Leptolyngbya sp. NIVA-CYA 255, a Promising Candidate for Poly(3-hydroxybutyrate) Production under Mixotrophic Deficiency Conditions

Cyanobacteria are a promising source for the sustainable production of biodegradable bioplastics such as poly(3-hydroxybutyrate) (PHB). The auto-phototrophic biomass formation is based on light and CO2, which is an advantage compared to heterotrophic PHB-producing systems. So far, only a handful of cyanobacterial species suitable for the high-yield synthesis of PHB have been reported. In the present study, the PHB formation, biomass, and elemental composition of Leptolyngbya sp. NIVA-CYA 255 were investigated. Therefore, a three-stage cultivation process was applied, consisting of a growth stage; an N-, P-, and NP-depleted phototrophic stage; and a subsequent mixotrophic deficiency stage, initiated by sodium acetate supplementation. The extracted cyanobacterial PHB was confirmed by FTIR- and GC-MS analyses. Furthermore, the fluorescent dyes LipidGreen2 and Nile red were used for fluorescence-based monitoring and the visualization of PHB. LipidGreen2 was well suited for PHB quantification, while the application of Nile red was limited by fluorescence emission crosstalk with phycocyanin. The highest PHB yields were detected in NP- (325 mg g−1) and N-deficiency (213 mg g−1). The glycogen pool was reduced in all cultures during mixotrophy, while lipid composition was not affected. The highest glycogen yield was formed under N-deficiency (217 mg g−1). Due to the high carbon storage capacity and PHB formation, Leptolyngbya sp. NIVA-CYA 255 is a promising candidate for PHB production. Further work will focus on upscaling to a technical scale and monitoring the formation by LipidGreen2-based fluorometry.


Introduction
Plastics have been an essential part of the modern human lifestyle since their discovery at the beginning of the 20th century. Unfortunately, their de facto non-existent biodegradability is a significant disadvantage [1]. The longevity causes accumulation in the terrestrial and marine ecosphere [2]. According to a recent forecast, 12 billion tons of plastic waste will remain in landfills in the environment by 2050, and approximately 12 billion tons of greenhouse gas carbon dioxide (CO 2 ) will be released [3]. Considering this trend, biodegradable and renewable alternatives to conventional plastics are becoming increasingly important. A promising class of bioplastics are polyhydroxyalkanoates (PHAs). PHAs are lipophilic polyesters that serve as C and energy storage compounds [4]. They are subcategorized according to their monomer length into short-chain-length (scl-PHAs), medium-chain-length (mcl-PHAs), and long-chain-length PHAs (lcl-PHAs) [5]. Scl-PHAs include the most abundant and most studied representative poly(3-hydroxybutyrate) (PHB). The formation of PHB has been reported in all domains of life, including Archaea, Bacteria, and Eukarya [6][7][8][9][10]. The material properties of PHB are similar to those of polypropylene,

Chemicals
Chemicals were purchased in analytical grade either from Carl Roth (Karlsruhe, Germany) or Sigma-Aldrich/Merck (Darmstadt, Germany) unless otherwise specified.

Biomass Concentration and Elemental Composition
Cyanobacterial growth was followed gravimetrically. Therefore, 6-9 mL of cell culture was centrifuged (Thermo Scientific, Waltham, MA, USA) at 10,000× g for 5 min, washed, frozen, and dried by lyophilization (Christ Martin, Osterode, Germany). Biomass concentration was calculated from the ratio of dried biomass to cell culture volume. The elemental composition of the biomass was determined by CHNS composition using a Vario MICRO Cube (Elementar Analysensysteme, Langenselbold, Germany) equipped with a thermal conductivity detector. Protein content was calculated from the N-content as described previously [47].

PHB and Glycogen Content
PHB and glycogen were quantified by HPLC, applying slightly modified conditions as reported elsewhere [48][49][50]. Briefly, 500 µL of 75% H 2 SO 4 or 500 µL of 7.5% H 2 SO 4 were added to the retained dried biomass for PHB or glycogen extraction, respectively, and heated at 95 • C for 60 min. The reaction mixture was diluted with distilled water before measurement. Standards were treated alike and used for calibration. Isocratic separation was performed with a Merck-Hitachi HPLC and ABOA SugarSep column (AppliChrom, Oranienburg, Germany) with 0.007 N H 2 SO 4 as mobile phase and a flow rate of 0.8 mL min −1 at 50 bar. Cisand trans-crotonic acid, as the hydrolysis products of PHB, were identified and quantified with a UV detector at 214 nm. As the hydrolysis product of glycogen, glucose was detected with a Merck-Hitachi, L-7490 refractive index detector.

