Cry1Ab/Ac proteins released from subspecies of Bacillus thuringiensis (Bt) and transgenic Bt-rice in different paddy soils

ABSTRACT As Bacillus thuringiensis (Bt) are entomopathogenic bacteria for use as biological insecticide and serve as a main source for genetically modified crops, it is imperative to investigate fate of Bt-toxins (Cry-proteins) released in soil. Four differently textured paddy soils, i.e. sandy loam, loam, silt loam and silty clay, were selected to investigate production and release of Cry-proteins by two Bt-strains (subspecies thuringiensis and kurstaki) and by transgenic Bt-rice (cultivar TT51-1). After amended with Cry-proteins, soil can adsorb proteins while sorption rate was variable among different paddy soils, with the highest amount of Cry-proteins in the silty clay soil. During the period of incubation with Bt-strains, the silty clay soil showed significant accumulation of Cry-proteins. Additionally, level of Cry-protein production was different between two Bt-strains. Inactivation of indigenous microorganisms in the soil samples led to slightly higher production of Cry-proteins by Bt-strains in all four soil types. Cry-proteins were released via root exudates of transgenic Bt-rice in sterile hydroponic culture and in non-sterile soils throughout the growth of plants. Protein release significantly varied among different paddy soils, up to 34.9 ng Cry-proteins g−1 soil in clayey soil. It indicates differently textured paddy soils reflect variable sorption, persistence, and accumulation of Cry-proteins.


Introduction
Bacillus thuringiensis (Bt) has been found broadly distributed in terrestrial environments (Milner 1994;Bravo 2018).A distinctive feature of Bt is its ability to produce Cry-proteins (i.e.Bt-toxins) exhibiting insecticidal activity against specific agricultural pests (Read et al. 2003;Badran et al. 2016).Attributes of Bt-toxins, including a narrow host range and harmlessness, have promoted development of insecticides derived from mixtures of cells and spores of Bt-strains and made Bt-based-insecticide a useful alternative to synthetic chemical pesticides (Crecchio and Stotzky 2001;Dohrmann et al. 2013).Various products and typical formulations of Bt-pesticides are commercially available for use as microbial pest control agent in agriculture (Hofte and Whiteley 1989;Melo et al. 2016), e.g.Dipel® (DF, dry flowable Bt kurstaki) and Neudorff® (StechmückenFrei, Bt israelensis, spray concentrates).
Cry genes from Bt were introduced into a variety of crop plants, referred to as Bt-crops, including Bt-maize, Bt-rice, and Bt-cotton, so as to provide protection against insect infestation and reduce the need for chemical pesticides (Crecchio and Stotzky 2001;Kramarz et al. 2007;Dohrmann et al. 2013;Little et al. 2017).Since first commercialization for agricultural use in 1996, Bt-crops have been cultivated in up to 30 countries in 2016 (James 2016).With the introduction of Bt-plants, Bt-toxins can be released into arable soil via several different pathways, e.g.via root exudation throughout cultivation (Miethling-Graff et al. 2010;Griffiths et al. 2017), via incorporation of Bt-plant residues after harvest (Baumgarte and Tebbe 2005;Icoz and Stotzky 2008;Xue et al. 2014), and probably some input from pollen (Losey et al. 1999;Pagel-Wieder et al. 2007).After their release, Cry-proteins bind on surface-active particles, such as clays and humic substances, thereby reducing the susceptibility of proteins to microbial degradation.This binding can also enhance the accumulation and persistence of Cry-proteins in soils (Hilbeck and Bigler 1999;Van Den Berg et al. 2013).In both field and laboratory studies, Bt-toxins have persisted in soil for up to 8 months (Tapp and Stotzky 1998;Saxena and Stotzky 2002;Zwahlen et al. 2003).A study conducted by Hopkins and Gregorich (2003) also indicated that Bt-toxins were still detectable in soil where Bt-corn has been grown successively for 4 years, even several months after the growing season.Additionally, the persistence of bound Bttoxins without loss of insecticidal activity may constitute a hazard to natural and agricultural ecosystem as a result of prolonged exposure to toxins, affecting microbe-mediated processes and functions in the soil ecosystem (Griffiths et al. 2005;Helassa et al. 2011;Fließbach et al. 2012;Neher et al. 2014).Despite a study (Icoz and Stotzky 2008) suggested the rapid degradation of Cry3Bb1 proteins from Bt-corn, the production, release, and fate of Cry-proteins from both Bt-strains and Bt-crops vary widely.Therefore, additional investigations appear to be necessary because of both the widespread application of Bt-pesticides and the continued commercial cultivation of Bt-crops (Wang et al. 2013;Ingber et al. 2018).
Unlike other Bt-crops grown in unsaturated arable soils, Bt-rice is mainly cultivated under submerged conditions and thereby on much more complex and dynamic soil systems.Cry1Ab/Ac gene was fused by the cry1Ab gene (GeneBank Accession No. X54939) and the cry1Ac gene (GeneBank Accession No. Y09787), encoding Cry1Ab/Ac proteins exhibiting high insecticidal activity against Lepidopeteran pests such as Chilo suppressalis and Scirpophaga incetulas.The insertion of exogenous cry1Ab/Ac gene into rice chromosomal DNA enables the production of Cry1Ab/Ac proteins within rice plants, i.e. transgenic Bt-rice event TT51-1 containing a synthetic version of cry1Ab/Ac gene.This cultivar was granted a biosafety certificate by China's Ministry of Agriculture in 2009 (Chen et al. 2014;Tu et al. 2000;Wu et al. 2013).The aim of this study is to investigate affinity of Cry1Ab/Ac proteins from selected Bt-strains in different paddy soils, to quantify Cry1Ab/Ac proteins produced by Bt in soil, and to analyze Cry1Ab/Ac proteins released in soil cultivated with Bt-rice.

