Article pubs.acs.org/JAFC
Transformation with TT8 and HB12 RNAi Constructs in Model Forage (Medicago sativa, Alfalfa) Affects Carbohydrate Structure and Metabolic Characteristics in Ruminant Livestock Systems Xinxin Li,†,§ Yonggen Zhang,*,§ Abdelali Hannoufa,# and Peiqiang Yu*,†,Δ †
College of Agriculture and Bioresources, University of Saskatchewan, 51 Campus Drive, Saskatoon, Saskatchewan, Canada S7N 5A8 College of Animal Science and Technology, Northeast Agricultural University, Harbin 150030, China # Agriculture and Agri-Food Canada, 1391 Sandford Street, London, Ontario, Canada N5V 4T3 Δ Tianjin Agricultural University, Tianjin 300384, China §
ABSTRACT: Lignin, a phenylpropanoid polymer present in secondary cell walls, has a negative impact on feed digestibility. TT8 and HB12 genes were shown to have low expression levels in low-lignin tissues of alfalfa, but to date, there has been no study on the effect of down-regulation of these two genes in alfalfa on nutrient chemical profiles and availability in ruminant livestock systems. The objectives of this study were to investigate the effect of transformation of alfalfa with TT8 and HB12 RNAi constructs on carbohydrate (CHO) structure and CHO nutritive value in ruminant livestock systems. The results showed that transformation with TT8 and HB12 RNAi constructs reduced rumen, rapidly degraded CHO fractions (RDCA4, P = 0.06; RDCB1, P < 0.01) and totally degraded CHO fraction (TRDCHO, P = 0.08). Both HB12 and TT8 populations had significantly higher in vitro digestibility of neutral detergent fiber (NDF) at 30 h of incubation (ivNDF30) compared to the control (P < 0.01). The TT8 populations had highest ivDM30 and ivNDF240. Transformation of alfalfa with TT8 and HB12 RNAi constructs induced molecular structure changes. Different CHO functional groups had different sensitivities and different responses to the transformation. The CHO molecular structure changes induced by the transformation were associated with predicted CHO availability. Compared with HB12 RNAi, transformation with TT8 RNAi could improve forage quality by increasing the availability of both NDF and DM. Further study is needed on the relationship between the transformation-induced structure changes at a molecular level and nutrient utilization in ruminant livestock systems when lignification is much higher. KEYWORDS: transformation of alfalfa, TT8, HB12, carbohydrate structure and nutrition, molecular spectroscopy
■
INTRODUCTION Alfalfa (Medicago sativa) is widely used as a forage for herbivores because of its high nutrient content and optimal structural to nonstructural carbohydrate ratio;1 however, the digestibility and utilization of alfalfa are hampered by its lignin content,2,3 because lignin can negatively affect rumen microbial degradation and digestion of feed by intestinal enzymes.4 These fermentation-resistance effects are associated with the lignin monomer composition, chemical functional groups, and crosslinking to wall polysaccharides.3,5−7 Lignin, the second most abundant component of secondary cell walls, is a complex structural phenylpropane polymer consisting of three monolignols: p-hydroxyphenyl (H), guaiacyl (G), and syringyl (S). Monolignols are synthesized from the phenyl phenylpropanoid/shikimic acid pathway,8−10 which is regulated by AC-rich elements, lignin activators from the MYB family, and three NAC transcription factors, as well as environmentally controlled lignin repressors.9 The physical properties of lignin are connected with its composition differences. A higher S:G ratio, resulting in a higher methoxyl content, could reduce forage digestibility.8,11 Genetic engineering has proven to be an innovative approach for regulating the lignin biosynthesis pathway by altering lignin composition, concentration, and/or lignin conformation.8,10,12 Recently, the proanthocyanidin regulatory gene TRANSPARENT TESTA 8 (TT8) and a homeodomain leucine zipper © 2015 American Chemical Society
