Improving Ruminal Degradibility Of Oil Palm Fronds Using Enzyme Extracts From White Rot Fungi CURRENT

Background: Oil palm fronds (OPF) when pretreated with white rot fungi (WRF) shows increased rumen degradability but with significant biomass loss. Thus, effects of pre-treated OPF with enzyme extracts from WRF on rumen degradability were studied in vitro. The enzyme extracts were prepared by inoculating OPF with three WRF, i.e Ceriporiopsis subvermispora, Lentinula edodes and Ganoderma lucidum, for 15, 30 and 45 days with either ammonium sulphate, (NH4)2SO4 and sodium nitrate, NaNO3 added to the culture media for each inoculation period. After preparation of enzyme extracts, the enzyme activities were determined. OPF was then pre-treated with enzyme extracts in a citrate buffer (pH 5.0) in a forced air oven at 40 oC during 5 days. Further, the in vitro rumen degradation of OPF pre-treated with enzyme extracts, with respect to the short chain fatty acid (SCFA) production, was determined after 24 h incubation. Activity of lignolytic (laccase and MnP), cellulolytic (CMCase and avicelase) and hemicellulolytic (xylanase) enzymes were measured in all of the extracts irrespective on the inoculation period. Results: Treatment of OPF with enzyme extracts from G. lucidum after 45 days of inoculation showed a numerical increase (13%) in total SCFA and apparently rumen degradable carbohydrates (ARDC) after 24 h in vitro incubation, without any loss of biomass. However, this increase was not clearly correlated to results of the enzyme assays. Conclusion: This study indicates pre-treatment of OPF with enzyme extracts from specific WRF to be promising to enhance the ruminal degradability of OPF without simultaneous loss of biomass.


Background
The high lignin-(hemi)cellulose complex in the lignocellulosic agricultural by-products cause a serious impairment of their degradability by rumen microbes [1]. Thus, upgrading such lignocellulosic byproducts could be of interest to increase their nutritional value for livestock production. Cereal straws (e.g. from rice, oat, wheat) as well as oil palm fronds (OPF) are examples of lignocellulosic byproducts. Oil palm fronds, consisting of leaf and petioles, are generally regarded as agriculture waste from the oil palm industry and widely used in ruminant diets in Malaysia [2] despite their considerable lignin content (about 200 g/kg DM, e.g. [3]).
White rot fungi (WRF) have been reported as the most effective basidiomycetes to degrade lignin and improve degradability of fibrous material such as wheat straw [4], [5], rice straw [6] and oat straw [7].
Biological pre-treatment of OPF with WRF also showed promising results to breakdown lignocellulosic bounds in OPF, thereby increasing in vitro rumen apparent degradability of carbohydrates by up to 31% [3]. Although lignin degradation by WRF might stimulate rumen fermentation, there are still major drawbacks of this approach due to non-selective degradation of lignin, which was accompanied by cellulose and hemicellulose losses, and hence decreased carbohydrate availability for ruminal microbial degradation [8]. Treatment of wheat straw with enzymatic extracts from fungal strains rather than with the fungi itself might minimize DM losses [4]. In that study, enzymatic extracts isolated from four fungal strains (i.e. 2 strains of Trametes versicolor, Bjerkandera adusta and Fomes fomentarius),, enhanced lignin degradation and increased rumen neutral detergent fibre (NDF) degradability of wheat straw, up to 13% as compared with non-treated wheat straw.
However, effectiveness of WRF to degrade lignin and improve rumen fermentation is substratespecific as well as dependent on conditions of the solid state fermentation (e.g. origin of the additional N source, duration of the solid state fermentation). Hence, in the current study, enzyme extracts were screened for their effectiveness to improve rumen degradability of OPF. These extracts were prepared using WRF which showed potential to improve rumen degradability of OPF [3] and the effectiveness of the extracts was compared with that of the fungi through a mass balance approach.
Lignolytic, cellulolytic and hemicellulolytic enzyme activities of the extracts were also determined to assess whether they are linked to changes in rumen degradability of OPF.

Substrates
Oil palm fronds were provided by the Malaysian Agricultural Research and Development Institute (MARDI), Selangor, Malaysia. The fresh OPF were chopped to 1-2 cm length and air dried before being inoculated with WRF.