Selection of paddy soils and Bt-strains
Four different paddy soils varying in soil texture, i.e. sandy loam, loam, silt loam and silty clay, were selected (Table 1).Soils were air-dried, crushed, sieved with 2 mm mesh, and stored at ambient temperature (AMT, 22 °C) for further use.One subspecies Bt ssp.kurstaki (Btk) serving as a main source

Affinity of Cry-proteins from Btk to paddy soil
Liquid cultures of Btk were centrifuged at 12,000 × g for 10 min at 4 °C to remove cells.Resulting supernatants containing Cry-proteins (protein solution) were collected and the protein content was measured via ELISA (see below).An aliquot (1 mL) of protein solution was pipetted into 200 mg paddy soil in test tubes (2 mL), generating an initial protein addition of 0.92 μg per 200 mg of soil.
The mixture was vortexed and then rotated by end-over-end shaking at AMT for 24 h.Soil was separated by centrifugation at 12,000 × g for 10 min at AMT.The supernatant was collected, and its Cry1Ab/Ac proteins were analyzed.Proteins adsorbed in soil were calculated by subtracting proteins remained in the supernatant from proteins initially amended and the Adsorption Rate was calculated as: The soil pellet was then re-suspended in 1 mL of sterile H 2 O MQ and rotated by end-over-end shaking at AMT for 24 h.After removal of soil by centrifugation at AMT, the resulting supernatant was analyzed to quantify Cry-proteins desorbed from soil.The Desorption Rate was calculated accordingly: Release of Cry1Ab/Ac proteins in soils incubated with Bt-strains Liquid cultures of the two Bt-strains were centrifuged at 8,000 × g for 15 min at 4 °C to obtain cell pellets, which were then re-suspended in H 2 O MQ .Volume of suspensions containing different Bt-cells was adjusted on the basis of the optical density (OD 600 nm ) to achieve same cell numbers in each inoculum at the starting point.Incubations of Bt-strains were conducted with both sterile and non-sterile (untreated) soil materials at AMT. Sterile soil material was obtained by three consecutive autoclaving at 121 °C for 30 min.Approximately 40 g of soil was transferred into a 250 mL Duran bottle and subsequently amended with 17 mL of H 2 O MQ and 1 mL of the protein suspension.They were thoroughly mixed directly after inoculation and prior to each sampling as well.Bottles were kept closed during incubation and ventilated weekly under a sterile workbench.At different intervals of incubation (i.e. 10 d, 19 d, 24 d, and 61 d), soil samples were collected and then processed to have Cry1Ab/Ac proteins extracted by following the method of Liu et al. (2018) and Wang et al. (2006).This extraction method was applied in this study for use with all soil samples.

