Impact of particle size reduction and carbohydrate hydrolyzing enzyme treatment on protein recovery from rapeseed (Brassica rapa L.) press cake Katariina Rommi,* Ulla Holopainen, Sari Pohjola, Terhi K. Hakala, Raija Lantto, Kaisa Poutanen, Emilia Nordlund VTT Technical Research Centre of Finland Ltd. P.O. Box 1000, FI-02044 VTT, Finland *Corresponding author; phone: +358 20 722 4514; fax: +358 20 722 7001; email:
[email protected]
1
1
Abstract
2
The aim was to assess how particle size reduction and carbohydrate hydrolyzing
3
enzyme treatment influence protein recovery from rapeseed cold-press cake, and to
4
determine the effect of these pretreatments in protein extraction procedures varying in
5
ionic strength, pH and total solid content. Defatted press cake (median particle size 600
6
µm) was milled to 21-164 µm and 7 µm median particle sizes by pin disc milling and
7
air-flow milling, respectively. The milled press cake samples were treated with a
8
carbohydrate hydrolyzing enzyme preparation, after which proteins were extracted in
9
saline (pH 6) or alkaline (pH 12) buffer at 5% solid content, or in water at 20% solid
10
content. Particle size reduction of the press cake did not influence enzyme action or
11
protein yield, suggesting that protein release from the press cake is not physically
12
limited by cell walls or internal cell structures. As an exception, protein release from the
13
aleuronic cells appeared to be hindered by intact cell walls. Enzyme treatment improved
14
protein recovery, more substantially when the extraction was carried out in water at
15
20% solid content than in saline or alkaline conditions at 5% solid content. The enzyme
16
mediated its positive effect most probably by reducing the water holding capacity of the
17
press cake, thereby facilitating solid-liquid separation, and releasing anionic compounds
18
which improved protein solubility through electrostatic stabilization. The results suggest
19
that carbohydrate hydrolyzing enzymes are beneficial for rapeseed protein extraction at
20
reduced water content or when no salt or alkali is added to increase protein solubility.
21
Keywords: milling, pectinase, microscopy, differential scanning calorimetry
2
22
Introduction
23
Rapeseed cold-press cake, the coproduct of food oil production, has gained growing
24
interest in the past decade as a potential plant-based protein source. In addition to the
25
current use as animal feed, production of food protein ingredients from the press cake
26
could generate added value for the manufacturing industry. Rapeseed protein has
27
favorable features for human use, including well-balanced amino acid composition and
28
suitable technological properties (Bell and Keith 1991; Salleh et al. 2002). The direct
29
use of rapeseed press cake as food is, however, hindered by its high carbohydrate
30
content and presence of secondary plant metabolites such as phytates and phenolics
31
having negative effects on the digestibility and sensory properties. For partial removal
32
of the interfering compounds, proteins are commonly extracted from the press cake
33
using alkali or salt solutions (Bérot et al. 2005; Ghodsvali et al. 2005) and precipitated
34
by decreasing the pH or ionic strength of the extracts. Furthermore, the use of proteases
35
and/or cell wall degrading enzymes has been recognized to facilitate the extraction
36
processes (Niu et al. 2012; Rommi et al. 2014; Sari et al. 2013; Tang 2010).
37
Consideration of the technological and economic feasibility of the production processes
38
is a prerequisite for successful commercialization of rapeseed protein production.
39
Although 70-90% protein yields have been obtained by water-intensive alkaline
40
extraction procedures (Ghodsvali et al. 2005; Zhou et al. 1990), water-lean processes
41
are suggested to be favorable for sustainable protein production (Von Der Haar et al.
42
2014). New possibilities for more feasible protein extraction processes could be opened
43
by increased understanding of the influence of cell wall structures and protein-
44
carbohydrate interactions on protein release from the press cake. Most of rapeseed
45
protein is stored in the embryo (i.e. cotyledons and embryonic axis) cells which are
3
46
surrounded by cell walls comprising of a cellulose, hemicellulose and pectin network
47
(Pustjens et al. 2013). Since cold pressing only partially disrupts the embryonic cell
48
walls, integrity of the tissue and cell wall structures in the press cake is expected to
49
influence protein release (Rommi et al. 2014).
