (1974). Das faserprotein kollagen. Die Angewandte Makromolekulare Chemie, 40(1), 485-492. Bueker, M., Bueker, G., & Grolig, G. (2009). Collagen concentrate, use thereof and also process for production thereof, Kalle GmbH, Patent: US20090162502 A1. BVL. (2011). Verfahren zur Probenahme und Untersuchung von Lebensmitteln. In: Amtliche Sammlung von Untersuchungsverfahren nach § 64 LFGB, § 35 vorläufiges Tabakgesetz, § 28b GenTG. Chan, V. C., Ramshaw, J. A., Kirkpatrick, A., Beck, K., & Brodsky, B. (1997). Positional preferences of ionizable residues in Gly-X-Y triplets of the collagen triple-helix. Journal of Biological Chemistry, 272(50), 31441-31446. De Gennes, P. G. (1976). Dynamics of Entangled Polymer Solutions. I. The Rouse Model. Macromolecules, 9(4), 587-593. Gelse, K., Pöschl, E., & Aigner, T. (2003). Collagens—structure, function, and biosynthesis. Advanced Drug Delivery Reviews, 55(12), 1531-1546. Glimcher, M. J., & Katz, E. P. (1965). The organization of collagen in bone: The role of noncovalent bonds in the relative insolubility of chicken bone collagen. J Ultrastruct Res, 12((5/6)), 705-729. Hiemenz, P. C., & Lodge, T. P. (2007). Dynamics of dilute polymer solutions. In P. C. Hiemenz & T. P. Lodge (Eds.), Polymer Chemistry (2 ed., pp. 335). Boca Raton: CRC Press.
16
Highberger, J. H. (1939). The Isoelectric Point of Collagen. Journal of the American Chemical Society, 61(9), 2302-2303. Huggins, M. L. (1942). The Viscosity of Dilute Solutions of Long-Chain Molecules. IV. Dependence on Concentration. Journal of the American Chemical Society, 64(11), 2716-2718. Irmscher, S. B., Böjthe, Z., Herrmann, K., Gibis, M., Kohlus, R., & Weiss, J. (2013). Influence of filling conditions on product quality and machine parameters in fermented coarse meat emulsions produced by high shear grinding and vacuum filling. Journal of Food Engineering, 117(3), 316-325. Irmscher, S. B., Rühl, S., Herrmann, K., Gibis, M., Kohlus, R., & Weiss, J. (2015). Determination of Process-Structure Relationship in the Manufacturing of Meat Batter Using Vane Pump-Grinder Systems. Food and Bioprocess Technology, 8(7), 1512-1523. Joseph, R. L. (2003). Beef. In B. Caballero (Ed.), Encyclopedia of Food Sciences and Nutrition (2 ed., pp. 412-417). Oxford: Academic Press. Kadler, K. E., Holmes, D. F., Trotter, J. A., & Chapman, J. A. (1996). Collagen fibril formation. Biochemical Journal, 316(Pt 1), 1-11. Kahn, L. D., & Witnauer, L. P. (1966). The viscometric behavior of solubilized calf skin collagen at low rates of shear. J Biol Chem, 241(8), 1784-1789. Katz, E. P., Francois, C. J., & Glimcher, M. J. (1969). The molecular weights of the alpha chains of chicken bone collagen by high speed sedimentation equilibrium. Biochemistry, 8(6), 2609-2615. Keogh, M. K., & O’Kennedy, B. T. (1998). Rheology of Stirred Yogurt as Affected by Added Milk Fat, Protein and Hydrocolloids. Journal of Food Science, 63(1), 108112. Kraemer, E. O. (1938). Molecular Weights of Celluloses and Cellulose Derivates. Industrial & Engineering Chemistry, 30(10), 1200-1203. Laemmli, U. K. (1970). Cleavage of Structural Proteins during the Assembly of the Head of Bacteriophage T4. Nature (London), 227(5259), 680-685. Lee, C. H., Singla, A., & Lee, Y. (2001). Biomedical applications of collagen. International Journal of Pharmaceutics, 221(1–2), 1-22. 17
Macosko, C. W. (1994). Rheology: Principles, Measurements, and Applications. Hoboken: WILEY-VCH. Maurer, A. J. (2003). Poultry | Chicken. In B. Caballero (Ed.), Encyclopedia of Food Sciences and Nutrition (2 ed., pp. 4680-4686). Oxford: Academic Press. Oechsle, A. M., Häupler, M., Gibis, M., Kohlus, R., & Weiss, J. (2015). Modulation of the rheological properties and microstructure of collagen by addition of co-gelling proteins. Food Hydrocolloids, 49(0), 118-126. Oechsle, A. M., Häupler, M., Weigel, F., Gibis, M., Kohlus, R., & Weiss, J. (2016). Modulation of extruded collagen films by the addition of co-gelling proteins. Journal of Food Engineering, 171, 164-173. Oechsle, A. M., Landenberger, M., Gibis, M., Irmscher, S. B., Kohlus, R., & Weiss, J. (2015). Modulation of collagen by addition of Hofmeister salts. International Journal of Biological Macromolecules, 79(0), 518-526. Oechsle, A. M., Wittmann, X., Gibis, M., Kohlus, R., & Weiss, J. (2014). Collagen entanglement influenced by the addition of acids. European Polymer Journal, 58(0), 144-156. Osburn, W. N. (2002). Collagen Casings. In A. Gennadios (Ed.), Protein-Based Films and Coatings (pp. 445-466). Boca Raton: CRC Press. Ramshaw, J. A. (1986). Distribution of type III collagen in bovine skin of various ages. Connective Tissue Research, 14(4), 307-314. Rao, M. A. (2007). Flow and Functional Models for Rheological Properties of Fluid Foods. In M. A. Rao (Ed.), Rheology of Fluid and Semisolid Foods (pp. 27-58): Springer US. Reiser, K., McCormick, R. J., & Rucker, R. B. (1992). Enzymatic and nonenzymatic cross-linking of collagen and elastin. Federation of American Societis for Experimental Biology, 6(7), 2439-2449. Sakai, R., Ikeda, S., & Isemura, T. (1967). Soluble collagen of chicken leg tendon; its denaturation temperature and hydrodynamic properties. Bulletin of the Chemical Society of Japan, 40(12), 2890-2894.
18
Sarbon, N. M., Badii, F., & Howell, N. K. (2013). Preparation and characterisation of chicken skin gelatin as an alternative to mammalian gelatin. Food Hydrocolloids, 30(1), 143-151. Strawich, E., & Glimcher, M. J. (1983). Differences in the extent and heterogeneity of lysyl hydroxylation in embryonic chick cranial and long bone collagens. J Biol Chem, 258(1), 555-562. Wu, J., Li, Z., Yuan, X., Wang, P., Liu, Y., & Wang, H. (2011). Extraction and isolation of type I, III and V collagens and their SDS-PAGE analyses. Transactions of Tianjin University, 17(2), 111-117.