Fungal strains
Three WRF strains, i.e Ceriporiopsis. subvermispora CBS 347.63 (purchased from The Centraalbureau voor Schimmelcultures (CBS), Utrecht, The Netherlands), Lentinula. edodes UF 20911 and Ganoderma. lucidum UF 20706 (isolated from decaying plant material collected in the North of Portugal and characterized by Dias [9]) were used to obtain the enzymatic extracts. Fungi were cultured on potato dextrose agar plates and incubated at 28 o C for 10 days.
2.3 Inoculation of OPF with WRF to produce enzyme extracts Inoculation of OPF with WRF to produce enzyme extracts were performed in quadruplicate under solid state fermentation, as described by [10]. Briefly, the enzymes were extracted in 250 ml Erlenmeyer flasks with culture media content (each flask contained OPF, 15 g; glucose, 22.5 mg and 45 ml deionized water). The control flasks only contained these substrates whereas additionally, either of two nitrogen sources (ammonium sulphate, (NH 4 ) 2 SO 4 ,2.4 mM or sodium nitrate, NaNO 3 , 2.4 mM) were added in treatments and referred to as NH 4 -medium and NO 3 -medium, respectively, throughout this study. The addition of nitrogen source is to stimulate the fungi's growth, thus, enhancing their enzyme production. Before inoculating OPF with WRF, the flasks were autoclaved at 121 o C for 20 minutes. After the cooling process, three 10 mm agar plugs, taken from each agar plate with fungi, were added to each flask under sterile conditions. Flasks were then incubated at 28 o C and OPF inoculated with one of the fungi was harvested after 15, 30 and 45 days of inoculation. Afterwards, contents of the culture flasks were suspended in 150 ml of deionized water and incubated on a rotary shaker (100 rpm) at 28 o C for 3 h. Extracts were filtered (Whatman GF/A) and 0.06 g polyvinyl polypyrrolidone (PVPP) was added before being centrifuged at 12 000 x g for 10 min. The supernatant were recovered and used for enzyme activity determination and pre-treatment of OPF.

Determination of enzymatic activities
Enzymatic activities were determined on quadruplicate samples using a Helios UV-vis spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Lignolytic enzyme activities such as laccase, manganese peroxidase (MnP) and lignin peroxidase (LiP) were monitored using 0.1 to 0.4 ml supernatant of the culture samples and the respective buffered substrate in 1.5 ml total reaction volume. Laccase activity was determined according to [10] by measuring the oxidation of 2.0 mM 2,2'-azino-bis-3-ethylbenzthiazoline-6-sulphonic acid (ABTS) buffered with 100 mM citrate-phosphate (pH 4.0) and the formation of an ABTS cation radical was monitored at 420 nm (ɛ 420 = 36.0 mM -1 cm -1 ). Manganese peroxidase activity was determined according to the modified method of Heinfling et al. (1998) by the formation of Mn 3+ -tartrate (ɛ 238 = 6.5 mM -1 cm -1 ) from 0.10 mM MnSO 4 using 100 mM tartrate buffer (pH 5.0) and 0.10 mM H 2 O 2 . Lignin peroxidase activity was monitored at pH 3.0 according to [11] and the formation of veratraldehyde was monitored at 310 nm (ɛ 310 = 9.3 mM -1 cm -1 ). For cellulolytic enzyme assays, activities of carboxymethylcellulase (CMCase) and Avicel digesting cellulase (avicelase) were measured according to the IUPAC recommendations [12], using substrate solutions of 10 g/l carboxymethylcellulose and 10 g/l avicel (cellulose microcrystalline) in a 50 mM citrate buffer (pH 4.8) at 50 C for 30 min and 3 h, respectively. The reducing sugars released were determined by dinitrosalicylic acid (DNS), using glucose as a standard. As for hemicellulolytic enzyme, xylanase, the activity was determined under similar conditions as for CMCase, except that 10 g/l of xylan solution was used as the substrate. The release of reducing sugars were determined by DNS, using xylose as a standard [10]. were then put in a forced air oven at 40 C for 5 days. Every 24 h, 1 ml H 2 O 2 were added to each flask to stimulate the production of MnP. After pre-treatment, the contents of the flasks were transferred to a plastic container for freeze-drying. The OPF residues obtained after freeze-drying were used to determine DM losses and to perform in vitro ruminal incubations. Besides pre-treatments with three WRF, OPF pre-treated with non-enzyme extracts (citrate buffer only) were included as a control.