Release of Cry1Ab/Ac protein from transgenic Bt-rice
Seeds of two rice cultivars, i.e. transgenic Bt-rice TT51-1 and its isogenic non-Bt-rice Minghui 63 (Oryza sativa ssp.indica) were germinated for 4-5 d.Afterwards, two seedlings of each cultivar were transferred in one glass tube (2.8 cm diameter × 20 cm height) filled either with freshly puddled soil (four paddy soils described above) or with sterile hydroponic cultures (Yoshida 1976).Dark grey foam was placed surrounding seedlings, about 2 cm from top of tubes.The tubes were placed in a box filled with water.A panel with holes fit for tubes was introduced and fixed on top of the box container.The upper section of each test tube (ca. 2 cm depth) was painted black to avoid light exposure (Supporting Information, Figure S1).
The plants were grown in a plant growth chamber (MLR-351, Sanyo, Japan) under controlled conditions (day/night temperature 25/20 °C, humidity 60 ± 2%, 14 h/10 h photoperiod).Plants were irrigated with sterile H 2 O MQ to control the amendment of nutrients which were replenished with nutrient solution as needed.When rice plants reached to a height of 40 cm (i.e.approximately 50 d after transplanting), both the soil and nutrient solution were sampled to measure the content of Cry1Ab/Ac proteins.

Analysis of Cry1Ab/Ac proteins
Analysis of Cry1Ab/Ac proteins were processed with an ELISA Kit (AP 003, Envirologix, Portland, ME, USA).The absorbance was measured at the wavelength of 450 nm by using ELISA Microplate Reader (EMax® Endpoint, Molecular Devices, Sunnyvale, CA, USA).Detailed method has been described in previous studies (Liu et al. 2018).

Statistics
Statistical analysis was performed with the software SPSS Statistics 20 (International Business Machines Corporation, Armonk, New York, USA) using one-way-and two-way-analysis of variance (ANOVA) and Duncan test at p < 0.05 level.

Affinity of Cry1Ab/Ac proteins to different paddy soils
Cry1Ab/Ac proteins (0.92 μg) were added to 200 mg of soil.Amounts of the proteins adsorbed varied among the differently textured paddy soils in a range of 0.22-0.54μg (Figure 1).It was significantly higher in the paddy soil of silty clay texture (p < 0.05), showing the strongest affinity.This might be explained by the fact that clayey soils are less prone to hydrophobicity as expressed by soil water repellency compared to sandy soils, owing to their larger specific surface area.Therefore, they are more easily saturated by hydrophobic organic molecules than clayey soils (Doerr et al. 2000;Helassa et al. 2011).
Desorption rate of Cry1Ab/Ac proteins, desorbed by washing with H 2 O MQ , was expressed as percentage of proteins bound in soil (Table 2).Variable desorption rate ranging from 8.0% to 26.8% was indicative of differences in protein affinity among different paddy soils, presumably due to patch-controlled electrostatic interactions (Sander et al. 2010).As proposed by Lesins and Ruckenstein (1988), electrostatic interactions govern Cry-protein adsorption to charged hydrophilic mineral surfaces.The nonuniform distribution of surface charge on Cry-protein give rise to patchcontrolled electrostatic attraction with like-charged sorbents.Higher pH attenuates electrostatic attraction and will therefore increase desorption in agricultural soil (Sander et al. 2010).