19
7
FIGURE CAPTIONS
Figure 1
SDS-PAGE pattern of the molecular weight markers (bovine myosin at 245 kDa and rec. E.coli β-galactosidase at 123 kDa), bovine telopeptidepoor collagen, chicken skin collagen, and chicken bone collagen from left to the right. (The contrast was enhanced.)
Figure 2
Exemplary ζ-potential measurement of bovine telopeptide-poor collagen, chicken skin collagen, and chicken bone collagen as a function of pH (i=1, n=4).
Figure 3
Reduced viscosity ηred (black symbols) and inherent viscosity ηinh (white symbols) calculated from Equation (1) and (2), respectively, of bovine telopeptide-poor collagen, chicken skin collagen, and chicken bone collagen as a function of the collagen concentration c (i=2, n=2).
Figure 4
Complex viscosity η* calculated from Equation (3) of telopeptide-poor collagen, chicken skin collagen, and chicken bone collagen as a function of the angular frequency ω (i=2, n=2).
Figure 5
Optical microscopic images of 4 g/100 g (A) bovine telopeptide-poor collagen, (B) chicken skin collagen, and (C) chicken bone collagen (scale bar = 50 µm).
Figure 6
Scanning electron microscopy images of 4 g/100 g (A) bovine telopeptidepoor collagen, (B) chicken skin collagen, and (C) chicken bone collagen (scale bar = 50 µm).
Figure 7
Schematic illustration of chicken skin collagen, chicken bone collagen, and bovine hide split collagen treated to obtain telopeptide-poor collagen in vivo and in gel form. The intrinsic viscosity [η] and the dynamic power law factor n* decreased, while the complex viscosity η* and the dynamic consistency index k* increased following the order telopeptide-poor collagen, chicken skin collagen, and chicken bone collagen.
20
8
FIGURES
Figure 1
21
Telopeptide-poor collagen Chicken skin collagen Chicken bone collagen
-potential (mV)
30
20
10
0
-10
3
4
5
6
pH Figure 2
22
7
Telopeptide-poor collagen Chicken skin collagen Chicken bone collagen
inh
4
3
red(L/g),
(L/g)
5
2
1
0 0.0
0.1
0.2
0.3
Collagen concentration (g/L) Figure 3
23
0.4
Complex viscosity * (Pa s)
Telopeptide-poor collagen Chicken skin collagen Chicken bone collagen
10000
1000
100
10 1
10
Angular frequency Figure 4
24
100
(rad/s)
Telopeptidepoor collagen
Chicken skin collagen
Chicken bone collagen
Figure
5
25
Telopeptidepoor collagen
Chicken skin collagen
Chicken bone collagen
Figure 6
26
Figure 7
27
9
TABLES
Table 1
Chemical composition of bovine telopeptide-poor collagen, chicken skin collagen and chicken bone collagen. Telopeptide-poor
Composition
collagen (g/100 g)
Chicken skin collagen
Chicken bone collagen
(g/100 g)
(g/100 g)
Dry matter content
9.88 ± 0.40
34.26 ± 1.28
41.29 ± 1.46
Ash content
4.14 ± 0.32
3.86 ± 0.35
2.77 ± 0.08
Protein content*
6.47 ± 0.37
25.23 ± 2.59
31.25 ± 1.23
Non-protein nitrogenous content
n. d.
0.06 ± 0.01
n. d.
Hydroxyprolin content
0.90 ± 0.02
2.56 ± 0.22
4.28 ± 0.17
Connective tissue content
7.22 ± 0.13
20.89 ± 1.49
34.70 ± 0.85
Connective tissue degradation products
0.01 ± 0.01
n. d.
n. d.
Lipid content
0.08 ± 0.07
0.91 ± 0.08
0.74 ± 0.06
Not detectable (n. d.), *Factor 5.5 was applied to calculate the protein from the nitrogen content.
28
Table 2
Intrinsic viscosity [η] of bovine telopeptide-poor collagen, chicken skin collagen, and chicken bone collagen determined according to Equation (1) and (2) deduced from the y-intercept of Figure 3.
Collagen type
[η]a (L/g)
R² a
[η]b (L/g)
R² b
Telopeptide-poor collagen
3.26
0.67 n. s.
3.16
0.89 n.s.
Chicken skin collagen
0.31
0.99**
0.28
0.99*
Chicken bone collagen
0.13
0.99**
0.13
0.99**
**p <0.01, *p <0.05, not significant (n. s.) p >0.05 a
The intrinsic viscosity [η] was determined according to Huggins (1942).
b
The intrinsic viscosity [η] was determined according to Kraemer (1938).
29
Table 3
Dynamic consistency index k* and the dynamic power law factor n* calculated from Equation (4) and deduced from a double logarithmic complex viscosity η* versus angular frequency ω plot (Figure 4) of bovine telopeptide-poor collagen, chicken skin collagen, and chicken bone collagen.
Collagen type
k* (Pa s n*
)
n*
R²
Telopeptide-poor collagen
320
0.31
0.99***
Chicken skin collagen
606
0.17
0.99***
Chicken bone collagen
15146
0.11
0.99***
***p < 0.0001
30
S2213-3291(16)30002-8 http://dx.doi.org/doi:10.1016/j.foostr.2016.02.001 FOOSTR 44
To appear in: Received date: Revised date: Accepted date:
22-7-2015 22-1-2016 15-2-2016
Please cite this article as: Oechsle, Anja Maria., Akg¨un, Dila., Krause, Franziska., Maier, Christiane., Gibis, Monika., Kohlus, Reinhard., & Weiss, Jochen., Microstructure and PhysicalndashChemical Properties of Chicken Collagen.Food Structure http://dx.doi.org/10.1016/j.foostr.2016.02.001 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
MICROSTRUCTURE AND PHYSICAL–CHEMICAL PROPERTIES OF CHICKEN COLLAGEN Anja Maria Oechsle1, Dila Akgün1, Franziska Krause1, Christiane Maier1,
Monika
Gibis1, Reinhard Kohlus2, Jochen Weiss1,* 1
Department of Food Physics and Meat Science, Institute of Food Science and
Biotechnology, University of Hohenheim, Garbenstrasse 21/25, 70599 Stuttgart, Germany 2
Department of Food Process Engineering and Food Powders, Institute of Food Science
and Biotechnology, University of Hohenheim, Garbenstrasse 25, 70599 Stuttgart, Germany
Submitted to Food Structure in July, 2015 Running Title: Microstructure and Physical–Chemical Properties of Chicken Collagen To
whom correspondence should be addressed: [email protected], Tel.: +49
711 459 24415, Fax: +49 711 459 24446
1
Graphical abstract
Highlights
Chicken skin collagen resembles the fibrous bovine telopeptide-poor collagen most.
