Introduction

Pathogenic infections are a major concern in the aquaculture industry leading to huge economic losses, bacterial infections like vibriosis, edwardsiellosis, and aeromonasis being a major contributor. Numerous Vibrio species such as Vibrio harveyi and Vibrio parahaemolyticus are accounted for causing vibriosis, the most prevalent infection in halophilic environment affecting wide range of organisms such as teleosts and marine invertebrates. Infection in marine fishes such as crab, shrimp, shellfish, lobster, and oysters is reported to show acute hepatopancreatic necrosis syndrome and even death (Kim and Bang 2008). Edwardsiellosis, caused by Edwardsiella tarda, a Gram-negative and facultative anaerobic bacterium, has been reported worldwide in economically important fresh water, brackish water, and marine fish species (Srinivasa Rao et al. 2001). The pathologic symptoms of enterohemorrhagic septicemia caused by E. tarda are commonly observed in different species of diseased fish (Benli and Yildiz 2004; Abraham et al. 2015). Abscesses, inflammation, and cutaneous ulcerations have been observed in channel catfish and striped mullet infected with E. tarda (Park et al. 2012). Yet another bacterial genus that results in severe economic losses to aquaculture industry is Aeromonas. The bacterium is inhabitant of the gastrointestinal tract and causes motile Aeromonas septicemia (MAS), hemorrhagic septicemia, ulcerative or red-sore disease, especially under stress conditions (Piotrowska and Popowska 2014). All the three bacterial infections spread very fast in contained culture and result in heavy mortality of fish in culture.

Thus, these bacterial infections afflict the aquaculture industry and are generally treated with antibiotics. However, antibiotic treatment is undesirable for food fishes as it leads to accumulation of the drug in fish and subsequently in the food chain. Also, bacteria often develop drug resistance upon continuous exposure to the antibiotics. Lack of any current commercial subunit vaccine further hinders successful treatment and eradication of these bacterial pathogens (Harikrishnan et al. 2003). Alternative strategies that would modulate the innate immunity of the fish would be an effective way to control infection. First line of defense plays an essential role in defending the host from a bacterial infection. Although the innate immunity is not specific against the attacking pathogen, however, it is crucially relevant in lower vertebrates and teleost fishes like L. rohita, wherein continued exposure to the pathogens in water poses an imminent threat to their survival. Antimicrobial peptides/proteins are produced by almost all species of life as an essential part of innate immunity. These possess broad-spectrum bactericidal activity and low propensity for resistance development. However, purification of these peptides from natural sources is time consuming and not economically viable which would also be accompanied by the exhaustion of the native source. Obtaining antimicrobial peptides using synthetic chemistry is a cumbersome process and has a limitation of length for accurate synthesis. These barriers can be overcome by the new generation of recombinant antimicrobial peptides which is cost effective and gives rise to more homogenous and consistent preparations (Marr et al. 2006).

The lipoproteins present in the plasma membrane are complex moieties of various lipid and protein components. These are classified into four major groups on the basis of lipid and apolipoproteins ratio, namely chylomicrons, very low-density lipoprotein, low-density lipoprotein, and high-density lipoprotein (Kondo et al. 2005). In vertebrates, exchangeable apolipoproteins comprise ApoA, ApoC, and ApoE characterized by low molecular weight and repeated amphipathic alpha helix regions for reversible lipoprotein binding, whereas non-exchangeable apolipoproteins comprise ApoB characterized by amphiphathic beta strand domain for irreversible lipid binding (Segrest et al. 1998; Saito et al. 2004). In addition, atypical apolipoproteins including ApoD, ApoF, ApoH, ApoI, ApoM, ApoN, and ApoO classes also play significant role in lipid metabolism (Page et al. 2001).