Production of Cry1Ab/Ac proteins in paddy soils
After inoculation of Bt-strains, Cry1Ab/Ac proteins were produced and released in paddy soils; additionally, increases in protein amounts were clearly indicative of strain growth in soil throughout the time of incubation (Figure 2).Two-way ANOVA was carried out to determine the relationships among sterile treatment, Bt-strains and protein amount.The sterile status was associated with Cryprotein amount (p < 0.05, Supporting Information, Table S1).Bt-strains were also associated with Cryprotein amount.
During the period of 0-19 d, Cry-proteins were accumulated in soils incubated with either Btt or Btk.Thereafter, different trends were observed between the two Bt-strains.During the period of 19-24 d, the amount of Cry-proteins released from Btt was prone to slightly decrease in soils, while proteins were continuously released from Btk and their amount increased rapidly.Such different trends may be interpreted as differences in strain intrinsic attributes, which potentially influenced their capability of producing Cry-proteins.Another experiment may help to confirm this by showing distinctive differences in Cry-protein production between the two strains (Supporting Information, Figure S2).The amount of proteins from Btk was approximately five times higher than that from Btt by the end of cultivation, although the volume of inoculated cells has been adjusted to the same level for both subspecies at the starting point.Similar findings were also reported by Mohan and Gujar (2001) that Cry-protein yields differed among various subspecies and even between same strains from different sources.They found that Cry-protein production of Btt was 28.8 μg g −1 while that of Btk was 33.7 μg g −1 and that Cry-protein content differed between Btk from different sources, i.e.Pasteur Institute, Paris, France (31.6 μg g −1 ) and BGSC, Columbus, USA (33.7 μg g −1 ).As a result of their intrinsic attributes, strains may be influenced in terms of their capability of producing Bt-toxins during their growth.
During the period of 24-61 d, the amount of Cry1Ab/Ac proteins released in paddy soils from Btk did not show remarkable changes; in contrast, the amount of proteins from Btt clearly declined.Such observations may partly be attributed to protein degradation (Tapp and Stotzky 1998;Crecchio and Stotzky 2001).When the rate of degradation overwhelmed the rate of production, the amount of Cry-proteins would start to decrease.Douville et al. (2005) investigated the persistence of Cryproteins in soil and estimated that their amount would decrease by 50% after 9-10 d.Sterilizing soil material was performed prior to the inoculation of Bt-strains so as to inhibit indigenous microbial activity, causing effects on production of Cry1Ab/Ac proteins in paddy soils.The amount of Cry-protein was slightly higher in the sterile soil material than that in the non-sterile one (Figure 2).That is to say, the inactivation of soil indigenous microorganisms led to an increasing production of Cry-proteins after inoculation of Bt-strains.On one hand, the reason could be that the growth of Bt-strains in sterile soil material was stimulated due to a lack of competition from indigenous microorganisms, therefore leading to a higher production of Cry-proteins.On the other hand, it could be attributed to the utilization of proteins as a source of carbon and energy by microorganisms in soils.Moreover, the inactivation of microorganisms could significantly inhibit degradation of Bt-toxins both in soil and in surface water (Douville et al. 2005), suggesting that decreases in the amount of Cry-proteins was most probably caused by protein biodegradation in soil ecosystem.
The production of Cry1Ab/Ac proteins varied among differently textured paddy soils after inoculation of Bt-strains.The amount of Cry1Ab/Ac proteins from Btt was up to 4. 29, 3.63, 7.33, and 24.4 ng g −1 soil in the sandy loam, loam, silt loam, and silty clay, respectively.A similar trend was also observed for Btk.Such trends indicated that soil texture together with other physicochemical parameters may influence the release of Cry-proteins by Bt-strains.It might be due to influences on cell growth in paddy soil microcosm; also, it might be attributed to strong affinity of Cry-proteins in clayey soil and, thus, protection against degradation.A slower decline of Bt-toxins in clayey soil than sandy soil, reported by Helassa et al. (2011), indicated that soil with higher clay contents display higher capability of maintaining Bt-toxins.Proteins can react with clay minerals and organomineral surfaces of soils are often reported to play a dominant role due to their large specific surface area and surface charge (Quiquampoix 1987a(Quiquampoix , 1987b;;Quiquampoix and Burns 2007).Following the assumptions based on findings about Cry-protein -adsorption and -desorption discussed previously, we may argue that the silty clay soil material (i.e.TL, 41.0% clay) exhibits strong binding for Cry1Ab/Ac proteins, protecting against microbial degradation and therefore leading to higher accumulation of proteins than the other three soil materials.
Release of Cry1Ab/Ac proteins via root exudates into paddy soils and hydroponic cultures Cry1Ab/Ac proteins were released via transgenic Bt-rice in hydroponic cultures, with the amount of 0.30 ng mL −1 , clearly indicative of Cry-protein presence.As assumed, the majority of Cry-proteins were released not from sloughed and damaged root cells but via root exudates, due to the fact that there was no discernible root debris of plants which were grown in hydroponic cultures (Saxena et al. 1999(Saxena et al. , 2002)).In all four paddy soils cultivated with Bt-rice, Cry1Ab/Ac proteins were detected in the rhizosphere after 50 d; their concentrations ranged from 8.55 to 34.9 ng g −1 soil (Table 3) with significant differences among different soils (p< 0.05).However, Cry-proteins were not detected in submerging surface water of soil during cultivation with transgenic rice.Our findings are partly in agreement with that reported by Rui et al. (2005) who noted that two Bt-corn events could release up to 200 and 300 ng Cry-proteins g −1 soil, but contradict the results reported by Wang et al. (2006) who observed that Cry1Ab proteins were not detectable in the rhizosphere under cultivation of Bt-rice cultivar KMD.Actually, the release of Bt-toxins into soil via root exudates has been frequently reported with respect to such transgenic crops as Bt-cotton and Bt-maize while the toxin amount varied widely (Tapp and Stotzky 1998;Saxena et al. 1999;Saxena and Stotzky 2001;Baumgarte and Tebbe 2005).Such variable observations may be partly attributed to the level of protein production in root materials of different events, as both the insertion position of cry genes that cannot be precisely controlled and the promoter which is used in gene manipulation can affect the production of Bt-toxins in Bt-crops (Clark et al. 2005).Cry-proteins are expressed at various levels in different cultivars (Tu et al. 2000;Saxena and Stotzky 2001;Tian et al. 2013).The more Cry-proteins were produced within root tissues of Bt-rice cultivar (TT51-1), the more the proteins would be realised in the rhizosphere during cultivation.
Different paddy soils displayed variable behavior in Cry-protein accumulation during cultivation of transgenic rice; protein amount was three-fold higher in the silty clay soil material than the sandy loam.It may be interpreted as the strong binding of Cry-proteins to clay particles and, therefore, protection of proteins against microbial degradation (Crecchio and Stotzky 1998;Stotzky 2000).As rice were grown in a plant growth chamber under controlled conditions and harvested before maturity, the findings need to be considered with caution.They may differ under field conditions; in addition, one may expect more release and higher accumulation of Cry-proteins in paddy soil ecosystem when Bt-rice reaches the stage of maturity.As reported, Cry-proteins released into soil via root exudates of Bt-rice were analyzed at different growth stages of transgenic rice and protein amount was higher at the stage of panicle initiation than that at the stage of early tillering (Liu et al. 2018).This may be attributed to the levels of Cry-proteins produced within root materials as discussed above.It may also be explained by the fact that Cry-proteins bound to soil become protected against microbial degradation and thus persist in soil for some time.