Chicken bone collagen reveals firm fragments and a low intrinsic viscosity proposing insufficient extraction.
The shear thinning behavior increases with the degree of crosslinks within the collagen.
The intrinsic viscosity is linked inversely to the collagen complex viscosity.
2
ABSTRACT The global consumption of sausages has increased immensely over the last few years, while bovine collagen has become scarce. Thus, collagen from alternative sources is being considered for application in the food industry. Therefore, chicken skin and bone collagen were characterized and compared with the bovine telopeptide-poor collagen aiming at the feasibility of producing pure, co-extruded chicken sausages. Hence, the chemical composition, microstructure and rheological properties of the different collagen samples were examined and SDS-PAGE, mass spectroscopy, and ζ-potential analysis were conducted. Weak bands in the SDS-PAGE gel indicated only partial maceration of chicken bone collagen, whereas chicken skin and telopeptide-poor collagen revealed distinct bands indicating collagen type I and III. This was also verified by mass spectroscopy. Large fragments were visible in optical microscopy for chicken bone collagen, whereas chicken skin collagen revealed a delicate network. Moreover, the highest dynamic consistency index, at 14146 Pa s n*, was determined for chicken bone collagen, followed by chicken skin and telopeptide-poor collagen at 606 and 320 Pa s n*, respectively. By contrast, the intrinsic viscosity was the highest for telopeptide-poor collagen (3.16 - 3.26 L/g), whereas chicken bone collagen exhibited the lowest value at 0.13 L/g, suggesting poor swelling behavior. Moreover, telopeptide-poor collagen featured the highest dynamic power law factor, suggesting the least crosslinks, serving a negative standard rather than an actual chicken collagen analog. Finally, chicken skin collagen displayed the most suitable source of collagen for the coextrusion process compared to the well established bovine hide split collagen. Keywords: Chicken skin collagen; chicken bone collagen; bovine telopeptide-poor collagen; intrinsic viscosity
3
1
INTRODUCTION
Collagen represents a family of proteins that serve as major structural components of connective tissues in vertebrates (Lee, Singla, & Lee, 2001). It has unique physical properties, including uniformity, tensile strength, flexibility, biocompatibility, and biodegradability. Therefore, it is used for a wide range of applications, e.g. as scaffold in tissue engineering, for implantations or wound dressing in surgical operations, as a capsule matrix material or binder in pharmaceutical applications, and for the production of gels and films in food (Osburn, 2002). A promising use of collagen in the food industry is the manufacture of coextruded sausage casings, due to the rapidly rising cost of natural, intestinal-derived casings (Barbut, 2010). Moreover, continuous processes allow the hygienic production of sausages in large quantities with consistent end product quality attributes (Irmscher et al., 2013; Irmscher et al., 2015; Osburn, 2002). Currently, bovine hide split collagen is used in the co-extrusion process (Bueker, Bueker, & Grolig, 2009). However, the global consumption of meat products, in particular sausages, is rising and supplies of bovine collagen are getting low. Moreover, the need of kosher and halal products is increasing constantly. As a result, chicken collagen could supply the need, since it might be easily gained as a by-product from skins, cartilages, bones, and feet from the poultry manufacturing industry. In addition to good availability, fibril-forming collagens are predestined for the production of edible sausage casings, as the stable network enables shrinking and stretching to accommodate the contraction and expansion of meat batter during processing (Osburn, 2002). Bovine and chicken skins predominantly contain the fibril-forming collagen types I and III (Abedin & Riemschneider, 1984; Bräumer, 1974; Gelse, Pöschl, & Aigner, 2003; Ramshaw, 1986). On a molecular basis, fibril-forming collagens feature an uninterrupted helical region with alternating polar and nonpolar domains leading to lateral alignment of molecules in a quarter staggered array (Reiser, McCormick, & Rucker, 1992). Collagen type I is a heterodimer composed of two identical α1-chains and one α2-chain (Bräumer, 1974; Gelse et al., 2003; Ramshaw, 1986), whereas collagen type III is a homotrimer of three α1(III)-chains and usually occurs in the same fibril with type I collagen (Kadler, Holmes, Trotter, & Chapman, 1996). Collagen stability and structure is based on hydrogen bonds between polar residues of 4-hydroxyprolin and 5-hydroxylysin, the formation of 4
hydration networks, and electrostatic interactions (Gelse et al., 2003). The latter ones emerge between ionizable side groups present in 15 – 20 % of all amino acid residues either in the X or in the Y position of the Gly-X-Y triplets (Chan, Ramshaw, Kirkpatrick, Beck, & Brodsky, 1997). Chicken collagen is less crosslinked due to the degree of covalent crosslinks increasing with advancing age due to lysyl oxidase-initiated crosslinks (Bailey & Shimokomaki, 1971). Thus, chickens are slaughtered after approximately 6 – 7 weeks (Maurer, 2003), while the average age of beef is 18 – 36 months (Joseph, 2003). The performance of telopeptide-poor collagen as a model system for less crosslinked collagen types, such as chicken collagen, was investigated in former studies of the author. We investigated the influence of acids, Hofmeister salts, and co-gelling proteins on native and telopeptidepoor bovine collagen rheology and microstructure (Oechsle, Häupler, Gibis, Kohlus, & Weiss, 2015; Oechsle, Landenberger, et al., 2015; Oechsle, Wittmann, Gibis, Kohlus, & Weiss, 2014). Thus, the aim of this study was to characterize chicken bone and skin collagen, which was provided by the industry partners due to their low-cost availability, and to evaluate their suitability for the extrusion process compared to a telopeptide-poor collagen. Therefore, results obtained from chemical analysis, microscopy, ζ-potential, mass spectrometry, and sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDSPAGE) analyses were compared. In this context, we hypothesized that chicken collagen behaves according to the telopeptide-poor collagen in terms of rheological and microstructural characteristics. Furthermore, we postulated that chicken skin collagen fibers stay intact, whereas the demineralization of bone leads to a break down of the collagen fibers, leading to weaker gels than telopeptide-poor collagen.