Apolipoprotein A-I/ApoA-I is the major component of high-density lipoproteins (HDL) also called “good cholesterol” (Oram 2003). It belongs to flexible exchangeable superfamily so it can exist in lipid-free, lipid-rich, and lipid-poor states (Beck et al. 2013). It has central role in reverse cholesterol transport pathway, activator of lecithin:cholesterol acyl transferase, and therefore exhibits atheroprotective activity. The risk of coronary artery disease is proportionate to the levels of ApoA-I, the risk being amplified in patients suffering from any microbial infection and has been proposed to be a biomarker for coronary artery disease (Rahim et al. 2016). Apart from crucial cardioprotection, its diverse protective and defensive roles have been well elucidated (Oram 2003; Chroni et al. 2003). HDL is an important part of innate immune system, plays critical role in fighting pathogens, and displays extensive antiviral activity by preventing virus penetration (Singh et al. 1999). HDL has been reported to modulate protection against trypanosomes and Gram-negative bacteria (Hubsch et al. 1995; Perez-Morga et al. 2005). It is capable of neutralizing bacterial lipopolysaccharide (LPS) and inhibits inflammatory cytokines (Ulevitch et al. 1981). Antimicrobial activity has been demonstrated for several mammalian and teleostean ApoA-I. Human ApoA-I has been shown to possess antiviral activity against herpes simplex virus, human immunodeficiency virus and xenotropic virus, and bactericidal activity against Yersinia enterocolitica (Srinivas et al. 1991; Wu et al. 2002; Alonso-Villaverde et al. 2003; Biedzka-Sarek et al. 2011). Like mammalian ApoA-1, ApoA-I from different fish species such as Carp (Concha et al. 2004; Dietrich et al. 2014), rainbow trout (Villarroel et al. 2007), striped bass (Johnston et al. 2008), cat fish (Pridgeon and Klesius 2013), Hong Kong grouper (Qu et al. 2014), and amphioxus (Wang et al. 2019) has been shown to confer antibacterial activity against selected bacteria. Ontogenic expression analysis of Labeo rohita ApoA-I (LrApoA-I) transcript has been reported post different infection conditions, and a significant upregulation of ApoA-I transcript post ectoparasite Argulus siamensis has been shown earlier (Mohapatra et al. 2016). Though this study made a statement about the minimum bactericidal concentration of LrApoA-I against A. hydrophila and E. tarda, its structural-functional characterization and detailed analysis of antibacterial activity against different marine and other freshwater bacteria have not been reported. ApoA-I is rich in hydrophobic amino acid residues (Lyssenko et al. 2012; Nagao et al. 2014). Recombinant expression of hydrophobic proteins using bacterial system often results in their expression as inclusion bodies with minimal biological activity even after refolding. Therefore, the current study was undertaken to develop strategies that would result in soluble expression of bioactive recombinant ApoA-I of L. rohita in E. coli, to assess its antibacterial potential against a number of both marine and freshwater bacteria that infect fish, and to understand its structure-function relationships.

Materials and methods

Bacterial strains, vectors, and chemicals

Aeromonas hydrophila EUS112 (MTCC#12301) was kindly provided by Dr. I. Karunasagar, College of Fisheries, Mangalore (India) and all other bacterial strains, namely A. culicicola (MTCC#3249), A. liquefaciens (MTCC#2654), A. sobria (MTCC#1608), Vibrio parahaemolyticus (MTCC#451), V. harveyi (MTCC#7954), Edwarsiella tarda (MTCC#2400), Bacillus subtilis (MTCC#441), and Staphylococcus aureus (MTCC#740), were procured from the Microbial Type Culture Collection (MTCC), Institute of Microbial Technology, Chandigarh, India. Escherichia coli DH5α and BL21 (λDE3) pLysS host strains for cloning and expression respectively were obtained from GIBCO-URL, USA. Expression plasmid vectors pET22b+ and pET28a+ were obtained from Novagen, USA. Constituents for bacterial culture medium were obtained from HiMedia Laboratories Pvt. Ltd., Mumbai, India. All the other reagents used in the study were of molecular biology grade and were obtained either from Sisco Research Laboratory, India, or Sigma-Aldrich Chemical Co., USA,

Generation of Apolipoprotein A-I expression construct

Synthetic L. rohita ApoA-I gene construct, without the signal sequence, codon optimized for expression in E. coli, was designed based on complete cds of ApoA-I mRNA (GenBank Acc. no.KC934748.1). The sequence of the synthetic LrApoA-I gene has been submitted to GenBank (Accession no.MK370726.1). The synthetic ApoA-I gene (729 bp), cloned in expression vectors pET22b+ (Novagen, USA) at Nde I Xho I sites at the 5′and 3′ ends respectively, was obtained from Genscript, USA. The resultant construct was designated as pET22.LrApoA-I.

