One-step synthesis and characterization of CNGs
Glass vials containing DAO (20 mg), DEX (100 mg), or a mixture of DAO and DEX with mass ratios (DAO:DEX) of 1.0:0.5 (20:10 mg), 1.0:1.5 (20:30 mg), or 1.0:5.0 (20:100 mg) were separately heated at 180 ℃ for 3 h to obtain CNGs. After heating, DAO formed a sticky brown film with poor water solubility (Additional file 1: Fig. S1) as oxidation of the amine groups and polymerization of DAO following carbonization during heating led to the formation of a nitrogen-doped polymeric graphene film with poor hydrophilicity. In contrast, after heating and dissolution in water, DEX yielded a transparent product, probably because of the lower degree of carbonization. The pyrolytic products obtained from the various mixtures of DAO and DEX were brown in color and exhibited high aqueous solubility. Moreover, the solubility of the product increased with an increasing ratio of DEX, owing to its intrinsic high hydrophilicity.
We next characterized the size of CNGs. Transmission electron microscopy (TEM) revealed that, after heating, the mixtures of DAO and DEX formed nanogel structures of 100−750 nm in diameter and polymeric frameworks on the surface of the nanogels (Fig. 2A). For simplicity, these as-prepared nanogels are denoted DAO/DEX0.5-CNGs, DAO/DEX1.5-CNGs, and DAO/DEX5.0-CNGs, based on mass ratios (DAO:DEX) of 1.0:0.5, 1.0:1.5, and 1.0:5.0, respectively, as used in the mixture. The as-prepared DAO/DEX5.0-CNGs exhibited the highest product yield (ca. 71%), mainly owing to high water solubility (Additional file 1: Table S1). Furthermore, cryogenic electron microscopy (cryo-EM) and scanning electron microscopy (SEM) revealed the presence of spherical CNGs (Fig. 2B–D). Moreover, the lattice structure in the interior of CNGs is discernible in the high-resolution TEM images (Fig. 2A, inset) and by their distinctive fluorescence properties (Fig. 2E–G) that confirm the formation of ultrasmall few-layered graphene quantum dots within as-formed CNGs [25]. The hydrodynamic diameter/ζ-potential of the as-prepared DAO/DEX0.5-CNGs, DAO/DEX1.5-CNGs, and DAO/DEX5.0-CNGs determined by dynamic light scattering were 209 nm/13.2 mV, 322 nm/6.1 mV, and 580 nm/5.8 mV, respectively (Fig. 2H). The higher ratio of polymeric and neutrally charged DEX in DAO/DEX-CNGs resulted in a larger size and smaller charge.
Next, we characterized the spectral properties of CNGs. The UV-visible absorption spectra of all as-prepared CNGs showed broad bands around 230−450 nm (Additional file 1: Fig. S2A). The absorption band around 250−290 nm was attributed to the π→π* transition of aromatic/alkenyl C = C bonds or C = N bonds, supporting the formation of graphitic carbon clusters, whereas the shoulder band at around 300 − 380 nm was probably due to n→π* transitions of C = O and C = N bonds [26]. All DAO/DEX-CNGs displayed similar emission maxima profiles around 460 nm when excited at 365 nm (Additional file 1: Fig. S2B). Additionally, the DAO/DEX-CNGs exhibited excitation wavelength-dependent fluorescence emission properties (Fig. 2F–H and Fig. S2C), mainly owing to the formation of different sizes of polycyclic aromatic or graphene clusters [27]. However, partially carbonized CNGs retained many emissive traps, which lowered their quantum yield (QY) to < 1% (at excitation/emission maxima of 365 and 460 nm; in comparison with the quinine sulfate reference) (Additional file 1: Table S1) [26].
Synthetic mechanism of CNGs
Some characteristics of the DAO/DEX-CNGs resembled those of small-sized carbon dots (CDs; < 10 nm), such as their excitation wavelength-dependent fluorescence emission. However, the large size of the DAO/DEX-CNGs (> 100 nm) prevented their classification as typical CDs. The CNGs comprised a cross-linked polymer (gel) structure with abundant functional groups and were embedded with graphene-like CQDs. X-ray photoelectron spectroscopy (XPS; Additional file 1: Fig. S3) and Fourier transform infrared spectroscopy (FT-IR; Additional file 1: Fig. S4 and Table S2) demonstrated the presence of diverse functional groups in the CNGs, including O − H, N − H, C − O, C = O, C − N, C = C, and C = N. Furthermore, several functional groups from the precursors DAO and DEX (e.g., O − H, N − H, C − O, and C − N) were preserved. The presence of C = C and C = N in the CNG suggests the formation of aromatic rings and that the nitrogen atoms were incorporated (doped) as pyridinyl, pyrrolyl, and amide moieties in the heterocyclic ring systems [28].