2 2.1
MATERIALS AND METHODS Materials
The raw material of the bovine telopeptide-poor collagen was kindly provided by the Kalle (Wiesbaden, Germany) and Protein Consulting (Singhofen, Germany) prepared the telopeptide-poor collagen by splitting off telopeptides and intermolecular crosslinks from 5
native collagen to obtain single collagen triple helices (Bueker et al., 2009). Chicken collagen was extracted using a procedure similar to that used for bovine collagen using chicken bone and skin as raw materials (Protein Consulting, Singhofen, Germany). For each samples one batch was provided by the industry partners that comprised the extraction product of several extraction procedures. 2.2
Chemical characterization
The standard analysis methods according to German law were used for the collagen characterization, according to § 64 Lebensmittel- und Futtermittelgesetzbuch (BVL, 2011). Dry matter content was determined according to L 06.00-3 (BVL, 2011). The ash (minerals) determination was conducted according to L 06.00-4 (BVL, 2011). The protein content in collagen samples was quantified applying the Kjeldahl method L 06.00-7, whereas a combination of two methods was used (L 07.00-41 and L 07.00-57) for nonprotein nitrogenous substances in the samples (BVL, 2011). The amount of collagen degradation products as well as hydroxyproline and connective tissue content in the collagen samples were determined according to L 06.00-8 (BVL, 2011). The lipid content was assessed by the method L 06.00-6 (BVL, 2011). 2.3
Sample preparation
The collagen samples were diluted with 0.001 M phosphoric acid to a concentration of 6 g/100 g based on their protein content and mixed with an Artisan 5KSM 150 equipped with a flex edge beater 5KFE5T (KitchenAid, St. Joseph, Minnesota, USA) for 5 min at the highest setting. Samples were subsequently adjusted to pH 3 with phosphoric acid (Th. Geyer AG, Renningen, Germany) prior to storage over night to enable the hydration of the collagen molecules. Phosphoric acid at pH 3 was found in a previous study to yield the lowest critical overlap concentration by promoting collagen interactions (Oechsle et al., 2014). Before measurement, the 6 g/100 g suspensions were diluted to concentration needed for the analysis as indicated in sections 2.4 - 2.7, homogenized for 5 min (300 rpm) using a Stomacher Circulator 400 from Seward (West Sussex, UK), and the pH was adjusted if necessary.
6
2.4
SDS-PAGE
Sodium dodecyl sulfate polyacrylamide gel electrophoresis was performed according to the method of Laemmli (1970). Solubilized collagen samples were mixed in a ratio of 1:1 (v/v) with the sample buffer containing 0.5 M tris(hydroxymethyl)aminomethane-HCl, pH 6.8 (Applichem, Darmstadt, Germany), 20 g/L SDS, 250 mL/L glycerol (VWR, Fontenay-sous-Bois, France), 50 mL/L β-mercaptoethanol (Sigma Aldrich, Munich, Germany), and 0.1 g/L Bromphenol Blue (Merck, Darmstadt, Germany). Consequently, the sample solution was denatured at 95 °C for 4 min. An aliquot of 10 µl of 1 g/L protein suspension was loaded onto Precast mini-Protean TGX gel (Bio-Rad Laboratories, Inc., Richmond, VA, USA) and subjected to electrophoresis at 200 V for 35 min using a MiniProtean Tetra System (Bio-Rad Laboratories Inc., Richmond, VA, USA). Subsequently, the gel was stained with Coomassie Brilliant Blue R-250 (Bio-Rad Laboratories Inc., Richmond, VA, USA) for 30 min, and afterwards, destained with 150 mL/L methanol and 100 mL/L acetic acid solution until the bands were clear. Roti-Mark prestained T852 Marker (Carl Roth, Karlsruhe, Germany) was applied to determine the molecular weight of the proteins. The contrast of the image was enhanced by imageJ 1.48v software (Wayne Rasbund, National Institutes of Health, Bethesda, MD, USA). 2.5
ζ-potential
The ζ-potential of 0.1 g/100 g collagen suspensions were measured between pH 3 and 7 using a particle electrophoresis instrument (Nano, ZS, Malvern Instruments, Malvern, UK) to determine the isoelectric point. The titration was conducted automatically with a multipurpose titrator (MPT-2, Malvern Instruments, UK). The pH was adjusted by 0.1 M hydrochloric acid (Carl Roth, Karlsruhe, Germany) and 0.1 M sodium hydroxide (Carl Roth, Karlsruhe, Germany). 2.6
Rheology
Experiments were conducted using a modular compact rheometer Physica MCR 502 (Anton Paar, Ostfildern, Germany) at 5 °C. Flow curves of dilute suspensions were conducted using a single gap cylinder geometry CC27 with 28.92 mm cup diameter and 26.66 mm bob diameter (Anton Paar, Ostfildern, Germany). The dynamic viscosity of the solvent ηs and 0.1 – 0.4 g/L collagen suspensions η were characterized to calculate the 7
reduced viscosity ηred and inherent viscosity ηinh in order to obtain the intrinsic viscosity [η] according to Huggins (1942) and Kraemer (1938) with Equation (1) and (2), respectively: 𝜂 − 𝜂𝑠 ) 𝑐→0 𝑐→0 𝑐 𝜂𝑠 𝜂 𝑙𝑛 (𝜂 ) 𝑠 [𝜂] = 𝑙𝑖𝑚(𝜂𝑖𝑛ℎ ) = 𝑙𝑖𝑚 ( ) 𝑐→0 𝑐→0 𝑐
(1)
[𝜂] = 𝑙𝑖𝑚(𝜂𝑟𝑒𝑑 ) = 𝑙𝑖𝑚 (
(2)
Moreover, 4 g/100 g collagen gels were analyzed by oscillating measurements using a plate-plate geometry PP25 with 25 mm diameter (Anton Paar, Ostfildern, Germany). Strain sweeps were performed at 1 Hz in order to determine the linear viscoelastic range. Consequently, frequency sweeps were performed in the linear viscoelastic range applying 1 % strain from 1 to 100 rad/s. Moreover, the complex viscosity η* was determined as a function of the angular frequency ω applying Equation (3) (Macosko, 1994): 2
2 1/2
𝐺′′ 𝐺′ 𝜂∗ = [( ) + ( ) ] 𝜔 𝜔
(3)
The power law relation of the complex viscosity η* as a function of the angular frequency ω according to Equation (4) was assessed in order to fit the dynamic consistency index k* and the dynamic power law factor n* (Keogh & O’Kennedy, 1998). 𝜂∗ = 𝑘 ∗ 𝜔 𝑛 2.7
∗
−1
(4)
Microscopy
Optical microscopy was conducted to visualize the microstructure of 4 g/100 g collagen gels. Unstained gel samples were compressed in between the microscope slide and the cover glass in order to minimize the collagen layer thickness and to visualize the collagen microstructure. Microscopy images were taken with an an Axio Scope optical microscope (A1, Carl Zeiss Microimaging GmbH, Göttingen, Germany) equipped with a digital camera (AxioCam ICc3, Jena, Germany). A magnification of 5x was used. The samples were dipped into liquid nitrogen and freeze-dried (Lyovac GT 2, Finn Aqua Santasalo-Sohlberg, Hürth, Germany and Lyovac GT 4, LH Leybold, Köln, Germany) for approximately 48 h for the scanning electron microscopy (SEM). The samples were 8
sputtered with gold and palladium (20:80) for about 8 min with a sputtering system (SCD 040, Balzers, Bingen, Germany). The images were taken with a scanning electron microscope from Zeiss (DSM 940, Zeiss, Hamburg, Germany) under vacuum at 5000 kV. A magnification of 500x was applied. 2.8
Statistical analysis
We prepared and measured the samples for the measurements at least twice using at least duplicate samples for each experiment. The number of measurements i and number of repetitions of one dilution n are indicated in the figure captions. Means, standard deviations, and the significance of determination were calculated from themeasurements using Excel (Microsoft, Redmond, WA, USA) and SigmaPlot 12.3 (Systat Software, Inc., San Jose, CA, USA). 3 3.1
RESULTS AND DISCUSSION Collagen type and composition