Expression of ApolipoproteinA-I in E. coli

Plasmid pET22.LrApoA-I clone was transformed into E. coli BL21(λDE3) pLysS cells. The transformants were selected on LB-agar plates containing ampicillin (100 μg/ml) and analyzed for expression essentially as described earlier (Karan et al. 2016). However, the expression level was negligible and could not be visualized by Commasie Brillinat Blue staining. Therefore, the synthetic ApoA-I gene was excised out from pET22.LrApoA-I by NdeI and XhoI digestion and subcloned into pET28a+. The transformants were selected on kanamycin (50 μg/ml)-LB agar plates and designated as pET28.LrApoA-I. The recombinant plasmid was transformed into E. coli BL21(λDE3) pLysS and induced for expression as described earlier (Karan et al. 2016). Briefly, the transformed E. coli BL21(λDE3) pLysS cells harboring pET28.LrApoA-I were grown until the log phase culture absorbance at 600 nm reached 0.6, following which the E. coli cells were induced with 1 mM isopropyl 1-thio-β-D-galactopyranoside (IPTG) and allowed to grow further for 4 h at 37 °C. The cells were harvested by centrifugation at 3000×g for 5 min at 4 °C, resuspended in lysis buffer (10 mM Tris-HCl, pH 7.5) and lysed by sonication as described earlier (Solanki et al. 2017). Uninduced and induced cell lysates were analyzed by SDS-PAGE (12%) for expression analysis. To assess the localization of expression, the cell lysates were centrifuged at 12,000×g for 20 min at 4 °C to separate the soluble proteins from insoluble fraction and the subcellular localization of rLrApoA-I was analyzed by SDS-PAGE (12%). To obtain maximum expression of soluble protein, the cells were induced at different temperatures and for different time points at the optimum temperature as indicated in the legends to the respective figure.

Equal protein concentrations were loaded for uninduced, induced cell lysates, soluble, and insoluble cellular fractions using Pierce BCA protein estimation kit (Thermo Fischer Scientific, USA).

Western blot analysis

Western blot analysis using anti-His monoclonal antibody was performed essentially as described earlier (Kaushik et al. 2018). The rLrApoA-I was resolved on the SDS-PAGE (12%) and transferred onto the nitrocellulose membrane using transfer buffer (25 mM Tris-HCl, pH 7.6, 192 mM glycine with 20% methanol) The membrane was then treated with blocking solution phosphate-buffered saline containing 0.1% Tween 20 (PBST) and 5% non-fat milk (Difco Laboratories Inc, USA) overnight at 4 °C. This was followed by incubation with primary anti-His monoclonal antibody (Sigma-Aldrich Chemical Co., USA) diluted in PBST (1:3000) for 1 h at room temperature (RT), followed by three PBST washes of 10 min each. Secondary antibody (anti-mouse HRP, Sigma-Aldrich Chemical Co., USA) was then added (1:5000 in PBST) and incubated for 1 h at RT. After washing, the blot was extensively with PBST as described above, the blot was developed using Pierce ECL Western blotting substrate (Thermo Scientific, USA). The image was acquired in a Biospectrum 500 Imaging system (UVP, Cambridge UK).