In vitro rumen incubation and analysis
In vitro ruminal incubations (24 h) were carried out to provide evaluation of the degradability of pretreated OPF with enzyme extracts. The in vitro batch incubations were performed by including the three replicates per treatment (as described in 2.5) in glass gastight serum flasks as described by [13]. Briefly, 0.250 g of pre-treated OPF were added to each flask. Flasks containing 0.250 g of nontreated OPF were also included as negative control. One milliliter of distilled water was added and the flasks were flushed four times with CO 2 before the addition of rumen fluid. Rumen fluid collection was done from three fistulated Texel sheep, fed with good quality hay at maintenance level, and was (Shidmadzu Corporation's Hertogenbosch, The Netherlands) according to [14]. Net production of SCFA was calculated by subtracting the amounts at the 0 h time point from amounts found after 24 h of incubation. Apparently rumen degradable carbohydrates (ARDC) were calculated as: ARDC (mg) = (Acetate/2 + Propionate/2 + Butyrate) x 162/1000, with 162 as the assumed molecular weight (g) of 1 mol of fermented carbohydrates (Demeyer 1991).
Acetate, propionate and butyrate were expressed as net micro molar productions.

Statistical analysis
All data were statistically analyzed using SPSS (2006). All the parameters measured were evaluated separately using a general linear model (univariate). For enzyme activities, the parameters measured were used to assess the main and interaction effect of fungi (A i = 1-3 ), inoculation day (B j = 1-3 ) and medium (C k = 1-3 ), according to Y ijk = μ + A i + B j + C k + AB ij + AC ik + BC jk + ABC ijk + ξ ijk where Y ijk is the response; μ the overall mean; A i the effect of fungi (C. subvermispora, L. edodes or G. lucidum); B j the effect of inoculation day (15, 30 or 45 days); C k the effect of medium (no addition N source, NH 4 medium or NO 3 medium); AB ij the interaction between fungi and inoculation day; AC ik the interaction between fungi and medium; BC jk the interaction between inoculation day and medium; ABC ijk the interaction between fungi, inoculation day and medium; and ξ ijk the residual error. As for the in vitro incubation, the parameters measured were used to assess whether pre-treatment with enzyme extracts improved rumen degradability of OPF, through the comparison of OPF pre-treated with each of the enzyme extracts with OPF pre-treated with the citrate buffer (control) by a Dunnet post-hoc test, according to Y ij = μ + D i + ξ ij where Y ij is the response; μ the overall mean; D i the effect of OPF pre-treated with enzyme extracts; and ξ ij the residual error. Effect with P<0.05 were considered statistically significant.