Conclusions
Soil heterogeneity has greater effects on the level of proteins production compared to other factors, i.e. strain subspecies.Cry-proteins could be adsorbed in paddy soils, among which the clayey soil render stronger affinity.As a result of firm binding, Cry-proteins become protected against microbial degradation and, therefore, can accumulate and persist in the soil ecosystem, imposing impacts on soil microbial communities.Overall, soil characteristics should be taken into consideration when assessing production, release, accumulation, and persistence of Cry-proteins via both Bt-pesticides and transgenic Bt-crops.These factors should be properly evaluated in studies under both laboratory and field conditions.

Figure 1 .
Figure 1.Sorption of Cry1Ab/Ac proteins produced by Bt ssp.kurstaki in differently textured paddy soils.Light grey: proteins adsorbed in the soil; dark grey: proteins extracted from the soil after adsorption; dashed line represents the initial amount of the proteins spiked in soils.Bars: standard error (n = 3).

Table 1 .
Physicochemical parameters of selected paddy soils.Wujiang County, Jiangsu, China; TX (silty texture): Wujia, Tai Xin County, Jiangsu, China.For further information on the origin of the selected paddy soils refer to Eickhorst and Tippkötter 2009 and Liu et al. 2018.ofcry1Ab/Ac genes for Bt-rice and another subspecies thuringiensis (Btt) were chosen for this study.Strains of these two subspecies were obtained from Leibniz Institute DSMZ-German Collection of Microorganisms and Cell, Germany, Btt (DSM-No.6029) and Btk (DSM-No.6102).Strains were cultivated in liquid medium (DSMZ medium 1: peptone 5.0 g, meat extract 3.0 g, MnSO 4 × 10H 2 O, 10.0 mg, sterile Milli-Q water (H 2 O MQ ) 1000 mL, pH 7.0), and set up at 25°C in the dark on a shaker with constant shaking of 70 rpm.

Table 2 .
Mean relative sorption of Cry1Ab/Ac proteins from Bacillus thuringiensis ssp.kurstaki in differently textured paddy soils.
The initial amount of Cry1Ab/Ac proteins added to paddy soils was 0.92 μg per 200 mg of soil; soil was collected 24 h after the addition of Cry-proteins; the results are presented as mean ± SE (n = 3); different letters behind values indicate significant difference between each soil (p < 0.05).
Yoshida et al. 1976 measure the content of Cry-proteins when rice plants reached to a height of 40 cm (i.e.approximately 50 days after transplanting).The results are presented as means ± SE (n = 3); *: described byYoshida et al. 1976; ND: not detectable as concentration of Cry1Ab/Ac proteins was lower than the detection limit of 0.05 ng mL −1 ; different letters behind values indicate significant difference between paddy soils (p <0.05).