Purification of rLrApoA-I from soluble fraction

The recombinant ApoA-I of L. rohita (rLrApoA-I) was purified using Hispur cobalt resin (Thermo Fischer Scientific, USA). The soluble rLrApoA-I failed to bind to the resin in its native state, due to inaccessibility of the histidine tag. Therefore, the soluble fraction was supplemented with urea (4 M) for partial unfolding and kept for binding with the resin at 25 °C for 2 h with end to end shaking. The slurry was centrifuged at 1200×g in a swinging bucket rotor for 2 min. The flow through was discarded and non-specific unbound proteins were removed by washing the resin twice with the wash buffer (10 mM Tris-HCl, pH 7.5). The rLrApoA-I was eluted using elution buffer (10 mM Tris-HCl buffer, pH 7.5, and 100 mM imidazole). Different fractions were analyzed on SDS-PAGE (12%). The fractions showing the rLrApoA-1 presence were pooled together and dialysed against10 mM Tris-HCl buffer (pH 7.5) using 3.5-kDa dialysis membrane (Spectrum Laboratories, USA). The rLrApoA-I was then concentrated using Amicon ultra-centrifugal device (Millipore, USA) and stored at − 20 °C in small aliquots. Protein concentration of the purified rLrApoA-I was estimated using BCA protein estimation kit as per the manufacturer’s recommendation. Commercial services of Sandor Proteomics, Hyderabad were availed to carry out MALDI-TOF-MS of the rLrApoA-I to confirm the authenticity of the protein produced in E. coli.

Circular dichroism spectroscopy

The secondary structure of the rLrApoA-I (0.2 mg/ml in 10 mM Tris–HCl, pH 7.5) at 25 °C was measured by circular dichroism (CD) spectroscopy in a Jasco model J-815 spectropolarimeter fitted with a temperature-regulated cuvette chamber spectropolarimeter as described by Yadav et al. (2016). Briefly, ten successive spectra were accumulated in the far UV region (200 to 250 nm) at a scanning speed of 50 nm/min in a cell with path length of 0.1 cm. The spectra were averaged followed by baseline correction. Mean residue weight ellipticities were expressed as degree × cm2 × dmol−1. Secondary structure contents of the folded rLrApoA-I from the CD measurements were calculated using the K2D3 software (http://cbdm-01.zdv.uni-mainz.de/~andrade/k2d3/). In order to assess the effect of temperature on the secondary structure of the rLrApoA-I, the rLrApoA-I was incubated at different temperatures (25 °C, 35 °C, 45 °C, 55 °C, 65 °C, 75 °C, and 85 °C) for 30 min, followed by far-UV circular dichroism spectroscopy.

Fluorescence spectroscopy

To assess the tertiary structure of the rLrApoA-I, fluorescence spectroscopy analysis of the rLrApoA-I (0.2 mg/ml in 10 mM Tris-HCl, pH 7.5) was carried out using a spectrofluorimeter fitted with a Peltier temperature controller (Cary Eclipse, Varian Optical Spectroscopy Instruments, USA) with excitation at 280 nm for tyrosine and tryptophan residues, and 295 nm for mainly tryptophan residues (slit width 10). Emission spectra were monitored in the wavelength range of 300 nm and 500 nm (slit width 10) at a scan speed of 30 nm/min. The height at maximal emission (emission maxima) was measured. To assess the effect of temperature on the tertiary structure of the rLrApoA-I, fluorescence spectra were recorded at different temperatures. Temperature kinetics of rLrApoA-I unfolding was analyzed by fluorescence spectroscopy of the protein (excitation at 280 nm, emission between 300 and 500 nm) incubated at different temperatures for 30 min. The kinetics of unfolding is represented by plotting peak fluorescence intensity at 334 nm versus incubation temperature.

Antimicrobial activity of recombinant Apolipoprotein A-I (rLrApoA-I)

The antimicrobial activity assay of rLrApoA-I was performed as described earlier (Johnston et al. 2008; Verma et al. 2018) against the following bacterial species: A. hydrophila, A. liquefaciens, A. sobria, A. culicicola, B. subtilis, E. coli, S. aureus, V. harveyi, and V. parahaemolyticus. A single colony of different bacteria was inoculated in Luria Bertani broth and allowed to grow overnight. The bacterial culture was then diluted in LB to a final colony count of 1 × 105 cfu/ml. Bacterial culture (1 × 104 cfu in 100 μL, in triplicate) was aliquoted in 96-well plate. The rLrApoA-I was filter sterilized (0.22 μm) and serially diluted in culture medium. One hundred microliters of different dilutions of the rLrApoA-I was then added to the bacterial culture in 96-well plate to attain a final concentration of 0 to 1 mg/ml to determine minimal bactericidal concentrations (MBC).The bacteria were then allowed to grow for 12 h in a plate shaker at 37 °C. Bacterial growth was monitored by measuring the absorbance at 600 nm at every 2 h. Effect of temperature on biological activity of rLrApoA-I was studied by heating rLrApoA-I for 1 h at varying temperature prior to addition with the bacterial culture.