Enzymatic activities of the enzyme extracts
Enzymatic activities of the extracts obtained from inoculation of OPF with any of the three WRF are presented in Figure 1 and 2. In the current study, activity of the lignolytic enzymes, laccase and MnP was detected (Figure 1 and 2, respectively), whereas LiP was not detected for any of the fungi over any of the inoculation periods tested. Among the three fungi, L. edodes and C. subvermispora showed the highest and lowest activity of laccase, respectively ( Figure 1 and Table 1). Laccase activity developed gradually over the incubation period showing the highest activity after 45 days of inoculation ( Figure 1 and Table 1). A higher laccase activity was observed for N-supplemented media (Table 1), irrespective of the N-source used. On the contrary, the MnP activity decreased over the inoculation period. For G. lucidum, duration of the inoculation affected laccase and MnP only to a limited extent showed the lowest MnP activity among the three fungi tested ( Figure 2 and Table 1).
Again, N supplemented to the media generally increased MnP activity ( Table 1).
All fungi extracts showed cellulolytic enzyme activities such as CMCase (endoglucanase) and avicelase (exoglucanase). Carboxymethylcellulase activity in C. subvermispora extracts was highest ( Figure 2a and Table 1). The activity of CMCase differed over the inoculation period with the highest and lowest activity being observed after 15 and 30 days of inoculation, respectively ( Figure 2a and Table 1). Lower CMCase activity was observed for N-supplemented media, irrespective of the Nsource ( Table 1). As for avicelase (Figure 2b), its activity was very low, as compared with CMCase.
The highest and lowest activity of avicelase was observed for C. subvermispora and L. edodes, respectively (Table 1). Nevertheless, the production of avicelase activity was relatively stable over the incubation period except for a slight decrease observed after 30 days of inoculation (Table 1). With respect to the type of medium, the extract obtained after fungi inoculation in a NO 3 -enriched medium showed a lower activity, although the difference is minor and generally, activities of this enzyme are low. The hemicellulolytic enzyme, xylanase, was also produced during the inoculation period by all fungi, with L. edodes showing the highest activity. The production of xylanase was increased after 45 days inoculation compared to 15 and 30 days inoculation (Figure 2c and Table 1). The lower xylanase activity was observed for N-supplemented media, irrespective of the N-source (Table 1).
In general, no differences were observed in enzyme activities between two N sources which stimulated lignolytic and reduced fibrolytic enzyme activity, respectively (Table 1). Nevertheless, interaction effects indicate this might vary according to inoculation day and fungal strain (Table 1).
For instance, enzyme activities in extracts of G. lucidum only showed minor differences between inoculation media, particularly for the longer inoculation period, except for laccase activity.
3.2 In vitro rumen degradation of OPF pre-treated with enzymatic extracts Oil palm fronds pre-treated with enzyme extracts obtained after 15, 30 and 45 days with inoculation of C. subvermispora, L. edodes and G. lucidum did not significantly change total SCFA production and ARDC, as compared with OPF pre-treated with citrate buffer (control; Table 2). Nevertheless, OPF pretreated with enzyme extracts obtained after 45 days of inoculation with G. lucidum showed a numerical increase by 13% in total SCFA production and ARDC as compared with the control. This increase was shown for the enzyme extract from G. lucidum inoculated in media containing NO 3 or without extra N source. As for the individual SCFA production, the enzymatic treatment generally resulted in decreased acetate and increased propionate and butyrate proportion of pre-treated OPF as compared with the control.

Enzyme activities of the extracts
The production of lignolytic enzymes by WRF when inoculated with fibrous material has been reported by several authors [15], [4], [16]. Laccase and MnP are considered to be the most common lignolytic enzymes produced by WRF [17], [18] which is in line with the lignolytic enzymes determined in the current study. Nevertheless, laccase and MnP activities reported in the current study (e.g L. edodes) were higher as compared with other studies using the same protocol (0.144 and 0.160 U/ml, respectively) through inoculation of wheat straw during 15 days by Trametes versicolor [4].
Furthermore, in the current study, a higher activity of laccase as compared to MnP was observed, which is in contrast with other studies (e.g [19], [20]) in which 10 times higher levels of MnP activity were reported than laccase activity after 28 days inoculation of wheat straw with WRF (e.g Nematoloma frowardii).. These differences mainly might be due to the production of enzyme extracts in different growth media and using different strains of fungi. Indeed, variation in production of lignolytic enzymes (e.g. laccase and MnP) between fungal strains was also observed in the current study with L. edodes being the highest producer, followed by G. lucidum and C. subvermispora. In the current study, LiP was not detectable in the enzyme extract. Some former studies reported some WRF to produce all three lignolytic enzymes (e.g laccase, MnP and LiP) while others produce only one or two of them [20].
The higher endoglucanase (CMCase) as compared with exoglucanase (avicelase) activity indicated a common feature of cellulolytic enzyme activity [21], [4]. [10] reported 10 times higher CMCase than avicelase activity in Irpex lacteus cultivated on wheat straw, which is in line with the current study (10 to 20 times higher). The higher xylanase activity as compared with CMCase suggested a possible preference of the currently studied WRF to degrade hemicellulose rather than cellulose [6].
Accordingly, in a study with six fungal species, hemicellulose content was much lower than cellulose in fungi-treated wheat straw as compared with untreated wheat straw [22].