class I HB12 gene were shown to have naturally low expression levels in low-lignin tissues of Brassica napus and high levels in high-lignin tissues (Hannoufa, unpublished data). TT8 gene encodes a basic helix−loop−helix domain protein,13,14 which could potentially interact with MYB proteins, involved in both the regulation of lignin and flavonoid biosynthesis pathways.15,16 In contrast, HB12 is a homeodomain-leucine zipper (HDZip) gene, which is induced by water deficit and abscisic acid (ABA) signaling.17,18 It is now possible to use RNAi-mediated approaches to suppress single or multiple genes in secondary cell wall synthesis.10 Hence, we recently developed alfalfa plants with transformation with TT8 and HB12 RNAi constructs. Down-regulation of HB12 and TT8 genes through RNAi technology19 could affect alfalfa inherent structure at a molecular level. Advanced molecular spectroscopy, synchrotron radiation, and globar sourced Fourier transformed infrared spectroscopy are powerful analytical tools that could reveal inherent carbohydrate and protein structure relevant to nutrient profiles for livestock.20−22 To date, there is no study on inherent structure changes of alfalfa induced by TT8 and HB12 Received: Revised: Accepted: Published: 9590
April 3, 2015 September 24, 2015 September 30, 2015 October 22, 2015 DOI: 10.1021/acs.jafc.5b03717 J. Agric. Food Chem. 2015, 63, 9590−9600
Article
Journal of Agricultural and Food Chemistry
GACCAGTTTCATC; reverse). For TT8, we used TA29816_246c (CACCTTCTGCAGCACTTTCACCTG; forward) and TA29816_602r (GCGTTAATAGGGTCATCGACA, reverse). The cDNA fragments were directionally cloned into a Gateway entry vector and then recombined into the pHellsgate12 RNAi destination vector25 via LR Clonase. The constructs were used to transform Agrobacterium tumefaciens LBA4404 by electroporation, and the resulting A. tumefaciens strains were used to transform alfalfa using the Agrobacterium-mediated transformation method described in Aung et al.19 Putative transformants were selected on media containing kanamycin, and the presence of the transgene was verified by PCR using CaMV35S-specific primer (CAATCCCACTATCCTTCGCAAGACCC) as forward primer and TA22700_634r (HB12) and TA29816_602r (TT8) as gene-specific reverse primers. Processing and Chemical Analysis. Alfalfa samples were analyzed for DM, ash, and ether extract (EE) according to the AOAC.26 Neutral detergent fiber (NDF), acid detergent fiber (ADF) (ANKOM A200 Filter Bag Technique, ANKOM Technology, Fairport, NY, USA), and acid detergent lignin (ADL) were determined according to the procedure described by Van Soest.27 The NDF, ADF, and ADL contents are common parameters among the others that are used to assess forage quality and used for diet formulation. The starch content was analyzed using the Megazyme method.28 Sugar was analyzed according to the method of Dubois et al.29 The contents of nonfiber carbohydrate (NFC), total carbohydrate (CHO), hemicellulose, and cellulose were calculated according to methods described in refs 27 and 30. Prediction of Rumen Carbohydrate Degradation Profiles: Nutrient Utilization and Availability in the Rumen Using the Updated CNCPS Approach. Rumen degradable carbohydrate fractions were predicted using model equations based on the updated CNCPS system.31 These included RDCA4 (rumen degraded carbohydrate CA4 (sugar) fraction), RDCB1 (rumen degraded carbohydrate CB1 (starch) fraction), RDCB2 (rumen degraded carbohydrate CB2 (soluble fiber) fraction), RDCB3 (rumen degraded carbohydrate CB3 (available NDF) fraction, and total degraded CHO fraction (TRDCHO). Rumen undegradable carbohydrate fractions included RUCA4 (rumen escaped carbohydrate CA4 fraction), RUCB1 (rumen escaped carbohydrate CB1 fraction), RUCB2 (rumen escaped carbohydrate CB2 fraction), RUCC (rumen escaped carbohydrate CC fraction), and total escaped CHO fraction (TRUCHO). Degradation rates (Kd) for alfalfa forage and their passage rates out of rumen (Kp) for the fractions were obtained from the cattle feeds database of Agricultural Modeling and Training System-Cattle-Professional (AMTS, Cornell, Ithaca, NY, USA; AMTS, 2010).32 In Vitro Digestibility of Dry Matter and Fiber at 30 and 240 h. The 30 and 240 h in vitro digestibility of DM (ivDM30, ivDM240) and NDF (ivNDF30, ivNDF240) of different genotypes of alfalfa populations were determined according to the procedure described in the Operator’s Manual Daisy II incubator D200 and D200I (ANKOM Technology, Fairport, NY). This in vitro measurement has been also recommended by the newly updated CNCPS 6.5 system. Rumen fluid was collected through the ruminal cannula from two lactating Holstein cows fed a standard Total Mixed Ration (TMR) and housed at the Rayner Dairy Research and Teaching Facility farm, University of Saskatchewan, Saskatoon, Canada. The samples incubated for 240 h were re-inoculated at 120 h with the same amount of rumen liquid and buffer solutions.33 Two experimental runs in each incubation time were carried out. After 30 and 240 h of inoculation, fiber bags containing alfalfa samples were removed and rinsed with cold water until water effluent was clear using APS standard washing procedure. The incubated residues were dried and the DM and NDF contents determined. The equations used for calculating 30 and 240 h in vitro digestibility of DM and NDF were adapted from ref 34 and Operator’s Manual Daisy II incubator D200 and D200I (ANKOM Technology) . Advanced Molecular Spectroscopic Technique. Noninvasive and nondestructive Fourier-transformed infrared-vibration (FT/IR) spectroscopy experiments were carried out at the APS molecular
in relation to nutrient chemical profiles, rumen degradation, and intestinal digestion in alfalfa populations for ruminants. The objectives of this study were (I) to investigate the effect of transformation of alfalfa with TT8 and HB12 RNAi constructs on (1) nutrient chemical profiles, (2) carbohydrate structural characteristics, (3) predicted nutrient availability in terms of rumen degradable and undegradable fractions using the updated CNCPS system, and (4) in vitro digestibility at 30 and 240 h of incubation; and (II) to quantify relationships between carbohydrate (CHO) structure profiles and nutritional values in ruminant livestock system. The hypothesis of this study was that transformation with TT8 and HB12 RNAi constructs affecting lignin and flavonoid biosynthesis pathways could alter the inherent carbohydrate structure and the digestibility of alfalfa forage in ruminant livestock systems by depressing the biosynthesis of antiquality cell wall metabolites.
■
MATERIALS AND METHODS
Chemicals. In this study, all of the chemicals used in our laboratories were of reagent grade. The detailed chemical information and manufacturers were reported in our previous publication by Khan et al.,23 as follows: Sodium hydroxide (purity > 99%), trichloroacetic acid (purity = 99.6%), sodium tungstate (Na2WO4·2H2O; purity = 100%), sulfuric acid (purity = 70%), hydrochloric acid (purity = 38%), ethyl ether (purity = 99%), sodium borate (Na2B4O7·10H2O; purity = 99.5%), sodium phosphate (NaH2PO4·H2O; purity = 99.5%), kjeltabs (0.15 g of CuSO4 and 5.0 g of K2SO4; N < 0.005%), boric acid (purity > 99.5%), acetone (purity ≥ 99.5%), potassium phosphate (purity > 99%), sodium sulfite anhydrous (purity = 98.4%), and potassium hydroxide (purity = 85%) were purchased from Fisher Scientific (Ottawa, ON, Canada). Thymol (C10H14O; purity = 99.5%) and enzymes pepsin and pancreatin were purchased from Sigma-Aldrich