PHA Extraction
PHA was extracted from 10-20 mg dried biomass with 10 mL chloroform in sealed glass extraction tubes for 60 min at 70 • C. Extracts were separated from residual biomass by hot filtration, precipitated with 5 mL cold ethanol, filtered, washed with 20 mL acetone and 40 mL water, and dried at 60 • C overnight.

FTIR-Analysis
Crude extracted PHA was analyzed for conformity with Fourier-transform infrared spectroscopy (FTIR) using a Bruker Tensor 27 (Bruker Corp., Billerica, MA, USA) equipped with an attenuated total reflection unit (ATR). Interferograms were taken between 550 and Biomolecules 2022, 12, 504 4 of 13 4000 cm −1 . Samples were scanned 30 times at a 4 cm −1 resolution. The resulting pattern of functional groups were compared to the PHB standard.

Monomer Composition
GC-MS was applied for the determination of the PHA composition. Analysis of standard poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHBHV) and extracted PHA-polyester was performed using a previously reported method [8]. Briefly, 5 mg polyester was simultaneously transesterified and hydrolyzed in sealed headspace vials using 4 mL 1,2-dichloroethane, 2 mL 4:1 (vol/vol) propanol HCl-mixture, and 100 µL of 20 g L −1 benzoic acid as internal standard. Propanolysis was carried out at 120 • C for 4 h. Afterward, 4 mL water was added, mixed, and maintained until phase reseparation. Then, 1000 µL of the lower (organic) phase was transferred for GC-MS analysis. Separation occurred on a Stabilwax column (Restek, Bad Homburg, Germany) using He as carrier at a flow rate of 1.44 mL min −1 , and a gradient of 120 • C (3 min), 140 • C at 3 • C min −1 , 230 • C at a heating ramp of 50 • C min −1 , and 240 • C at 10 • C min −1 . A standard curve of PHBHV (12 wt% 3HV, Sigma-Aldrich/Merck, Darmstadt, Germany) was used for the qualification of propionyl-3HB and propionyl-3HV monomers. The resulting chromatograms and mass to charge ratio (m/z) were analyzed to identify monomer composition.

FAME and Lipid Classes
Simultaneous extraction and transesterification for fatty acid methyl ester (FAME) quantification were conducted in sealed headspace vials using 3 mL 3 N methanolic HCl and 4 mL hexane. Subsequently, 3 mL H 2 O was added and homogenized. After phase separation, the upper phase was transferred via a syringe filter into a corresponding vial.
GC-MS analysis was conducted with He as mobile phase at 48 kPa, 250 • C, and a flow rate of 7.7 mL min −1 , and SGE BPX 70 column as stationary phase (Fisher Scientific, Schwerte, Germany) using a QP2010Plus (Shimadzu, Kyoto, Japan).

Fluorescence PHB Quantification with Nile red and LipidGreen2
The fluorescence staining method was carried out based on our own previous studies [52]. Briefly, 1500 µL PBS buffer (137 mM NaCl, 2.7 mM KCl, 10 mM Na 2 HPO 4 , 1.8 mM KH 2 PO 4 , 5 mM EDTA, and 5 mM Tris, pH 7.5) was mixed with 500 µL cell culture in a standard 1 cm cuvette. After 5 min incubation, the fluorescence emission of the PHB-dye-complex was detected with a Perkin Elmer LS45 (Perkin Elmer, Waltham, MA, USA). Slits were set to 10 nm, and a gain of 700 V was applied. The excitation wavelength was set to 440 nm for LipidGreen2 and 525 nm for Nile red. The fluorescence was recorded by an emission scan over 200 nm, starting at 460 nm and 545 nm for LipidGreen2 and Nile red, respectively. Emission intensities at 510 nm (LipidGreen2) and 610 nm (Nile red) were plotted against HPLC-obtained PHB values.