Disk diffusion assay for determination of zone of inhibition by r Lr ApoA-I

The overnight grown A. hydrophila, A. liquefaciens, A. sobria, A. culicicola, V. harveyi, V. parahaemolyticus, and E. tarda culture were spread on LB agar plate to obtain bacterial lawn. Different dilutions of rLrApoA-I were prepared in PBS so as to maintain the application volume constant. Sterile disks (6 mM, HiMedia Laboratories Pvt. Ltd. India) impregnated with different concentrations of rLrApoA-I (0 μg, 25 μg, 50 μg, 100 μg in a fixed volume of 10 μl) was pressed aseptically onto the inoculated agar plate to ensure contact. The bacteria were then allowed to grow at 25 °C for 16 h. To assess the effect of temperature on antibacterial activity of rLrApoA-I, the rLrApoA-I was incubated for 1 h at varying temperature prior to use in disk diffusion assay.

Statistical analysis

The significance level of the change (p value) brought about by the rLrApoA-I was analyzed using two-way analysis of variance (ANOVA) method. For this purpose, GraphPad Prism 7 software (Chicago, IL, USA) was used.

Gene sequences

The nucleotide sequence of the codon-biased synthetic ApoA-I gene of L. rohita has been submitted to NCBI GenBank (Accession no. MK370726.1)

Results

Cloning and Expression of ApolipoproteinA-I

Synthetic codon-biased rLrApoA-I gene cloned into pET22b+ failed to express (data not shown). Therefore, subcloning of the 729 bp NdeI–XhoI fragment encoding the ApoA-I gene, excised from the pET22.LrApoA-I, was done in plasmid pET28a+ digested with the same enzymes and transformed into E. coli BL21 (λDE3) pLysS cells for expression.

In order to get expression of the recombinant protein, we carried out induction at different temperatures and analyzed for expression and localization of the rLrApoA-I. As shown in Fig.1a, no expression was observed at 15 °C. SDS-PAGE analysis of the uninduced and induced cell lysates of cultures grown at different temperature showed the presence of rLrApoA-I with an apparent molecular weight of ~ 31 kDa only in the induced cell lysate (Fig. 1a) at other temperatures indicating tight regulation of expression. Maximum expression was obtained in the culture induced at 37 °C. Localization analysis revealed that all the induction temperatures, the rLrApoA-I expressed predominantly as soluble protein (Fig.1b), though negligible amounts were seen in the insoluble fraction in the culture induced at 37 °C. Time kinetics of expression showed that expression of the recombinant protein increased with time until 4 h, which nearly remained constant until 6 h (Fig. 1c).These data resulted in identification of conditions that led to maximum expression of the soluble rLrApoA-I, and for purification of the large amounts of the protein, the culture was induced with 1 mM IPTG for 4 h at 37 °C. Western blot analysis of the cell lysates induced under these conditions using anti-His antibody confirmed the protein to be rLrApoA-I as a single band at the expected position appeared only in the induced cell lysates (Fig. 1d).