Co. (St. Louis, MO, USA). Heat-stable α-amylase (17400 liquefon units/mL); NDF solution containing 30 g of sodium dodecyl sulfate, 18.61 g of ethylenediaminetetraacetic disodium salt, 6.81 g of sodium borate, 4.56 g of anhydrous sodium phosphate diabasic, and 10 mL of triethylene glycol, and ADF solution containing 20 g of cetyltrimethylammonium bromide in 1 N sulfuric acid were purchased from Ankom Technology (Macedon, NY, USA).23 Plant Material. Alfalfa (Medicago sativa) clone N4.4.224 was used as the wild type control and as the recipient for transformation with HB12 and TT8 RNAi constructs. The alfalfa clone was obtained from Dr. Daniel Brown (Agriculture and Agri-Food Canada, AAFC, London, ON, Canada). All alfalfa plants were grown under greenhouse conditions at 21−23 °C and 16 h light per day with halogen lights having been applied after 6:00 p.m. Light intensity of 380−450 W/m2 (∼500 W/m2 at noon) and a relative humidity of 70% were maintained throughout the growth period. Harvests of individual plants were conducted at early-to-mid vegetative stage. Plants were stored in bags. Each bag represented one cut of one plant grown in a pot in the greenhouse. The HB12 had 11 bags in total, which were divided into two replicated samples. The TT8 genotype had five bags, which were also divided into two replication samples. Harvests from each genotype were freeze-dried individually for each plant and ground through a 1 mm screen (Retsch ZM-1, Brinkmann Instruments Ltd., Ontario, Canada) at the Department of Animal and Poultry Science, University of Saskatchewan. Two replicate samples of each genotype population were drawn from individual plants (combining different individual plants within each genotype). Alfalfa populations were named TT8-RNAi alfalfa (n = 2), HB12 RNAi alfalfa (n = 2), and control alfalfa (n = 2). Generating RNAi Constructs and Transformation of Alfalfa. RNA was extracted from the fourth internode of M. sativa plants in the vegetative phase of development. After cDNA synthesis, 171 and 336 bp fragments, of HB12 and TT8, respectively, were amplified. For HB12, we used primers TA22700_463c (CACCGCATGGAAAGTGCATCAGAA; forward) and TA22700_634r (CAAAATTCATAA9591
DOI: 10.1021/acs.jafc.5b03717 J. Agric. Food Chem. 2015, 63, 9590−9600
Article
Journal of Agricultural and Food Chemistry
Table 1. Effect of Transformation with TT8 and HB12 RNAi Constructs on Chemical Profiles in Model Forage of Alfalfa (Medicago sativa)a transformation with TT8 and HB12 RNAi constructs (T) item
HB12, n = 2
TT8, n = 2
SEMb
91.5 ± 0.20 7.9 b ± 0.08 1.6 ± 0.17 0.6 ± 0.17 92.1 a ± 0.08
91.7 ± 0.08 8.2 a ± 0.01 1.9 ± 0.11 0.9 ± 0.11 91.9 b ± 0.01
92.0 ± 0.19 8.3 a ± 0.06 1.5 ± 0.25 0.5 ± 0.25 91.7 b ± 0.06
0.12 0.04 0.13 0.13 0.04
0.12 0.02 0.20 0.20 0.02
0.12 0.01 0.61 0.61 0.01
69.7 a ± 0.16 3.7 a ± 0.23 5.1 ± 1.31 28.6 b ± 0.70 23.6 ± 1.29 2.4 ± 0.08 5.0 ± 0.57 21.2 ± 1.20 43.2 a ± 0.21
70.5 a ± 0.35 1.4 b ± 0.17 2.9 ± 0.54 33.2 a ± 0.78 26.3 ± 0.12 4.2 ± 0.39 6.9 ± 0.65 22.1 ± 0.27 39.5 b ± 1.27
67.6 b ± 0.33 1.8 b ± 0.47 3.0 ± 0.39 31.0 a ± 0.76 23.5 ± 1.36 3.5 ± 1.29 7.6 ± 2.12 20.0 ± 2.64 38.1 b ± 0.37
0.21 0.23 0.60 0.53 0.77 0.55 0.94 1.19 0.55
< 0.01 0.01 0.13 0.02 0.13 0.20 0.29 0.53 0.02
0.09 < 0.01 0.06 0.01 0.25 0.12 0.15 0.93 < 0.01
no genetic modification, control (CT), n = 2
chemical profile (% DM ± STD) DM (%) ash EE FA OM carbohydrate profilec (% DM ± STD) CHO starch sugar NDF ADF ADL hemicellulose cellulose NFC
P
contrast, P, CT vs T
a Means with different letters within the same row differ significantly (P < 0.05). bSEM, standard error of the mean; the Tukey−Kramer method was used for multitreatment comparisons. cCHO, carbohydrate; NFC, nonfiber carbohydrate; EE, ether extract; OM, organic matter; NDF, neutral detergent fiber; ADF, acid detergent fiber; ADL, acid detergent lignin.