Bioimaging
To observe morphological changes and visualize cyanobacterial PHB granules, incident light and fluorescence microscopic images were captured with Olympus BX41 (Olympus, Tokyo, Japan) equipped with a XC50 camera (Olympus, Tokyo, Japan). The staining was performed as described in Section 2.9. An excitation filter of 460-490 nm was used, and images were analyzed with the cellSens software package (Olympus) as explained by the manufacturer's instructions.

Data Analysis and Statistics
Data plotting, analysis, and statistics were performed with OriginPro (OriginLab Co, Northampton, MA, USA). A t-test at α = 0.1 was used to determine differences between PHB content obtained by either HPLC or fluorimetry. Fluorescence data were normalized to the corresponding analytical values to obtain a normal distribution. All measurements were performed as independent triplicates.

Morphology and Growth
Morphological changes between the different deficiency conditions were observed by microscopy at the end of the phototrophic stage ( Figure S1). Biomass was monitored during the phototroph-deficient and mixotrophic stage. The highest final cell concentration of 1.48 g L −1 and the highest growth rate (64.4 mg L −1 d −1 ) were reached in BG 11 C at the end of the phototrophic stage. Deficient conditions caused lower productivity ( Figure 1). Leptolyngbya sp. reacted to the P limitation with meager growth rates of 6.9 mg L −1 d −1 and 9.4 mg L −1 d −1 for BG 11 P− and BG 11 NP− , respectively. N limitation did not affect growth in the same magnitude. However, Leptolyngbya sp. cultured in BG 11 N− was 44% lower in growth when compared to BG 11 C , resulting in 36.3 mg L −1 d −1 and 1.0 g L −1 final biomass concentration at the end of the phototrophic stage. Supplementation with sodium acetate (mixotrophic stage) caused a decrement in biomass formation in all cultures. The decrease in BG 11 P− was minimal since the growth rate was already low. The most significant decrease was demonstrated by BG 11 C , which lost 48% of biomass during the mixotrophic cultivation phase, resulting in a final biomass concentration of 0.85 g L −1 compared to 0.68 g L −1 of BG 11 N− .

PHB-Formation and Biomass Composition
PHB and glycogen formation were studied during the mixotrophic stage, which was accompanied by a decrease of glycogen in all cultures. The intracellular glycogen was built up in the phototrophic stage and metabolized entirely in BG11 P− and BG11 C (Figure 2).

PHB-Formation and Biomass Composition
PHB and glycogen formation were studied during the mixotrophic stage, which was accompanied by a decrease of glycogen in all cultures. The intracellular glycogen was built up in the phototrophic stage and metabolized entirely in BG 11 P− and BG 11 C (Figure 2). Whereas BG 11 N− -starved cells showed the highest intracellular content of 217 mg g −1 glycogen, BG 11 NP− , BG 11 P− , and BG 11 C exhibited levels of 87, 58, and 7 mg g −1 , respectively.