Effects of pre-treated OPF with enzyme extracts on in vitro rumen degradability
A numerical increase of 13% in total SCFA production and ARDC for OPF pre-treated with enzyme extracts of G. lucidum obtained after 45 days of inoculation in NO 3 -medium or in medium without extra N-source did not correlate with differences in enzyme activities: lignolytic enzyme activities in these extracts were lower as compared with activities measured in extracts obtained from the other fungal strains tested in the current study. However, also in other studies no direct relationship between lignolytic enzyme activities and the extent of degradation was found, e.g. for wheat straw [23]. Eventually, the improved OPF degradability was induced by enzymes present in the extract of which the activity was not measured. Several studies have reported that other enzymes such as βglucosidase, feruloyl and acetyl esterases could play an important role during the degradation of lignocellulosic materials such as wheat straw [24]. [25] showed that feruloyl esterases have the ability to hydrolyze lignin-hemicellulose bounds, improving further exposure of structural carbohydrates to cellulolytic enzymes.
In our former study [3], we showed the NDF and ADF content of OPF decreased when colonized with WRF. These decreases in fiber fractions are consistent with degradation of lignin which is part of ADF and NDF, which then could allow accessibility of the structural carbohydrates by cellulolytic and hemicellulolytic bacteria and further enhance the degradability of OPF in the rumen. The decrease of cellulose indicates non-selective degradation by fungi of both lignin and cellulose [26]. Nevertheless, a decrease in cellulose content due to WRF treatment is one of the major drawbacks of this approach as it results in a biomass loss. This is illustrated in our previous study [3]. where a 10% and 31% increase in rumen degradability as provoked by G. lucidum and L. edodes, respectively, were associated with a net loss of degradable biomass (250 and 200 g/kg DM for G. lucidum and L. edodes, respectively). An overview of degradability and mass balance from non-treated OPF and OPF pretreated with fungi is presented in Table 3 and 4, respectively.
As illustrated in Table 4, OPF pre-treatment with G. lucidum does not result in a net increase in rumen degradable biomass despite the improvement in rumen degradability. For biomass losses at an order of magnitude of 20%, an increase in degradability of more than 30% is required (e.g LEW9) to ensure a net gain of degradable biomass. Even then, improvement in degradable biomass is limited.
Therefore, pre-treatment of OPF with enzyme extracts from WRF seems promising to overcome the biomass loss problem. Indeed, [4] showed that enzymatic extracts from WRF enhanced lignin degradation and further increased rumen NDF degradability of wheat straw (up to 13%), with minimal DM losses. Their improvement in degradability by pre-treatment with enzyme extract from WRF is in agreement with the current study, where the pre-treatment of OPF with an enzyme extract of G.
lucidum showed an increase in OPF degradability by up to 13%. This improvement was higher than OPF pre-treated with G. lucidum (10%; [3]) with an advantage of not losing biomass.
Mass balances for this pre-treatment approach are shown in Table 5, including the recovery of the OPF which was used for the production of the enzyme extract. For this OPF, degradability characteristics of fungal pre-treated material [3] were assumed. Nevertheless, this recovery would not be required when an efficient enzyme production on a larger scale is applied, requiring relatively less OPF. To assess the mass balance of the OPF pre-treated with enzyme extracts, both the biomass required for enzyme extraction as well as minor losses occurring during pre-treatment of OPF with enzyme extract were considered. The current lab scale enzyme production and OPF pre-treatment required 14 and 86% of the OPF biomass, respectively (Table 5).
To produce the enzyme extract, 15 g OPF and 195 ml water was used in each flask. As we had 4 replicates, the total amount of OPF and water used was 60 g and 780 ml, respectively. After filtration, the enzyme extracts from the four replicates were pooled and only 20 ml were used for pre-treatment of OPF with enzyme extract, which represented the use of 1.5 g of OPF (20 ml/780 ml = 2.5% x 60 g = 1.5 g OPF was needed to produce 20 ml enzyme extract), for the total treatment of 11 g of OPF, which represents 14% extra OPF needed to produce the enzyme extract (20 ml enzyme extract needed to pre-treat 11 g OPF from which OPF material with a 13% higher in degradability was reached).
When a possible recovery of this biomass is considered with a degradability improvement of 10% (as obtained for direct fungal treatment of OPF, [3]), a 9% gain of degradable biomass was calculated (Table 5), which is higher as compared to the increase that was reached by the best strategy with direct WRF pre-treatment of OPF in our former study (e.g +4.8% by LEW9). Moreover, given the relative great sensitivity of the direct fungal treatment (e.g inoculation period is very critical), a more standardized treatment with an enzyme extract (produced under more standardized (industrial) conditions) might be of interest. Moreover, the advantage of minimized DM losses is mainly of interest when upscaled enzyme production would be optimized and would require only minimal amounts of biomass.