Figure 1A shows the proposed mechanism of formation of the CNGs. Time-course TEM measurements revealed that large irregular gel-like structures formed within 5 min through the crosslinking reaction of DAO and DEX at 180 ℃ (Additional file 1: Fig. S5). During the heating process, the primary amino group within DAO at both terminals acted as a crosslinking agent for DEX polysaccharides to form inter- and intra-crosslinking polymers (or supramolecules) with micrometer sizes through dehydration. The dehydration reaction between the aldehyde groups of the DEX and the amino group of DAO may have resulted in the formation of a Schiff base, followed by rearrangement to form the Amadori product [29]. Then, during the 5–10 min of the pyrolysis reaction, the supramolecular structures partially decomposed into smaller fragments. Further heating produced the smaller and semi-spherical nanostructures through the condensation reaction while in situ partial carbonization occurred. During the subsequent period of heating (1−3 h), spherical nano-colloidal structures formed as a result of further condensation and carbonization. We observed that the as-formed CNGs still featured polymeric frame structures on their surfaces (Additional file 1: Fig. S5). Large-sized carbon particles (> 1 µm) with poor aqueous solubility (< 10 µg mL−1) were obtained upon overheating (⁓ 4 h), probably owing to extreme pyrolysis; thus, we limited heating to 3 h. We also compared DEX mixed with different linear alkyl diamines (NH2(CH2)nNH2; n = 2, 4, 6, 8, 10, 12) to prepare CNGs in our current work. It is interesting to note that CNGs were not formed with NH2(CH2)nNH2 (n < 6) and the size of the CNGs was controlled by changing the lengths of the alkyl diamines (n ≥ 6) as well as the molecular weight of DEX.
Antibacterial activities of DAO/DEX-CNGs
We first tested the antibacterial potency of DAO/DEX-CNGs against four strains of non-resistant bacteria (Escherichia coli, S. aureus, Pseudomonas aeruginosa, and Salmonella enteritidis) and one strain of multidrug-resistant bacteria (methicillin-resistant S. aureus, MRSA). Our previously reported antibacterial CQDs prepared from spermidine (Spd-CQDs; diameter of ca. 6 nm) with a high zeta potential (ζ = +45 mV) exhibited effective antibacterial ability only in low ionic strength solution (5 mM sodium phosphate, pH 7.4) (Additional file 1: Fig. S6) [21]. In contrast, the DAO/DEX5.0-CNGs reported herein displayed potent antibacterial activity on all tested bacteria, even in high ionic strength solution, such as phosphate-buffered saline (PBS). Unlike DAO/DEX-CNGs, the Spd-CQDs tend to aggregate and then precipitate in PBS solution owing to electrostatic screening. Among the tested mass ratios, the DAO/DEX5.0-CNGs displayed superior bacteriostatic activity to the other DAO/DEX-CNGs in PBS solution. We attribute this to the polymeric features of the large-sized DAO/DEX5.0-CNGs exerting strong interaction effects with bacteria despite charge screening in the high ionic strength solution. Indeed, we observed that the DAO/DEX5.0-CNGs featured a Velcro-like property, whereby they rapidly bound to E. coli and S. aureus membranes after only 1 min of incubation in PBS solution (Additional file 1: Fig. S7).