Fig. 1
figure 1

Expression analysis of recombinant ApoA-I in E. coli BL21(λDE3) pLysS cells harboring pET28.ApoA-I. a Effect of temperature on rLrApoA-I expression. SDS-PAGE (12%) analysis of the cell lysates induced at different temperatures (15 °C, 25 °C, 37 °C, 40 °C). U and I indicate the uninduced and induced cell lysates at the respective temperature. b Localization of expression of the rLrApoA-I induced at different temperatures. SDS-PAGE (12%) analysis of the soluble (S) and insoluble (I) fractions of the cell lysates of E. coli BL21(λDE3) pLysS cells harboring pET28.ApoA-I induced at different temperatures. c Time kinetics of rLrApoA-I expression. SDS-PAGE (12%) analysis of E. coli BL21(λDE3) pLysS cells harboring pET28.ApoA-I induced at 37 °C for different time periods (1 to 12 h). U indicates the uninduced cell lysates. Arrow points to the rLrApoA-I band. d Western blot analysis of the rLrApoA-I using anti-His antibody. U and I indicate the uninduced and induced cell lysates of E. coli BL21(λDE3) pLysS cells harboring pET28.ApoA-I. The arrow points to the rLrApoA-I protein of ~ 31 kDa detected only in the induced cell lysates. e The figure shows SDS-PAGE (12%) analysis of the rLrApoA-I purified using metal affinity chromatography. M in all the panels denote migration of protein molecular weight markers

Full size image

Purification of ApolipoproteinA-I from soluble fraction

The hexa-histidine-tagged rLrApoA-I was purified from the soluble fraction using Hispur cobalt resin affinity chromatography (Fig. 1e). The rLrApoA-I was purified to near homogeneity (> 98%) as determined by densitometric analysis of the gel with an approximate yield of ~ 75 mg/L of soluble rLrApoA-I at shake flask level. MALDI-TOF-MS of the purified rLrApoA-I further confirmed the authenticity of the purified protein (Fig. S1).

Structural characterization of the rLrApoA-I

Far-UV circular dichroism spectroscopy of the rLrApoA-I revealed it to predominantly possess α-helical structure with characteristic double minima in ellipticity at 208 nm and 222 nm (Fig 2a). Analysis of the spectra using K2D3 software for secondary structure prediction showed the rLrApoA-I at ambient temperature (25 °C) to have 82.56% α-helical content, 1.24% β-strand, and 16.2% random coil. When the rLrApoA-I was incubated at different temperatures for 30 min, no significant change in the secondary structure was observed upto 45 °C (Fig. 2b). Incubation at 55 °C for 30 min resulted in significant loss of native tertiary structure with a shift from α-helical content to β-turns. Analysis of various secondary structure characteristics upon treatment at different temperatures is summarized in Table 1. The Tm of rLrApoA-I was calculated as 50 °C. After incubation at 85 °C, the rLrApoA-I contained only 2.48% α-helical content and 43.27% β-strand, remainder being contributed by random coil.

Fig. 2
figure 2

Structural analysis of the rLrApoA-I. a Circular dichroism (CD) spectrum of the soluble ApoA-I in the far-UV regions (200–250 nm) at room temperature (25 °C). Mean residue mass elipticity [θ] values are shown as degree × cm2 dmol−1. b Effect of temperature on the secondary structure of the rLrApoA-I. The rLrApoA-I was incubated at the indicated temperatures between 25 and 85 °C for 30 min, followed by CD spectral analysis in the far-UV regions. Mean residue mass elipticity [θ] values are shown as degree × cm2 dmol−1. c Intrinsic fluorescence spectra (contributed by both tyrosine and tryptophan residues) of the rLrApoA-I (excitation at 280 nm, emission between 300 and 500 nm). Kinetics of unfolding was determined by measuring intrinsic fluorescence at different temperatures. d Intrinsic fluorescence spectra of the tryptophan residue (Trp155) at different temperatures (excitation at 295 nm, emission between 300 and 500 nm). e Kinetics of temperature induced unfolding of rLrApoA-I. Peak fluorescence intensities obtained by analysis of intrinsic fluorescence (excitation at 280 nm, emission between 300 and 500 nm) at different temperatures are plotted against the temperature. RFI (AU) denotes relative fluorescence intensity in arbitrary units

Full size image
Table 1 Secondary structural components of the rLrApoA-I incubated at different temperatures for 30 min prior to CD spectroscopic analysis
Full size table