50
Most of the published rapeseed protein extraction studies have focused on improving
51
protein solubility as a function of pH and ionic strength, whereas the physical barriers of
52
protein release have received little attention. Particle size reduction by milling has been
53
reported to clearly enhance protein extraction from soybean flour (Rosenthal et al. 1998,
54
2001; Russin et al. 2007), but to our knowledge, respective studies on rapeseed press
55
cake have not been published. Milling is also known to benefit enzyme-aided
56
fractionation processes by increasing the accessible surface area of the substrate to the
57
enzyme and thus improving the hydrolysis efficiency, as has been demonstrated for
58
several lignocellulosic feedstocks such as brewer’s spent grain (Niemi et al. 2012), corn
59
grain (Blasel et al. 2006) and wheat straw (Silva et al. 2012). Furthermore, particle size
60
reduction has been reported to decrease the water holding capacity and viscosity of
61
biomass slurries, with positive impacts on processing especially at reduced water
62
content (Santala et al. 2013; Viamajala et al. 2009).
63
The aim of the present study was to obtain insight on the factors affecting protein
64
recovery from rapeseed press cake, with focus on determining the effect of particle size
65
reduction and carbohydrate hydrolyzing enzyme treatment on aqueous protein
66
extraction. These factors where evaluated in water-intensive saline or alkaline
67
conditions where the protein solubility and extract recovery was expected to be high, as
68
well as during water extraction at 20% solid content where the protein solubility and
69
extract separation efficiency was expected to be limited. 4
70
Materials and Methods
71
Materials and enzymes
72
Rapeseed press cake was obtained from Kankaisten Öljykasvit Oy (Turenki, Finland).
73
Oil was cold-pressed from turnip rape (Brassica rapa L.) seeds at 50–60°C and the
74
press cake was pelletized. A commercial carbohydrate hydrolase preparation, Pectinex
75
Ultra SP-L, was obtained from Novozymes A/S (Bagsvaerd, Denmark). The preparation
76
contained 2876 nkat of polygalacturonase, 135 nkat of endo- -glucanase, 22 nkat of
77
endoglucanase and 7 nkat endo-1,4- -xylanase activity per total protein (61 mg/mL) in
78
the enzyme preparation (Rommi et al. 2014).
79
Defatting and particle size reduction
80
The pelletized rapeseed press cake was ground at 2000 rpm using an SM 300 cutting
81
mill (Retsch GmbH, Düsseldorf, Germany) equipped with a 8 mm square hole bottom
82
sieve, and defatted by supercritical carbon dioxide (SC-CO2) extraction at 42°C
83
according to (Rommi et al. 2015). The defatted press cake (with 35.5% protein content
84
on dry matter basis) was further milled in a pin disc mill (100 UPZ-II Fine Impact Mill,
85
Hosokawa Alpine Ag, Ausburg, Germany) or in an air-flow mill (MKCL8-15J DAU,
86
Masuko Sangyo, Kawaguchi, Japan) which was set to 3 min residence time. Particle
87
size distribution in the press cake samples after grinding and milling was determined by
88
laser light (750 nm) diffraction using an LS320 particle size analyzer (Beckman Coulter
89
Inc., Indianapolis, IN, USA) and Beckman Coulter LS software, version 3.2. The
90
ground sample was sieved in a shaker through 2 mm holes to exclude particles over 2
91
mm (representing 2% of the total sample) which exceeded the maximum detection limit
92
of the analyzer. All samples were dispersed and run in ethanol, and before analysis
93
treated with ultrasound for 5 min to prevent aggregate formation. Duplicate 5
94
measurements were performed for each sample. The median particle size of the ground,
95
SC-CO2-defatted press cake was 600 µm, whereas further pin disc milling at 8900 rpm
96
produced coarse flour with a median (D50) particle size of 164 µm. Fine flour with a
97
median particle size of 21 µm was obtained by passing the press cake twice through the
98
pin disc mill at 17800 rpm. Ultrafine flour with a median particle size of 7 µm was
99
produced by dry milling at 7600 rpm in an air-flow mill (Table 1).