In Vitro Digestibility Analysis. The model used for the analysis was Yijk = u + trti + rj + eijk, where Yij was the dependent variable ij, u was the population mean for variable, trti was the effect of treatment, that is, alfalfa genotypes, as a fixed effect, rj was the experimental run as a random effect, and eijk was the random error associated with ijk. The differences between the control and HB12 and TT8 RNAi alfalfa populations were compared by Contrast statements in SAS. Three model assumptions in RCRD were checked: (1) the treatments have a common variance, (2) the model residual data are normally distributed, and (3) the random block effects are normally distributed. The normality check was carried out by using PROC UNIVARIATE with Normal and Plot options in SAS. Correlation Analysis of CHO Spectral Profiles with CHO Nutrient Supply. Correlations between CHO molecular structural spectral profiles and CHO chemical characteristics, predicted rumen degradable and undegradable CHO fractions, and in vitro degradability were analyzed using the PROC CORR procedure of SAS 9.4 (SAS Institute, Inc.)39 with a Spearman option after a normality test. Multiple Regression Analysis of CHO Spectral Profiles with CHO Nutrient Supply. Multiple regression analysis of CHO molecular structure spectral profiles with CHO nutrient supply was performed to select the best IR parameters that explain nutrient values using the PROC REG procedure of SAS 9.3 (SAS Institute, Inc.) with a reversed stepwise option.39 The following model was used for the multiple regression with model variable selection: model Y = spectral parameter 1 + spectral parameter 2 + spectral parameter 3 + spectral parameter 4 + ... + error. The model used a “STEPWISE” option with variable selection criteria: “SLENTRY = 0.05, SLSTAY = 0.05”. All variables left in the final prediction models were significant at the 0.05 level. Residual analysis was performed using the Univariate procedure of SAS with Normal and Plot options. Collinearity detection was conducted using the VIF option of SAS to eliminate the influence of correlated dependent variables to the prediction of independent variables. A Tukey−Kramer method was used to separate means among the treatments with the additional software “pdmix 800” macro. Statistical significance was declared at P < 0.05, and trends were noted at 0.05 ≤ P ≤ 0.10.
spectroscopy laboratory, University of Saskatchewan, Canada, to determine carbohydrate-related molecular structure spectral features. The alfalfa samples were freeze-dried and ground through a 0.50 mm screen with the same grinder used in chemical analysis. The spectral data were obtained from the mid-IR region (ca. 4000−800 cm−1) at a resolution of 4 cm−1 and 128 co-added scans with SpectraManager II software, using JASCO FT/IR-4200 (JASCO Corp., Tokyo, Japan). The molecular spectroscopy instrument was equipped with a ceramic infrared light source and a deuterated L-alanine doped triglycine sulfate detector (JASCO Corp., Tokyo, Japan), employing a MIRacle ATR accessory module, as well as a ZnSe crystal and pressure clamp (Pike Technologies, Madison, WI, USA). Each alfalfa sample was analyzed five times with five subsamples from each alfalfa sample. Univariate Molecular Spectral Analysis of Carbohydrate Spectral Data. The spectral data were analyzed by OMNIC 7.3 software (Spectra Tech, Madison, WI, USA). In accordance with published studies,35−38 the well-known molecular spectral parameters for carbohydrate-related functional groups were included in this study: (1) a broad structural CHO peak area (A_StCHO, region and baseline, ca. 1485−1188 cm−1) featuring three major peaks at ca. 1415, 1370, and 1315 cm−1; (2) a cellulosic