The chemical composition of the collagen samples is listed in Table 1. Telopeptide-poor collagen was already utilized in previous studies (Oechsle, Häupler, et al., 2015; Oechsle et al., 2014) to signify the effects on microstructure and rheology for less crosslinked collagen types, such as chicken collagen. The telopeptide-poor collagen sample used in this study, however, had a different chemical composition. In particular, the mineral content differed considerably, which might be attributed to variations in beef type, age and breeding conditions, since the extraction method was standardized according to the supplier. Due to the different drying levels of the samples extracted, chicken bone collagen revealed the highest dry matter content with 41.29 g/100 g, followed by chicken skin collagen and telopeptide-poor collagen with 34.24 g/100 g and 9.88 g/100 g, respectively. Accordingly, the chicken bone collagen served the highest connective tissue content that corresponds to the collagen concentration with 34.70 g/100 g, followed by chicken skin collagen and bovine telopeptide-poor collagen with 20.89 g/100 g and 7.22 g/100 g, respectively. We expected the protein and connective tissue content to be identical in each of the three collagen samples. The deviations might be ascribed to inaccuracies of the photometrical measurement of the sensitive connective tissue determination. Moreover, the protein concentration is calculated from the nitrogen 9
concentration using the factor 5.55 according to BVL (2011). This factor is based on the occurrence of nitrogen in proteins with higher nitrogen content such as gelatin; however this might lead to inaccuracies, since we did not determine the actual nitrogen concentration in the collagen samples by an independent method. According to the connective tissue content the samples were diluted to the concentration desired for analysis. Figure 1 depicts the SDS-PAGE gel with bovine telopeptide-poor collagen featuring two distinctive bands at approximately 123 kDa indicating α1(I)- and α2(I)-chain monomers of type I collagen. Furthermore, the band of the slightly larger α1(III)-chain was visible above, indicating collagen type III. Mass spectrometry analysis (supplementary data) verified the presence of α1(I)- , α2(I)-, and α1(III)-chains with 133, 129 and 138 kDA, respectively. Similar to telopeptide-poor collagen, chicken skin collagen delivered α1(I)and α2(I)-bands in the SDS-PAGE gel. Mass spectrometry revealed type I and III collagen, although no α1(III)-band was evident in the SDS-PAGE gel. In this way, Abedin and Riemschneider (1984) demonstrated collagen type I and III in chicken skin collagen. The absence of α1(III)-bands within the SDS-PAGE gel can be explained by the fact that α1(III)-chains might get chemically reduced during extraction and appear at the position of α1(I)-bands in SDS-PAGE (Wu et al., 2011). Chicken bone collagen revealed solely type I collagen via mass spectroscopy analysis indicated by α1(I) and α2(I)-chains, as reported by Strawich and Glimcher (1983). The SDS-PAGE gel exhibited only weak bands. However, large fragments were visible to the naked eye in the chicken bone collagen sample, indicating insufficient extraction, leading to incomplete solubilization during SDS sample preparation. Supporting our findings, Glimcher and Katz (1965) reported that very little of the collagen of demineralized bone, even in young rapidly growing animals, could be extracted in solutions normally used to extract collagen with existing tertiary structure from unmineralized bone. On the other hand, they reported that a large fraction of the chicken bone collagen was found to be extracted as gelatin, signifying the complexity of the extraction procedure. Overall, the protein sizes determined via mass spectrometry did not correspond to the protein sizes perceived by SDS-PAGE. This might be due to fibrillar collagen molecules featuring a higher electrophoretic mobility within the SDS-PAGE gel and, therefore, they appear smaller 10
than globular marker proteins, namely bovine myosin (245 kDa) and rec. E.coli βgalactosidase (123 kDa). For all three samples, β-bands and γ-bands were visible below and above 245 kDa representing dimers and trimers, respectively, as already observed by Oechsle, et al. (2015). 3.2
ζ-potential
Figure 2 demonstrates exemplarily a measurement of the ζ–potential as a function of the pH. Bovine telopeptide-poor collagen displayed an isoelectric point at pH 4.62 ± 0.07, followed by chicken bone collagen and chicken skin collagen at pH 5.09 ± 0.09 and 6.54 ± 0.32, respectively. The expected isoelectric point of bovine collagen was reported to be located at pH 4.7, which matched the data collected for bovine telopeptide-poor collagen (Highberger, 1939). However, both chicken collagen samples revealed higher isoelectric points, which might be ascribed to different animal species, leading to alterations in the amino acid side groups and net charge, since the extraction was based on the patent established for bovine collagen by Bueker, Bueker, and Grolig (2009). Moreover, only the ζ-potentials of the soluble fractions of the chicken bone collagen sample might have been detected due to its inhomogeneity. Consequently, the results should be considered with caution, since the solubilized collagen fraction might discriminate vigorously from the insoluble collagen bound to the bone matrix. Nonetheless, all three collagen samples exhibited a positive net charge at pH 3, signifying that all gels feature repulsive forces preventing collagen from precipitation and facilitating hydration. 3.3
Rheology
The reduced viscosity ηred and inherent viscosity ηinh are displayed in Figure 3 as a function of the collagen concentration with the y-intersect being the intrinsic viscosity [η] listed in Table 2. The intrinsic viscosity [η] is a measure of a polymer’s ability to increase the viscosity of a solvent (De Gennes, 1976). Telopeptide-poor collagen delivered the highest [η] with 3.16 – 3.26 L/g. Kahn and Witnauer (1966) determined an intrinsic viscosity of 2.7 L /g for calf collagen diluted in citrate buffer at pH 3.5. This is in rough accordance with the telopeptide-poor collagen. The slight variation might be due to different conditions such as pH and buffer, which also affect the conformation and the 11
interaction with the bulk media. By contrast, the chicken skin and bone collagen samples revealed 10- and 30-fold lower values of 0.28 – 0.31 L/g and 0.13 L/g, respectively. Hence, the telopeptide-poor collagen molecules might either exhibit the largest hydrodynamic volume or the lowest molecular weight in order to get a lower intrinsic viscosity, since these parameters are inversely correlated in Equation (5) (Hiemenz & Lodge, 2007). [𝜂] =
5 𝑁𝐴𝑉 𝑉ℎ 2𝑀
(5),