After demonstrating their superior antibacterial activity toward common pathogenic bacteria, we explored the antimicrobial action of DAO/DEX5.0-CNGs toward marine Vibrio. Figure 3A displays the results of colony formation assays for V. parahaemolyticus that were untreated or treated with DAO/DEX-CNGs or Spd-CQDs in artificial seawater (480 mM NaCl, 27 mM MgCl2, 30 mM MgSO4, 10 mM CaCl2, 10 mM KCl, and 2.0 mM NaHCO3). The DAO/DEX5.0-CNG-treated group showed > 95% inhibition of the bacteria. The minimal inhibitory concentration (MIC) of DAO/DEX5.0-CNGs (9−19 µg mL−1) for the tested Vibrio strains was much lower than that of DAO/DEX0.5-CNGs (95−160 µg mL−1), DAO/DEX1.5-CNGs (25−87 µg mL−1), and Spd-CQDs (> 500 µg mL−1) (Fig. 3B). Furthermore, the precursors (i.e., DAO and DEX) exhibited negligible antibacterial activity against representative bacteria (MIC > 1.0 mg mL−1) compared with DAO/DEX5.0-CNGs. These results suggest that these DAO/DEX5.0-CNGs effectively eradicate Vibrio bacteria.
Microscopic images revealed that the DAO/DEX5.0-CNGs, at ~ 500 nm, strongly bound to V. parahaemolyticus and led to a higher extent of bacterial aggregation than Spd-CQDs and other DAO/DEX-CNGs (Fig. 4A). Additionally, TEM imaging showed that the as-prepared DAO/DEX5.0-CNGs easily deposited onto the bacteria and damaged the integrity of bacterial membranes, thereby causing leakage of the cytoplasm (Fig. 4B).
In the 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) assay, the fluorescent product DCF (λemmax ≈ 530 nm), which is produced through the sequential reaction of DCFH-DA with cellular esterase and intracellular reactive oxygen species (ROS), was observed in DAO/DEX5.0-CNGs-treated V. parahaemolyticus (Additional file 1: Fig. S8). ROS generation in DAO/DEX5.0-CNG (10 µg mL−1)-treated V. parahaemolyticus was higher than that in untreated V. parahaemolyticus and V. parahaemolyticus treated with H2O2 (10 µg mL−1). The catalytic activity of DAO/DEX-CNGs that enables ROS generation arises mainly owing to specific ligands (i.e., C = O and –C–O–C) on the edges of the graphene-like structure [30]. Additionally, the fast electron transportation characteristics of the embedded graphene facilitates ROS generation [28, 30]. Furthermore, nitrogen-doping into the graphene structure as quaternary N and pyridinic N may also contribute to the catalytic formation of ROS [28, 31, 33]. This nitrogen-doping can boost the spin density and charge distribution of carbon atoms, thereby increasing the density of catalytically active centers on the graphene surfaces [28].
Biocompatibility of DAO/DEX5.0-CNGs
In the alamarBlue assay, DAO/DEX5.0-CNGs did not show significant cytotoxicity toward any tested cell line up to 100 µg mL− 1, which was > 5-fold higher than the MIC values of the bacteria (Additional file 1: Fig. S9A). Moreover, DAO/DEX5.0-CNGs exhibited negligible hemolysis up to 100 µg mL− 1 (Additional file 1: Fig. S9B).
After 1 week of feeding with commercial feed or DAO/DEX5.0-CNGs-mixed feed (1−100 µg g−1), even at the highest dose of DAO/DEX5.0-CNGs (100 µg g−1), the shrimp survival rate remained the same as that in the group fed commercial feed (Additional file 1: Fig. S10A). Thus, our results indicate that DAO/DEX5.0-CNGs additives do not cause severe toxicity in shrimp at up to 100 µg g−1. Histological results of the hepatopancreas from DAO/DEX5.0-CNGs-fed shrimp (up to 100 µg g−1) were the same as those from the control group. The hepatopancreas tissue slice samples exhibited a well-organized glandular tubular structure (T), including a star-shaped tubule lumen (Lum) lining with a single layer of normal epithelial cells and Blasenzellen cells (B-cells) with large apical secretory granules (Additional file 1: Fig. S10B). Thus, we have demonstrated that DAO/DEX5.0-CNGs is a highly biocompatible material for both human cells and shrimp, implying that DAO/DEX5.0-CNGs is a safe feed additive for shrimp aquaculture.