compound peak area (A_CELC, region and baseline, ca. 1294−1188 cm−1) centered at ca. 1244 cm−1; (3) a total CHO peak area (A_CHO, ca. region and baseline, ca. 1190−930 cm−1) featuring three major peaks at ca. 1150, 1100, and 1025 cm−1; and (4) a nonstructural CHO peak area (A_Non-stCHO, region and baseline, ca. 931−875 cm−1) with two peaks at ca. 918 and 895 cm−1. Statistical Analysis. Chemical and Nutrient Profiles and Molecular Spectroscopic Analyses. The chemical profiles and predicted rumen degradable and undegradable carbohydrate fractions, as well as carbohydrate-related spectra profile data were analyzed using the PROC MIXED procedure of SAS 9.4 (SAS Institute, Inc., Cary, NC, USA).40 The model used for the analysis was Yij = u + trti + eij, where Yij was the dependent variable, u was the population mean for variable, trti was the effect of treatment, that is, alfalfa genotypes, as a fixed effect, and eij was the random error. The differences between the control and HB12 and TT8 RNAi alfalfa populations were compared by SAS Contrast statements. Two model assumptions for CRD were checked: (1) the treatments have a common variance and (2) the model residual data are normally distributed. The normality check was carried out by using PROC UNIVARIATE with Normal and Plot options in SAS. 9592
DOI: 10.1021/acs.jafc.5b03717 J. Agric. Food Chem. 2015, 63, 9590−9600
Article
Journal of Agricultural and Food Chemistry
Table 2. Effect of Transformation with TT8 and HB12 RNAi Constructs on Nutrient Utilization and Availability in Terms of Predicted Rumen Degradable and Undegradable Carbohydrate Fractions in Model Forage of Alfalfa (Medicago sativa)a transformation with TT8 and HB12 RNAi constructs (T) item
no genetic modification, control (CT), n = 2
predicted rumen degradable carbohydrate fractions (% DM ± STD) RDCA4 4.5 ± 1.14 RDCB1 3.1 a ± 0.20 RDCB2 29.3 ± 1.14 TRDCHO 36.9 a ± 0.20 predicted rumen undegradable carbohydrate fractionse (% DM ± STD) RUCA4 0.7 ± 0.17 RUCB1 0.6 a ± 0.04 RUCB2 5.0 ± 0.20 RUCC 5.7 ± 0.19 TRUCHO 12.0 ± 0.18
HB12, n = 2
TT8, n = 2
SEMb
P
contrast, P, CT vs T
2.6 ± 0.47 1.2 b ± 0.14 30.0 ± 0.49 33.7 b ± 1.10
2.6 ± 0.34 1.5 b ± 0.40 28.5 ± 1.05 32.6 b ± 0.32
0.52 0.19 0.66 0.47
0.13 0.01 0.42 0.02
0.06 < 0.01 0.90 < 0.01
0.4 ± 0.07 0.2 b ± 0.03 5.1 ± 0.08 10.1 ± 0.95 15.9 ± 0.76
0.4 ± 0.06 0.3 b ± 0.08 4.9 ± 0.18 8.3 ± 3.08 13.9 ± 3.02
0.08 0.04 0.12 1.32 1.28
0.13 0.01 0.44 0.20 0.25
0.06 < 0.01 0.90 0.12 0.16
c
a Means with different letters within the same row differ significantly (P < 0.05). bSEM, standard error of the mean; the Tukey−Kramer method was used for multitreatment comparisons. cRumen degradable carbohydrate fractions include RDCA4 (rumen degraded CA4 fraction), RDCB1 (rumen degraded CB1 fraction), RDCB2 (rumen degraded CB2 fraction), and TRDCHO (total degraded CHO fraction). eRumen undegradable fractions include RUCA4 (rumen escaped CA4 fraction), RUCB1 (rumen escaped CB1 fraction), RUCB2 (rumen escaped CB2 fraction), RUCC (rumen escaped CC fraction), and TRUCHO (total escaped CHO fraction).
Table 3. Effect of Transformation with TT8 and HB12 RNAi Constructs on in vitro DM digestion (ivDM) and NDF Digestion (ivNDF) Parameters in Model Forage of Alfalfa (Medicago sativa)a transformation with HB12 and TT8 RNAi constructs (T) itemb ivDM30 (% DM) ivDM240 (% DM) ivNDF30 (% NDF) ivNDF240 (% NDF) iNDF240 (% DM)
no genetic modification, control (CT), n = 2 78.0 84.2 32.8 53.3 12.3
b ± 2.04 a ± 1.72 c ± 7.48 b ± 8.13 a ± 2.02
HB12, n = 2 80.1 79.6 53.2 55.3 13.9
b ± 1.62 b ± 1.63 b ± 5.58 b ± 0.82 a ± 0.63
TT8, n = 2
SEMc
P
contrast, P, CT vs T
84.6 a ± 0.83 84.3 a ± 1.44 62.4 a ± 2.84 67.8 a ± 2.06 9.5 b ± 0.50
0.79 0.80 2.81 2.43 0.63