with the Avogadro constant NAV, the hydrodynamic volume Vh and the molecular weight M. As mentioned before, extraction of the chicken bone collagen might have been too faint, leading to an inhomogeneous sample with larger fragments, which, in turn, might not be able to swell and, therefore, not be able to affect the bulk water. As a consequence, unswollen collagen structures do not form a homogeneous gel and are therefore unable to yield homogenous thin films after extrusion. Supporting the hypothesis stated above, Glimcher and Katz (1965) reported that chicken bone collagen is not able to swell in mild acidic conditions. However, they also postulated that very few intra- or intermolecular covalent crosslinks are present in chicken bone collagen, signifying strong noncovalent intermolecular forces playing the major role in its relative insolubility. Furthermore, the solubilized chicken bone collagen molecules might feature a smaller molecular weight that also contributes to a lower intrinsic viscosity (Katz, Francois, & Glimcher, 1969). Sakai, Ikeda, and Isemura (1967) determined a [η] of 1.25 g/L for soluble chicken leg tendon collagen which is higher than the [η] of both chicken collagen samples. The divergence might be ascribed to the fact that in addition to type I collagen, chicken tendon also contains type V and III collagen (Birk, Hahn, Linsenmayer, & Zycband, 1996), resulting in a higher swelling capacity, which, in turn, influences the intrinsic viscosity. Moreover, Sarbon, Badii, and Howell (2013) stated an intrinsic viscosity for chicken skin gelatin of 0.15 L/g, which is in agreement with the chicken bone collagen results, implicating that the bone collagen sample might be degraded. Supporting this hypothesis, Glimcher and Katz (1965) stated that 80 % of the collagen from chicken bone was extracted as gelatin featuring a smaller hydrodynamic volume, due to the random coil conformation, rather than the tertiary triple helical fiber structure of collagen. 12
Figure 4 demonstrates the complex viscosity η* that was measured in the linear viscoelastic region regarding the angular frequency of the collagen samples. Hereby, chicken bone collagen exhibited the highest η*, followed by chicken skin and telopeptide-poor collagen. The complex viscosity comprises a real and imaginary component that is associated with the storage and loss modulus, respectively, suggesting that the chicken bone collagen features the highest moduli. However, the results need to be considered carefully, since SDS-PAGE analysis and the intrinsic viscosity implemented collagen molecules being extracted insufficiently. Hence, large fragments might have misled the rheological analysis by increasing the torque. Therefore, η* appeared to be higher than the suspension actually was. Meeting expectations, the lowest η* was determined for telopeptide-poor collagen, which might be ascribed to the low molecular weight based on the absence of intermolecular crosslinks and telopeptide ends. This phenomenon was also observed in a previous study where the highly crosslinked native collagen delivered a lower [η] and a higher storage modulus than telopeptide-poor collagen (Oechsle et al., 2014). The dynamic consistency index k* and dynamic power law factor n*, as listed in Table 3, were deduced from the power law fit of Figure 4. As for the complex viscosity, the dynamic consistency index k* describes the consistency and was the highest for the chicken bone collagen, followed by chicken skin and telopeptide-poor collagen with 15146, 606, and 320 Pa s
n*
, respectively. The dynamic power law factor n* indicates a
shear-thickening (n* > 1), Newtonian (n* = 1) or shear-thinning (n* < 1) behavior (Rao, 2007). Chicken bone collagen revealed the highest shear-thinning behavior with n* = 0.11, followed by chicken skin collagen with n* = 0.13 and telopeptide-poor collagen with n* = 0.31. In a previous study (unpublished data), the highly crosslinked native collagen displayed a dynamic power law factor n* of 0.14. Consequently, a connection between increasing covalent crosslinks and decreasing shear-thinning behavior might exist for the collagen samples analyzed. As a result, chicken skin and bone collagen might contain more intra- and intermolecular crosslinks than telopeptide-poor collagen. Thus, telopeptide-poor collagen is actual a negative standard with the least crosslinks and complex viscosity.
13
Microstructure
3.4
The optical microscopic images of all three collagen samples display fibrous structures (Figure 5). Bovine telopeptide-poor collagen exhibited the most distinct fibrous network, followed by chicken skin collagen, and chicken bone collagen revealed the most inhomogeneous structure with large and very small fragments. This supports the hypothesis stated above claiming bone collagen molecules are related to a low intrinsic and high complex viscosity due to large fragments, falsifying the apparent viscosity. A fibrous network was also visible for chicken skin collagen, however, with thinner collagen fibers. Thus, a correlation between a more pronounced fibrous structure and decreasing shear thinning behavior might exist. This can be explained by the fact that the absence of crosslinks allows the collagen molecules to assemble in large flexible fibers that can rearrange reversely while deformation takes place. Thus, the chicken skin collagen, which was proposed to feature more covalent bonds, revealed the most delicate network. This was also evident in the SEM images (Figure 6). On the contrary, telopeptide-poor and chicken bone collagen demonstrated a coarser matrix with larger pores. This might be explained by the mobility of the non-crosslinked telopeptide-poor collagen molecules and the extracted bone collagen molecules to form larger structures and larger pores. Key insights
3.5
In summary, we can highlight a number of key insights obtained from this study, as illustrated in Figure 7:
(A) Telopeptide-poor collagen featured the highest dynamic power law exponent, suggesting that the least covalent crosslinks were present in this sample, thus, warranting it as a negative standard.
(B) Chicken skin collagen exhibited a fibrous microstructure and complex viscosity resembling telopeptide-poor collagen most. However, chicken skin collagen revealed thinner fibers and a lower ability to swell, as indicated by its lower intrinsic viscosity. 14
(C) Chicken bone collagen extraction was not sufficient, because large and firm fragments led to an inhomogeneous particulate suspension with both high complex viscosity and a dynamic consistency index, while the intrinsic viscosity was found to be low, suggesting poor swelling capacity.
4
CONCLUSION
This study displays a first approach to evaluate the application of various chicken collagen samples within the co-extrusion process. Conclusion can be drawn from the structure–function–process relation established in previous studies (Oechsle, Häupler, et al., 2015; Oechsle et al., 2016; Oechsle, Landenberger, et al., 2015; Oechsle et al., 2014), where a highly crosslinked native collagen and a weakly crosslinked telopeptide-poor collagen was extruded. There, gels having an insufficient viscosity due to a low entanglement of insufficiently long collagen fibers were found to not be coextrudable. Based on these results, chicken bone collagen may not be not suitable for extrusion, since the extracted fibers were short. By contrast, chicken skin yielded more entangled collagen fibers and gels having higher viscosities possibly providing the needed properties for extrusion and subsequent fixation. Moreover, film forming abilities may be further modulated by addition of salts or proteins. Thus, collagen obtained from chicken skin might serve the need for the production of pure co-extruded chicken sausages. Nevertheless, further research on chicken skin collagen is mandatory and the actual application needs to be tested to produce high-end products.