DAO/DEX5.0-CNGs protect shrimp from Vibrio infection
After seven days of feeding with commercial feed or the feed mixed with 10 or 100 µg g−1 DAO/DEX5.0-CNGs and V. parahaemolyticus challenge, infected shrimp displayed typical pathological signs, including lethargy, empty gut, paleness, and aqueous hepatopancreas, in the control group (i.e., feed without DAO/DEX5.0-CNGs additive) [31]. In contrast, only mild signs were observed in shrimp fed DAO/DEX5.0-CNGs. Additionally, the survival rates on day 7 post-challenge were significantly improved (p < 0.001) in the DAO/DEX5.0-CNGs fed groups compared with those in the control group (Fig. 5A and 5B). Specifically, the survival rate increased from 26% when fed with commercial feed to 73% when the feed was mixed with 100 µg g−1 DAO/DEX5.0-CNGs.
Pathogenic Vibrio easily colonizes the digestive system, where it starts to cause diseases [33]. We hypothesized that the protective effects of antibacterial DAO/DEX5.0-CNGs may arise via the suppression of the colonization of V. parahaemolyticus in the intestinal organs. Therefore, we collected midgut tissues from challenged shrimp to evaluate the in vivo antimicrobial potency of DAO/DEX5.0-CNGs. After sequential tissue homogenization, serial dilutions, and plating, the total number of Vibrio colonies in each sample was counted on thiosulfate citrate bile salts sucrose (TCBS) agar plates. In the control group, wherein shrimp were fed with commercial feed without Vibrio challenge, Vibrio was detected in the digestive canal (Fig. 5C(i) and 5D(i)); however, the number of Vibrio colonies increased significantly after Vibrio challenge (Fig. 5C(ii) and 5D(ii)). In shrimp that ingested DAO/DEX5.0-CNGs, the number of Vibrio colonies was even lower than that in the control group when 100 µg g−1 DAO/DEX5.0-CNGs was applied (Fig. 5C(iv) and 5D(iv)). Our results indicate that DAO/DEX5.0-CNGs exert a superior antibacterial capability, compared to previous antibacterial compounds which exert robust antibacterial effects, even in the gut of shrimp, thereby reducing Vibrio infection.
Although the number of Vibrio colonies in the midgut of whiteleg shrimp fed 10 µg g−1 DAO/DEX5.0-CNGs was higher than that in shrimp fed 100 µg g−1 DAO/DEX5.0-CNGs (Fig. 5C and 5D), the survival rates were similar between groups (Fig. 5B). We, therefore, analyzed hepatopancreas tissue using pathological methods. In the hepatopancreas tissue slides of shrimp challenged with V. parahaemolyticus, we observed typical pathological features of AHPND, including pigment loss in the connective tissue capsule, irregular shape of tubal lumen with epithelial cell detachment, loss of secretory granule in B-cells, hemocytic nodules, and hemocytic infiltration (HI) (Fig. 5E) [34]. When the shrimp were fed with DAO/DEX5.0-CNGs additives, their hepatopancreas exhibited relatively mild signs of AHPND. Moreover, the lower dose of DAO/DEX5.0-CNGs (10 µg g−1) still reduced the histopathological damage and HI compared with the control group (without DAO/DEX5.0-CNGs). These results reveal that the DAO/DEX5.0-CNGs may have functions, other than bactericidal effects, that protect shrimp from AHPND.
Immunomodulation of DAO/DEX5.0-CNGs in shrimps
Our previous work demonstrated that Spd-CQDs stimulate several immune-related genes, including lysozyme, anti-lipopolysaccharide factor, and cytosolic manganese superoxide dismutase genes, that prevent WSSV infection [24]. Many studies have also reported that nanomaterials regulate the immune systems of animals [35−37]. Therefore, we further evaluated the shrimp immune system after treatment with DAO/DEX5.0-CNGs to elicit their contribution toward immune protection. After feeding shrimp in the presence or absence of DAO/DEX5.0-CNGs (10 or 100 µg g−1) for three days, their hemocytes were collected, and the expression levels of immune-related genes including β-1,3-glucan-binding protein (LGBP), anti-lipopolysaccharide factor (ALF), lysozyme (LYZ), and cytosolic manganese superoxide dismutase (cytMnSOD), were analyzed by real-time quantitative reverse-transcription PCR (qRT-PCR). These genes represent either important sensors or effecters in the shrimp immune system against various bacterial pathogens [38−41]. Feeding shrimp DAO/DEX5.0-CNGs did not increase the expression of LGBP, ALF, or LYZ, and that of cytMnSOD was slightly decreased (Additional file 1: Fig. S11). Unlike the other genes, cytMnSOD is not only a responder when the immune system is challenged but is also sensitive toward exposure to toxic materials [41]. We suggest, therefore, that DAO/DEX5.0-CNGs does not act as an immune stimulator in the shrimp. Instead, the decrease in cytMnSOD expression might be explained by the antibacterial activity of DAO/DEX5.0-CNGs, which may reduce the gut microbial population and their toxin production.