PHB-Formation and Biomass Composition
PHB and glycogen formation were studied during the mixotrophic stage, which was accompanied by a decrease of glycogen in all cultures. The intracellular glycogen was built up in the phototrophic stage and metabolized entirely in BG11 P− and BG11 C (Figure 2). Whereas BG11 N− -starved cells showed the highest intracellular content of 217 mg g −1 glycogen, BG11 NP− , BG11 P− , and BG11 C exhibited levels of 87, 58, and 7 mg g -1 , respectively. In comparison, cultures grown in BG11 N− and BG11 NP− showed a similar glycogen consumption rate of approximately 80 mg g −1 . However, since the initial glycogen concentration in BG11 N− was higher than that of BG11 NP− , the glycogen was not completely metabolized until the end of the experiment.
Low PHB concentrations were present at the beginning of the mixotrophic stage in all cultures. Both BG11 P− and BG11 C revealed PHB levels lower than 5 mg g −1 . After the supplementation with sodium acetate, PHB content increased to a maximum of 74 mg g −1 after 4 d, and to 25 mg g −1 after 6 d for BG11 P− and BG11 C , respectively. The highest PHB concentration was observed after 22 d in BG11 NP− . Within 6 d of mixotrophy, the PHB content increased remarkably to 325 mg g −1 . Subsequently, the concentration decreased to 213 mg g −1 , as the glycogen concentration decreased. PHB formation in BG11 N− increased during the mixotrophic stage and reached 206 mg g −1 at the end of the experiment, which was comparable to the content of BG11 NP− (213 mg g −1 ).
In addition to PHB and glycogen monitoring, the final biomass was examined for FAME and lipid group composition, since the presence of triacylglycerides (TAG) was In comparison, cultures grown in BG 11 N− and BG 11 NP− showed a similar glycogen consumption rate of approximately 80 mg g −1 . However, since the initial glycogen concentration in BG 11 N− was higher than that of BG 11 NP− , the glycogen was not completely metabolized until the end of the experiment.
Low PHB concentrations were present at the beginning of the mixotrophic stage in all cultures. Both BG 11 P− and BG 11 C revealed PHB levels lower than 5 mg g −1 . After the supplementation with sodium acetate, PHB content increased to a maximum of 74 mg g −1 after 4 d, and to 25 mg g −1 after 6 d for BG 11 P− and BG 11 C , respectively. The highest PHB concentration was observed after 22 d in BG 11 NP− . Within 6 d of mixotrophy, the PHB content increased remarkably to 325 mg g −1 . Subsequently, the concentration decreased to 213 mg g −1 , as the glycogen concentration decreased. PHB formation in BG 11 N− increased during the mixotrophic stage and reached 206 mg g −1 at the end of the experiment, which was comparable to the content of BG 11 NP− (213 mg g −1 ). In addition to PHB and glycogen monitoring, the final biomass was examined for FAME and lipid group composition, since the presence of triacylglycerides (TAG) was previously described as a storage compound in cyanobacteria [37]. Indeed, all depleted cultures demonstrated a doubling in TAG content compared to the control BG 11 C (Table 1). TAG levels of 13.6 mg g −1 (BG 11 N− ), 15.6 mg g −1 (BG 11 P− ), 16.1 mg g −1 (BG 11 NP− ), and 7.2 mg g −1 (BG 11 P− ) were obtained. Polar lipids (PL) content was lowest in BG 11 P− with 7.4 mg g −1 . BG 11 C , BG 11 N− , and BG 11 NP− showed PL levels of 18.4 mg g −1 , 11.4 mg g −1 , and 10.3 mg g −1 , respectively. C16 and C18 are common fatty acids in neutral lipids such as TAG. The sum of C16 and C18 reached 5.8 mg g −1 in the control culture BG 11 C . The macronutrient-depleted cultures showed higher levels of 13.2 mg g −1 (BG 11 N− ), 17.9 mg g −1 (BG 11 P− ), and 13.7 mg g −1 (BG 11 NP− ). The lowest protein content was observed in BG 11 N− (13.5 wt%), resulting in a high C/N ratio of 13.0. BG 11 P− (26.1 wt%), and BG 11 NP− (22.0 wt%) demonstrated higher protein contents, which corresponded to half of the protein content of BG 11 C (44.8 wt%). Therefore, the C/N ratios of BG 11 P− and BG 11 NP− were significantly lower than BG 11 N− , but only minimally higher than BG 11 C .

Correlation of PHB and Fluorescence
The feasibility of fluorescent PHB detection using LipidGreen2 and Nile red staining for quantitative measurements was examined using a correlation of raw fluorescence and HPLC-obtained PHB content in accordance with a previous study [53]. A remarkably high agreement of LipidGreen2 fluorescence to PHB concentration was achieved. The combined coefficient (R 2 ) of BG 11 N− and BG 11 NP− was 0.9883 ( Figure S2). In contrast, Nile red fluorescence and HPLC-based PHB content were only correlated in the BG 11 N− culture (R 2 = 0.9484). For BG 11 NP− , no linear relationship (R 2 = 0.3375) was obtained since phycocyanin fluorescence interfered with the fluorescence of Nile red-stained PHB ( Figure S3). The regression models were used to calculate and compare the PHB levels ( Figure 3A