5
ACKNOWLEDGEMENTS
The authors would like to thank Gert and Marion Bueker (Protein Consulting) and Robert Wilfer and Katja Mader (Kalle GmbH) for providing the collagen samples, Barbara Maier for her assistance with the SEM imaging, and Stefan B. Irmscher (University of Hohenheim) for his support in all fields. Moreover, we acknowledge Jens Pfannstiel and the Proteomics Core Facility of the Life Science Center at the Universitiy of Hohenheim (Germany) for their support in mass spectrometry analysis. This research project was supported by the German Ministry of Economics and Energy (via AiF) and the FEI (Forschungskreis der Ernährungsindustrie e.V., Bonn): Project AiF 17478 N.
15
6
REFERENCES
Abedin, M. Z., & Riemschneider, R. (1984). Chicken skin collagen-molecular diversity and susceptibility to neutral proteinases. Pharmaceutical Industry, 46, 532-535. Bailey, A. J., & Shimokomaki, M. S. (1971). Age related changes in the reducible crosslinks of collagen. FEBS Letters, 16(2), 86-88. Barbut, S. (2010). Microstructure of natural, extruded and co-extruded collagen casings before and after heating. Italian Journal of Food Science, 22(2), 126-133. Birk, D. E., Hahn, R. A., Linsenmayer, C. Y., & Zycband, E. I. (1996). Characterization of collagen fibril segments from chicken embryo cornea, dermis and tendon. Matrix Biol, 15(2), 111-118. Bräumer, V. K. (1974). Das faserprotein kollagen. Die Angewandte Makromolekulare Chemie, 40(1), 485-492. Bueker, M., Bueker, G., & Grolig, G. (2009). Collagen concentrate, use thereof and also process for production thereof, Kalle GmbH, Patent: US20090162502 A1. BVL. (2011). Verfahren zur Probenahme und Untersuchung von Lebensmitteln. In: Amtliche Sammlung von Untersuchungsverfahren nach § 64 LFGB, § 35 vorläufiges Tabakgesetz, § 28b GenTG. Chan, V. C., Ramshaw, J. A., Kirkpatrick, A., Beck, K., & Brodsky, B. (1997). Positional preferences of ionizable residues in Gly-X-Y triplets of the collagen triple-helix. Journal of Biological Chemistry, 272(50), 31441-31446. De Gennes, P. G. (1976). Dynamics of Entangled Polymer Solutions. I. The Rouse Model. Macromolecules, 9(4), 587-593. Gelse, K., Pöschl, E., & Aigner, T. (2003). Collagens—structure, function, and biosynthesis. Advanced Drug Delivery Reviews, 55(12), 1531-1546. Glimcher, M. J., & Katz, E. P. (1965). The organization of collagen in bone: The role of noncovalent bonds in the relative insolubility of chicken bone collagen. J Ultrastruct Res, 12((5/6)), 705-729. Hiemenz, P. C., & Lodge, T. P. (2007). Dynamics of dilute polymer solutions. In P. C. Hiemenz & T. P. Lodge (Eds.), Polymer Chemistry (2 ed., pp. 335). Boca Raton: CRC Press.
16
Highberger, J. H. (1939). The Isoelectric Point of Collagen. Journal of the American Chemical Society, 61(9), 2302-2303. Huggins, M. L. (1942). The Viscosity of Dilute Solutions of Long-Chain Molecules. IV. Dependence on Concentration. Journal of the American Chemical Society, 64(11), 2716-2718. Irmscher, S. B., Böjthe, Z., Herrmann, K., Gibis, M., Kohlus, R., & Weiss, J. (2013). Influence of filling conditions on product quality and machine parameters in fermented coarse meat emulsions produced by high shear grinding and vacuum filling. Journal of Food Engineering, 117(3), 316-325. Irmscher, S. B., Rühl, S., Herrmann, K., Gibis, M., Kohlus, R., & Weiss, J. (2015). Determination of Process-Structure Relationship in the Manufacturing of Meat Batter Using Vane Pump-Grinder Systems. Food and Bioprocess Technology, 8(7), 1512-1523. Joseph, R. L. (2003). Beef. In B. Caballero (Ed.), Encyclopedia of Food Sciences and Nutrition (2 ed., pp. 412-417). Oxford: Academic Press. Kadler, K. E., Holmes, D. F., Trotter, J. A., & Chapman, J. A. (1996). Collagen fibril formation. Biochemical Journal, 316(Pt 1), 1-11. Kahn, L. D., & Witnauer, L. P. (1966). The viscometric behavior of solubilized calf skin collagen at low rates of shear. J Biol Chem, 241(8), 1784-1789. Katz, E. P., Francois, C. J., & Glimcher, M. J. (1969). The molecular weights of the alpha chains of chicken bone collagen by high speed sedimentation equilibrium. Biochemistry, 8(6), 2609-2615. Keogh, M. K., & O’Kennedy, B. T. (1998). Rheology of Stirred Yogurt as Affected by Added Milk Fat, Protein and Hydrocolloids. Journal of Food Science, 63(1), 108112. Kraemer, E. O. (1938). Molecular Weights of Celluloses and Cellulose Derivates. Industrial & Engineering Chemistry, 30(10), 1200-1203. Laemmli, U. K. (1970). Cleavage of Structural Proteins during the Assembly of the Head of Bacteriophage T4. Nature (London), 227(5259), 680-685. Lee, C. H., Singla, A., & Lee, Y. (2001). Biomedical applications of collagen. International Journal of Pharmaceutics, 221(1–2), 1-22. 17
Macosko, C. W. (1994). Rheology: Principles, Measurements, and Applications. Hoboken: WILEY-VCH. Maurer, A. J. (2003). Poultry | Chicken. In B. Caballero (Ed.), Encyclopedia of Food Sciences and Nutrition (2 ed., pp. 4680-4686). Oxford: Academic Press. Oechsle, A. M., Häupler, M., Gibis, M., Kohlus, R., & Weiss, J. (2015). Modulation of the rheological properties and microstructure of collagen by addition of co-gelling proteins. Food Hydrocolloids, 49(0), 118-126. Oechsle, A. M., Häupler, M., Weigel, F., Gibis, M., Kohlus, R., & Weiss, J. (2016). Modulation of extruded collagen films by the addition of co-gelling proteins. Journal of Food Engineering, 171, 164-173. Oechsle, A. M., Landenberger, M., Gibis, M., Irmscher, S. B., Kohlus, R., & Weiss, J. (2015). Modulation of collagen by addition of Hofmeister salts. International Journal of Biological Macromolecules, 79(0), 518-526. Oechsle, A. M., Wittmann, X., Gibis, M., Kohlus, R., & Weiss, J. (2014). Collagen entanglement influenced by the addition of acids. European Polymer Journal, 58(0), 144-156. Osburn, W. N. (2002). Collagen Casings. In A. Gennadios (Ed.), Protein-Based Films and Coatings (pp. 445-466). Boca Raton: CRC Press. Ramshaw, J. A. (1986). Distribution of type III collagen in bovine skin of various ages. Connective Tissue Research, 14(4), 307-314. Rao, M. A. (2007). Flow and Functional Models for Rheological Properties of Fluid Foods. In M. A. Rao (Ed.), Rheology of Fluid and Semisolid Foods (pp. 27-58): Springer US. Reiser, K., McCormick, R. J., & Rucker, R. B. (1992). Enzymatic and nonenzymatic cross-linking of collagen and elastin. Federation of American Societis for Experimental Biology, 6(7), 2439-2449. Sakai, R., Ikeda, S., & Isemura, T. (1967). Soluble collagen of chicken leg tendon; its denaturation temperature and hydrodynamic properties. Bulletin of the Chemical Society of Japan, 40(12), 2890-2894.