In the hemolymph, the expression of LGBP, ALF, LYZ, and cytMnSOD increased sharply within 24 h after V. parahaemolyticus challenge in shrimp fed normal feed (Additional file 1: Fig. S11). Such an acute and severe bacterial infection might cause systemic immune failure, also known as sepsis. The overstimulated immune system and oxidative stress during sepsis usually damage tissues and cause mortality, often to the point that even extensive antibiotic administration is unable to facilitate recovery [42]. Although the pathological mechanism of shrimp sepsis remains unclear, we did observe similar signs, such as overstimulation of the immune system in severe V. parahaemolyticus infection. The expression levels of the four immune-related genes were significantly reduced in shrimp fed DAO/DEX5.0-CNGs; those of LGBP and cytMnSOD even returned to that of the unchallenged condition (Additional file 1: Fig. S11). In addition to its antibacterial activity, immune inhibition by DAO/DEX5.0-CNGs might contribute to the higher survival rate of shrimp after acute and severe V. parahaemolyticus infection.
DAO/DEX5.0-CNGs sponge PirAB toxin
Although DAO/DEX5.0-CNGs reduced colonization by V. parahaemolyticus in the intestine of whiteleg shrimp, the PirAB toxin remains harmful to the hepatopancreas and lethal to the shrimp. Previous work has shown that the mortality of PirAB toxin-challenged shrimp increases with ALF knockdown [43]. ALF is a short antimicrobial polypeptide that binds to lipopolysaccharides and peptidoglycans from V. parahaemolyticus to alleviate AHPND [44]. Molecular modeling and docking studies have revealed that the lipopolysaccharide-binding sites of ALF also interact with PirB, thereby reducing PirAB toxicity [43]. The expression of ALF was significantly reduced in shrimp fed DAO/DEX5.0-CNGs followed by V. parahaemolyticus challenge, compared to the controls. Thus, DAO/DEX5.0-CNGs might act as a sponge to absorb the PirAB toxin through their polymeric nature and compensate for the need for ALF. To test this, we prepared recombinant PirA (0.1 mg) or PirB (0.1 mg) toxins and incubated them with different amounts of DAO/DEX5.0-CNGs (0.5−10 mg). After removing absorbed PirA or PirB by centrifugation, the supernatants were analyzed by western blotting. The DAO/DEX5.0-CNGs (5.0 mg) adsorbed > 80% of PirA (0.1 mg) and PirB (0.1 mg) toxin (Additional file 1: Fig. S12). Based on these observations, we believe that highly efficient trapping of lethal PirA/B toxins by DAO/DEX5.0-CNGs also contributes toward their excellent protective effects for whiteleg shrimp against AHPND.
Effects of DAO/DEX5.0-CNGs on the microbiota of shrimp
It is now recognized that the gut microbiota plays indispensable roles in several key physiological functions of shrimp [45]. The intestinal bacterial composition and their metabolites may highly affect nutrient acquisition and susceptibility to pathogenesis in shrimp [46]. Antibacterial agents that impact the gut microbiota may also affect shrimp. Therefore, different doses of DAO/DEX5.0-CNGs (0, 10, and 100 µg g−1) were fed to shrimp for seven days, after which their gut microbiota was analyzed by a standard Illumina 16S rRNA gene amplicon sequencing method. The prokaryotic populations in each sample were analyzed using tag-encoded high-throughput sequencing of the V3–V4 region of 16S rRNA gene amplicons. In total, 1,277,642 high-quality sequences were obtained, with an average number of 70,980 reads (ranging from 28,131 to 176,016 reads per sample). The read dataset of each library was randomly subsampled to ensure an even sampling depth (28,131 reads per library). In total, 4,432 operational taxonomic units (OTUs) were obtained; the number of OTUs detected in each sample was 138–316, with an average of 246.