Fluorescence spectroscopy of the rLrApoA-I was performed to analyze temperature dependence of the tertiary structure of the protein. An excitation wavelength of 280 nm was used to excite all the 8 Tyr residues and 1 Trp residues, whereas to selectively excite 1 Trp-155 at 155th position, an excitation wavelength of 295-nm wavelength was used. The rLrApoA-I gave emission maxima at 334 nm indicating intact tertiary structure when excited at 280 nm whereas emission maxima of 337 nm was observed when excited at 295 nm (Fig. 2 c and d, respectively). A gradual decrease in the emission maxima was observed with increase in temperature until 90 °C, indicating loss of tertiary structure. In order to study the steady state of unfolding of the rLrApoA-I, the protein was incubated for 30 min at each temperature followed by fluorescence spectroscopy. This analysis showed consequent decrease in fluorescence intensity at 334 nm with gradual increase of temperature at 280-nm excitation and red shift of 4 nm was observed when Trp-155 was excited at 295 nm (Fig. 2 c and d). The kinetics of the rLrApoA-I unfolding is evident from the gradual decrease in relative fluorescence intensity with increase in temperature (Fig. 2e).

Antimicrobial activity of recombinant Apolipoprotein A-I

In vitro antimicrobial activity assay using broth dilution assay showed that rLrApoA-I was able to inhibit growth of A. hydrophila, A liquefaciens, A. culicicola, A. sobria, V. harveyi, V. parahaemolyticus, and E. tarda in a concentration dependent manner (Fig. 3). While rLrApoA-I was able to inhibit the growth of A. culicicola significantly (p ≤ 0.001) at all the concentrations from 2 h onwards, it was able to inhibit growth of A. liquefaciens at 50 μg/ml and higher from 2 h onwards (p ≤ 0.001). Significant inhibition of growth could be observed from 4 h onwards for A. hydrophila (p ≤ 0.05 to p ≤ 0.001), A. sobria (p ≤ 0.001), V. harveyi (p ≤ 0.05 to p ≤ 0.001) at concentrations starting from 100 μg/ml. For V. parahaemolyticus and E. tarda, significant inhibition of growth could be observed from 8 h onwards at concentrations 100 μg/ml and 200 μg/ml of rLrApoA-I (p ≤ 0.001). At 500 μg/ml and 1000 μg/ml, the rLrApoA-I was able to significantly inhibit the growth of all the test bacteria from 2 h onwards (p ≤ 0.005 to p ≤ 0.001). No inhibition of growth was observed for B. subtilis, E. coli, and S. aureus (data not shown). Percentage growth inhibition of all the bacteria at 12 h of growth period at different concentration is shown in Fig. 4. As evident, a dose-dependent increase in growth inhibition could be seen with A. hydrophila, A. sobria, A. liquefaciens, V. harveyi, and V. parahaemolyticus. Minimum bactericidal concentration for A. hydrophila, E. tarda, V. parahaemolyticus, and V. harveyi was found to be 100 μg/ml or 3.23 μM whereas for A. liquefaciens, A. culicicola, and A. sobria, minimal bactericidal concentration was found to be 25 μg/ml or 0.81 μM. The antibacterial property of rLrApoA-I was further confirmed with an increase in zone of inhibition in the presence of rLrApoA-I on LB agar plate for A. hydrophila, A. liquefaciens, A. culicicola, A. sobria, V. harveyi, V. parahaemolyticus, and E. tarda culture. The zone of inhibition obtained with different bacteria in the presence of different concentrations of rLrApoA-I is summarized in Table 2.

Fig. 3
figure 3

a–f Growth profile of different bacteria (0–12 h) in the presence of different concentrations (25 to 1000 μg/ml) of the rLrApoA-I using broth dilution assay. The data represent mean ± SD of three independent experiments performed in triplicates. aAeromonas hydrophila, bAeromonas culicicola, cAeromonas sobria, dAeromonas liquefaciens, dVibrio harveyi, eVibrio parahaemolyticus, and fEdwardsiella tarda. The significance level (p values) of the change observed in growth of different bacteria treated with varying concentrations of the protein at different time intervals in comparison to the control cells treated with PBS only at the respective time point is shown in Fig. S2