Characterisation of Extracted Polyester
The characteristic wave numbers of PHB in the FTIR spectrum were presented by the strong carbonyl group (C=O) at 1726 cm −1 and asymmetric C-O-C stretching vibration at 1279 cm −1 , which are typical for ester bondings in PHB polyesters ( Figure 4). The patterns are consistent with earlier reports [54]. Other adsorption bands obtained at 1460 and 1378 cm −1 designated the -CH 2 and -CH 3 groups, respectively. The fingerprint region from 1130 to 979 cm −1 was denoted as the C-O and C-C stretching vibration [31]. With 97.2% agreement, the spectra of the isolated PHB from Leptolyngbya sp. NIVA-CYA 255 matched that of the PHB standard.
FTIR spectra do not accurately indicate the exact PHA type, as it is challenging to distinguish between the vibration pattern of different scl-co-polyesters. Therefore, the monomeric composition was analyzed by GC-MS ( Figure S4). Compared to the scl-co-polyester standard PHBHV, which constituted two peaks at 3.75 min (hydroxybutyrate, HB) and 4.8 min (hydroxyvalerate, HV), only HB could be detected in the extract of Leptolyngbya sp. NIVA-CYA 255. Due to the absence of HV, the formation of homopolymeric PHB could be confirmed.

Characterisation of Extracted Polyester
The characteristic wave numbers of PHB in the FTIR spectrum were presented by the strong carbonyl group (C=O) at 1726 cm −1 and asymmetric C-O-C stretching vibration at 1279 cm −1 , which are typical for ester bondings in PHB polyesters (Figure 4). The patterns are consistent with earlier reports [54]. Other adsorption bands obtained at 1460 and 1378 cm −1 designated the -CH2 and -CH3 groups, respectively. The fingerprint region from 1130 to 979 cm −1 was denoted as the C-O and C-C stretching vibration [31]. With 97.2% agreement, the spectra of the isolated PHB from Leptolyngbya sp. NIVA-CYA 255 matched that of the PHB standard. FTIR spectra do not accurately indicate the exact PHA type, as it is challenging to distinguish between the vibration pattern of different scl-co-polyesters. Therefore, the monomeric composition was analyzed by GC-MS ( Figure S4). Compared to the scl-copolyester standard PHBHV, which constituted two peaks at 3.75 min (hydroxybutyrate, HB) and 4.8 min (hydroxyvalerate, HV), only HB could be detected in the extract of

Characterisation of Extracted Polyester
The characteristic wave numbers of PHB in the FTIR spectrum were presented by the strong carbonyl group (C=O) at 1726 cm −1 and asymmetric C-O-C stretching vibration at 1279 cm −1 , which are typical for ester bondings in PHB polyesters ( Figure 4). The patterns are consistent with earlier reports [54]. Other adsorption bands obtained at 1460 and 1378 cm −1 designated the -CH2 and -CH3 groups, respectively. The fingerprint region from 1130 to 979 cm −1 was denoted as the C-O and C-C stretching vibration [31]. With 97.2% agreement, the spectra of the isolated PHB from Leptolyngbya sp. NIVA-CYA 255 matched that of the PHB standard. FTIR spectra do not accurately indicate the exact PHA type, as it is challenging to distinguish between the vibration pattern of different scl-co-polyesters. Therefore, the monomeric composition was analyzed by GC-MS ( Figure S4). Compared to the scl-copolyester standard PHBHV, which constituted two peaks at 3.75 min (hydroxybutyrate, HB) and 4.8 min (hydroxyvalerate, HV), only HB could be detected in the extract of