Similarly to previous reports [47, 48], the relative abundance of OTUs at the phylum level showed that the microbiota of control diet whiteleg shrimp was mainly composed of Proteobacteria (54% ± 3%) and Bacteroidetes (36% ± 11%) (Additional file 1: Fig. S13; S1 group). After four days of feeding with DAO/DEX5.0-CNGs additives (10 µg g−1), Proteobacteria (39% ± 3%) and Bacteroidetes (51% ± 2%) remained the dominant phyla in the shrimp intestines (S2 group). Upon feeding with a higher dose of DAO/DEX5.0-CNGs (100 µg g−1), the relative abundance of Proteobacteria (43% ± 4%) and Bacteroidetes (41% ± 8%) was slightly changed, but they still dominated.
Next, we analyzed samples from each group individually. We selected species (i.e., OTUs) that had an abundance greater than 1% and appeared in at least one of the samples. The top 30 species were then selected according to their ranking via their p-values from the two-tail Wilcoxon rank-sum test or Kruskal-Wallis test [49]. There were some individual differences in the top 30 species, even within the same experimental groups (Fig. 6; S1-1, S1-2, and S1-3). However, feeding with DAO/DEX5.0-CNGs resulted in no significant difference in the abundance of the dominant bacteria, including Tenaclbaculum, Alglbacter, Motilimonas, and Ruegeria, compared with that in the control diet group. It is also worth noting that Vibrio did not reach the 1% abundance threshold in three successive measurements of the S2 group. However, in S3-1 and S3-3, which were fed with higher concentrations of DAO/DEX5.0-CNGs, Vibrio still remained at 1.22% and 4.84% abundance, respectively.
Using t-distributed stochastic neighbor embedding (t-SNE) analysis to visualize the differences between each data set, samples in group S1–S3 were closely clustered (Additional file 1: Fig. S14). Certain individual variations, such as between S1-1 and S1-3, were even higher within the same group than in the intergroup comparison. Overall, DAO/DEX5.0-CNGs did not exhibit a strong impact on the dominant species in the shrimp gut microflora, including Vibrio. This is consistent with our observations that DAO/DEX5.0-CNGs did not result in a significant difference in antimicrobial activities of various bacteria. (Fig. 3B and Additional file 1: Fig. S6).
Previous reports have shown that feeding with an antibiotic, such as ciprofloxacin or sulfonamide, causes a significant decrease (> 50%) in shrimp intestinal OTUs, and the Shannon index analysis indicated that using antibiotics also decreases the diversity of the intestinal microflora [48]. However, in our case, there was no statistical difference in OTUs nor in the Shannon index in each group, including the S3 group which had been fed with the highest dosage of DAO/DEX5.0-CNGs (100 µg g−1) (Additional file 1: Fig. S15). Because of the non-selective bactericidal effect of DAO/DEX5.0-CNGs, the levels of major species and minor species were evenly decreased and preserved. Compared with antibiotics, it might be an add up for employing as a feed additive in shrimp.
When shrimp were challenged by AHPND-causing V. parahaemolyticus, the intestinal microflora was greatly influenced both at the phylum (Additional file 1: Fig. S13A) or the genus level (Additional file 1: Fig. S13B), even in the DAO/DEX5.0-CNGs fed groups. In general, the abundance of Proteobacteria increased significantly in all challenged groups, particularly those from the genus Vibrio. The t-SNE analysis also showed that the bacterial composition after Vibrio challenge was very different from the healthy groups (Additional file 1: Fig. S14). This phenomenon during an AHPND outbreak is named dysbiosis of the gut [33]. However, even in the DAO/DEX5.0-CNGs treated groups, the intestinal microflora could not be restored to the state of the healthy groups. Owing to their non-selective nature, DAO/DEX5.0-CNGs could only reduce the absolute number of pathogens (Fig. 5B). A possible explanation is that our experimental shrimp were kept in a water tank containing a high dose of V. parahaemolyticus, even after challenging. Pathogenic Vibrio in the culture environment may continue to affect the shrimp gut ecosystem via the oral or anal routes, and in turn, Vibrio remained dominant in the intestinal microflora. Additionally, even if Vibrio had been eliminated by DAO/DEX5.0-CNGs, its DNA may have still been included in the OTU count [50]. These are in agreement with the fact that after feeding DAO/DEX5.0-CNGs, the number of viable Vibrio colonies did decrease significantly (Fig. 5C) and the survival rate of whiteleg shrimp was also greatly improved.