100
Enzymatic hydrolysis
101
The ground (600 µm) and milled (7-164 µm) press cake samples (Table 1) were treated
102
with Pectinex Ultra SP-L using the procedure described by Rommi et al. 2015 with
103
justified modifications. Pectinex was dosed at 10 mg total protein / g dry press cake and
104
the treatment was continued for 4 h. The enzyme dose and treatment time were adjusted
105
based on the study of Rommi et al. (2015) in which hydrolysis was found to be feasible
106
at 40% solid content, but the selected enzyme dose (5 mg/g) and hydrolysis time (1 h)
107
were suggested to be insufficient. The samples were mixed with MQ water and enzyme
108
to 40% (w/w) total solid content and placed in a closed 800 mL stirred tank reactor
109
(Protoshop, Espoo, Finland) and incubated at 40°C with 40 rpm mixing. After
110
hydrolysis the slurries were immediately frozen, freeze-dried and gently ground in a
111
mortar. All enzyme treatments were performed in duplicate.
112
Protein extraction
113
Proteins were recovered from the enzyme-treated and non-enzymatically treated press
114
cake samples by a two-step saline extraction into 0.2 M sodium phosphate buffer, pH 6.
115
The samples were mixed with buffer to 5% (w/w) total solid content and agitated
116
horizontally in a shaking incubator at 230 rpm, 25 °C for 30 min. Extracts were
117
separated by centrifugation. The extraction was repeated for the residual solids and the 6
118
first and second extracts were combined and weighed. From the fine (21 µm) and
119
ultrafine samples (7 µm) after enzyme treatment or without enzyme treatment, protein
120
was additionally recovered by a two-step alkaline extraction into 0.2 M sodium
121
phosphate buffer, pH 12. In addition, the ground (600 µm) and fine (21 µm) press cake
122
samples were subjected to a two-step water extraction at 20% total solid content. The
123
extraction pH was defined by the natural pH of the substrate after enzyme treatment (pH
124
5.5) or without enzyme treatment (pH 6). The extraction was performed twice at 25°C
125
in MQ water for 30 min, and extracts were obtained by centrifugation. All extractions
126
were performed in duplicate for each of the two parallel enzyme-treated samples (N =
127
4) and in triplicate or quadruple for each non-enzymatically treated sample (N = 3). The
128
yield of solubilized protein was determined by multiplying the protein concentration of
129
extracts by the total liquid volume during extraction and dividing by the amount of total
130
protein. Respectively, the yield of recovered protein was determined by multiplying the
131
protein concentration of extracts by the recovered extract volume and dividing by the
132
amount of total protein. Protein yields from the water extraction at 20% solid content
133
were differentiated for the first extraction step (1-step extraction) and for the first and
134
second extraction steps (2-step extraction).
135
Analyses
136
Total nitrogen content of the defatted press cake and protein extracts was analyzed by
137
Kjeldahl according to the method by (Kane 1986). Protein concentration was calculated
138
from the total nitrogen content using a conversion factor of 6.25. The press cake
139
samples were prepared for microscopy according to (Holopainen-Mantila et al. 2013),
140
and the sections were stained with Acid Fuchsin and Calcofluor White as described by
141
(Dornez et al. 2011). The samples were examined and micrographs were obtained
7
142
according to (Holopainen-Mantila et al. 2013). Images taken from replicate sample
143
blocks were examined and representative images were selected for publication.
144
Denaturation enthalpy of proteins in the press cake samples was measured by
145
differential scanning calorimetry (DSC) (Mettler Toledo DSC820, Greifensee,
146
Switzerland).
147
hydrated with water to 75% moisture content, sealed and equilibrated for 1.5 h before
148
heating from 10°C to 150°C at a temperature increase rate of 10°C/min. The
149
measurements were performed in duplicate.
150
Statistical analysis
151
Statistical analysis of the protein solubilization and recovery yields from 3-4 replicate
152
experiments was carried out by general linear model multivariate analysis using SPSS
153
Statistics software (version 22, IBM, Armonk, NY). Level of significance was set at p