18
Sarbon, N. M., Badii, F., & Howell, N. K. (2013). Preparation and characterisation of chicken skin gelatin as an alternative to mammalian gelatin. Food Hydrocolloids, 30(1), 143-151. Strawich, E., & Glimcher, M. J. (1983). Differences in the extent and heterogeneity of lysyl hydroxylation in embryonic chick cranial and long bone collagens. J Biol Chem, 258(1), 555-562. Wu, J., Li, Z., Yuan, X., Wang, P., Liu, Y., & Wang, H. (2011). Extraction and isolation of type I, III and V collagens and their SDS-PAGE analyses. Transactions of Tianjin University, 17(2), 111-117.
19
7
FIGURE CAPTIONS
Figure 1
SDS-PAGE pattern of the molecular weight markers (bovine myosin at 245 kDa and rec. E.coli β-galactosidase at 123 kDa), bovine telopeptidepoor collagen, chicken skin collagen, and chicken bone collagen from left to the right. (The contrast was enhanced.)
Figure 2
Exemplary ζ-potential measurement of bovine telopeptide-poor collagen, chicken skin collagen, and chicken bone collagen as a function of pH (i=1, n=4).
Figure 3
Reduced viscosity ηred (black symbols) and inherent viscosity ηinh (white symbols) calculated from Equation (1) and (2), respectively, of bovine telopeptide-poor collagen, chicken skin collagen, and chicken bone collagen as a function of the collagen concentration c (i=2, n=2).
Figure 4
Complex viscosity η* calculated from Equation (3) of telopeptide-poor collagen, chicken skin collagen, and chicken bone collagen as a function of the angular frequency ω (i=2, n=2).
Figure 5
Optical microscopic images of 4 g/100 g (A) bovine telopeptide-poor collagen, (B) chicken skin collagen, and (C) chicken bone collagen (scale bar = 50 µm).
Figure 6
Scanning electron microscopy images of 4 g/100 g (A) bovine telopeptidepoor collagen, (B) chicken skin collagen, and (C) chicken bone collagen (scale bar = 50 µm).
Figure 7
Schematic illustration of chicken skin collagen, chicken bone collagen, and bovine hide split collagen treated to obtain telopeptide-poor collagen in vivo and in gel form. The intrinsic viscosity [η] and the dynamic power law factor n* decreased, while the complex viscosity η* and the dynamic consistency index k* increased following the order telopeptide-poor collagen, chicken skin collagen, and chicken bone collagen.
20
8
FIGURES
Figure 1
21
Telopeptide-poor collagen Chicken skin collagen Chicken bone collagen
-potential (mV)
30
20
10
0
-10
3
4
5
6
pH Figure 2
22
7
Telopeptide-poor collagen Chicken skin collagen Chicken bone collagen
inh
4
3
red(L/g),
(L/g)
5
2
1
0 0.0
0.1
0.2
0.3
Collagen concentration (g/L) Figure 3
23
0.4
Complex viscosity * (Pa s)
Telopeptide-poor collagen Chicken skin collagen Chicken bone collagen
10000
1000
100
10 1
10
Angular frequency Figure 4
24
100
(rad/s)
Telopeptidepoor collagen
Chicken skin collagen
Chicken bone collagen
Figure
5
25
Telopeptidepoor collagen
Chicken skin collagen
Chicken bone collagen
Figure 6
26
Figure 7
27
9
TABLES
Table 1
Chemical composition of bovine telopeptide-poor collagen, chicken skin collagen and chicken bone collagen. Telopeptide-poor
Composition
collagen (g/100 g)
Chicken skin collagen
Chicken bone collagen
(g/100 g)
(g/100 g)
Dry matter content
9.88 ± 0.40
34.26 ± 1.28
41.29 ± 1.46
Ash content
4.14 ± 0.32
3.86 ± 0.35
2.77 ± 0.08
Protein content*
6.47 ± 0.37
25.23 ± 2.59
31.25 ± 1.23
Non-protein nitrogenous content
n. d.
0.06 ± 0.01
n. d.
Hydroxyprolin content
0.90 ± 0.02
2.56 ± 0.22
4.28 ± 0.17
Connective tissue content
7.22 ± 0.13
20.89 ± 1.49
34.70 ± 0.85
Connective tissue degradation products
0.01 ± 0.01
n. d.
n. d.
Lipid content
0.08 ± 0.07
0.91 ± 0.08
0.74 ± 0.06
Not detectable (n. d.), *Factor 5.5 was applied to calculate the protein from the nitrogen content.
28
Table 2
Intrinsic viscosity [η] of bovine telopeptide-poor collagen, chicken skin collagen, and chicken bone collagen determined according to Equation (1) and (2) deduced from the y-intercept of Figure 3.
Collagen type
[η]a (L/g)
R² a
[η]b (L/g)
R² b
Telopeptide-poor collagen
3.26
0.67 n. s.
3.16
0.89 n.s.
Chicken skin collagen
0.31
0.99**
0.28
0.99*
Chicken bone collagen
0.13
0.99**
0.13
0.99**
**p <0.01, *p <0.05, not significant (n. s.) p >0.05 a
The intrinsic viscosity [η] was determined according to Huggins (1942).
b
The intrinsic viscosity [η] was determined according to Kraemer (1938).
29
Table 3
Dynamic consistency index k* and the dynamic power law factor n* calculated from Equation (4) and deduced from a double logarithmic complex viscosity η* versus angular frequency ω plot (Figure 4) of bovine telopeptide-poor collagen, chicken skin collagen, and chicken bone collagen.
Collagen type
k* (Pa s n*
)
n*
R²
Telopeptide-poor collagen
320
0.31
0.99***
Chicken skin collagen
606
0.17
0.99***
Chicken bone collagen
15146
0.11
0.99***
***p < 0.0001
30