Discussion
Cyanobacteria offer a promising and sustainable alternative to produce biomolecules, such as the polyester PHB. Currently, industrial production of PHB with cyanobacteria is not lucrative due to the production costs and long cultivation time. An initial production analysis assessed the current price of cyanobacterial PHB to be approximately EUR 24 kg −1 , which is 2-5 times higher than the present synthesis with heterotrophic bacteria [55,56]. To reduce PHB production costs, the cultivation conditions were envisaged to be optimized, and new species was studied. In preliminary experiments, Leptolyngbya sp. NIVA-CYA 255 was a suitable candidate for cultivation on a technical scale as the cultivation was robust and not sensitive to temperature fluctuations. Therefore, PHB synthesis and biomass composition were investigated. Studies indicated a significant contribution of PHB to cell dormancy in addition to its function as a storage lipid [57]. It has been demonstrated that PHB is involved in stress survival during environmental changes [58][59][60][61]. PHB is formed from the intracellular glycogen pool built up during the photoautotrophic stage, especially under N-deficiency [62,63]. The results of our study confirmed these findings. The highest glycogen pool was formed during N-deficiency and converted into PHB in the mixotrophic phase. Significantly lower glycogen concentrations were achieved in BG 11 P− and BG 11 C ( Figure 2). These results are consistent with earlier reports, where N-deficiency led to a higher total amount of C storage compounds (PHB, glycogen, lipids) in Synechocystis sp. PCC 6714 [64]. Consequently, a lack of P during the phototrophic stage led to a decreased glycogen productivity and thus reduced the formation of PHB. P-deficiency also had a substantial negative impact on growth, which was almost completely stopped in the P-deficient cultures BG 11 NP− and BG 11 P− (Figure 1). Nevertheless, the highest PHB content was obtained in BG 11 NP− ( Figure 2B). Acetate accelerated the PHB formation and ensured PHB increment within a short time. Although the PHB content of 20.6 wt% in N-deficient culture was lower than that of the NP-deficient culture (32.3 wt%), the higher biomass concentration resulted in a comparable PHB concentration of 141.5 mg L −1 compared to 167.3 mg L −1 (NP-deficiency). Therefore, N-limitation is beneficial for large-scale applications, since (i) biomass productivity does not stop during the N-deficient stage, (ii) high glycogen productivity ( Figure 2) was observed, and (iii) the dormant state (chlorosis) supports PHB production and cell survival. A prolonged cultivation period might result in even higher yields. High PHB concentrations were obtained in similar studies ( Table 2). Synechococcus sp. MA19 showed the highest content of 55 wt% under P-deficient conditions (Table 2). However, Synechococcus sp. MA19 is a thermophile, resulting in high cultivation temperatures of 50 • C. Due to the additional energy input, higher production costs have to be considered. Considering this fact, Leptolyngbya sp. mesophilic cultivation temperatures of 26 • C is advantageous for large-scale applications.
Although PHB is described as a typical storage lipid in cyanobacteria, the formation of lipid bodies has also been reported [62]. Therefore, lipid classes, including TAG as the main component of lipid bodies, were analyzed. Due to the low intracellular content (1.3-1.6 wt%), TAG's function can be understood as a general stress response rather than a storage compound.
The highest C/N ratio was obtained with BG 11 N− . This value represents a cumulative PHB and glycogen content parameter, and can be considered as a general C storage formation benchmark. However, C/N analysis must be performed from the dried biomass, which is time-consuming and not applicable for rapid monitoring approaches. Therefore, the eligibility of fluorometric monitoring was also studied. Unlike heterotrophic bacteria, cyanobacteria are subjected to major morphological changes under different stress conditions, including the formation or degradation of pigments. For fluorometric detection, interferences of pigment fluorescence should be considered. Nile red PHB emission interfered with phycobilin fluorescence in BG 11 NP− and therefore showed no correlation ( Figure S2).
LipidGreen2 revealed better applicability for PHB detection since the fluorescence emission maximum of 510 nm was outside the phycobilin and chlorophyll fluorescence emission. This resulted in high degrees of correlations for both BG 11 N− and BG 11 NP− cultures. To the best of our knowledge, this is the first report using cuvette-based fluorimetry coupled with LipidGreen2 for PHB monitoring in cyanobacteria. Due to the advantageous properties of the dye, future studies will focus on the process optimization and up-scaling of Leptolyngbya NIVA-CYA 255 using LipidGreen2 for the fast monitoring of intracellular PHB contents.

Conclusions
Leptolyngbya sp. NIVA-CYA 255 is a promising host for cyanobacterial PHB production. This study investigated PHB formation in a three-stage cultivation process, containing a growing stage, a macronutrient-depleted phototrophic stage, and a subsequent mixotrophic stage. Cultivation in N-and P-deficiency supplemented with acetate resulted in an intracellular concentration of 32.3 wt% PHB. At the end of the experiment, BG  cultures demonstrated comparable PHB concentrations. Since BG 11 N− showed the highest C storage capacities (PHB and glycogen), N-depletion seems to be the favorite strategy for PHB production in Leptolyngbya sp. NIVA-CYA 255 that can be easily monitored using LipidGreen2-fluorescence.
Funding: This research was funded by ERDF (European Regional Development Fund) and the state of Saxony-Anhalt (Germany), with the project DIGIPOL (numbers/2018/11/95487).

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable. Data Availability Statement: Datasets were generated and analyzed during the study.