Full size image
Fig. 4
figure 4

Percentage growth inhibition of different bacteria at 12 h of growth at 37 °C in liquid culture in the presence of different concentrations of rLrApoA-I. Aeromonas hydrophila (Ah), A. liquefaciens (Al), A. culicicola (Ac), A. sobria (As), Vibrio harveyi (Vh), V. parahaemolyticus (Vp), and Edwardsiella tarda (Et). Percentage growth inhibition is calculated by taking the absorbance of untreated cells grown at 37 °C for 12 h as 100% growth. The data represent mean ± SD of three different experiments performed in triplicate. The numbers 1, 2, 3, 4, 5, and 6 on X-axis denote rLrApoA-I concentrations at 25 μg/ml, 50 μg/ml, 100 μg/ml, 200 μg/ml, 500 μg/ml, and 1000 μg/ml, respectively. The letters a and b indicate p values ≤ 0.001 and ≤ 0.005, respectively

Full size image
Table 2 Zone of inhibition of different bacteria at varying concentrations of rLrApoA-I, after 16 h incubation at 25 °C, employing disc diffusion assay
Full size table

Since the rLrApoA-I caused significant growth inhibition (p ≤ 0.001) of A. hydrophila, A. culicicola, and V. harveyi, these bacteria were chosen to assess the effect of temperature on the antibacterial activity of rLrApoA-I. As shown in Fig. 5a–c, there was a significant decrease (p ≤ 0.001) in the growth of these bacteria in the presence of rLrApoA-I. No decrease in the growth inhibitory activity of the rLrApoA-I was observed even when it was pre-incubated at temperatures higher than the room temperature up to 85 °C prior to its activity assay for these bacteria. To assess if the antibacterial activity of the heated rLrApoA-I is due to rapid reanaturation during the assay, we compared the CD spectra of the protein sample heated at 85 °C for 30 min with that of the protein sample which was heated at 85 °C for 30 min and allowed to cool down to room temperature for 1 h. As evident from Fig. 5d, the spectra of the two samples (heated alone and cooled after heating) were identical and no renaturation of the protein was observed after cooling.

Fig. 5
figure 5

Effect of temperature on growth inhibition ability of the rLrApoA-I on aA. hydrophila (Ah), bV. harveyi (Vh), and cA. culicicola (Ac). The rLrApoA-I was incubated at different temperatures for 1 h prior to its use in the bacterial growth curve assay at a concentration of 200 μg/ml. The Ah, Vh, and Ac indicate cells treated with corresponding volume of PBS. The points with significant change level of p ≤ 0.001 with respect to PBS-treated control cells are boxed in grey dotted line box. d Comparative analysis of CD spectra (in the far-UV region) of the rLrApoA-I heated to 85 °C and cooled down to room temperature with that of the rLrApoA-I heated to 85 °C. CD spectrum of the control rLrApoA-I at 25 °C is also shown. Mean residue mass elipticity [θ] values are shown as degree × cm2 dmol−1

Full size image

Discussion

Apolipoprotein A-I is the major protein of HDL (70%) and plays a critical role in primary immune response pathway. It has host-defensive role and fights against viral, bacterial, and parasitic infections. The mature human ApoA-I is obtained after proteolytic processing of the preproprotein as a single polypeptide chain of 243 amino acids and consists majorly of amphipathic alpha helices (Law and Brewer Jr 1984; Phillips 2013). ApoA-I has been well studied in human and has also been identified in various mammalian and non-mammalian species rat (Rattus rattus), rabbit (Oryctolagus cuniculus), chicken (Gallus gallus domesticus), salmon (Salmo salar), cod (Gadus morhua), rainbow trout (Oncorhyncus mykiss), seabream (Sparus aurata), skate (Raja erinacea), brown bullhead (Ameriusne bulosus), and sea turtle (Chrysemys spicta) (Denslow et al. 1994; Llewellyn et al. 1998; Duggan and Callard 2001; Duggan et al. 2000; Magnadottir and Lange 2004).

Immunomodulatory role of the ApoA-I of L. rohita and its ontogenic expression under different infection conditions have been reported (Mohapatra et al. 2016). However, being a eukaryotic protein, the bottlenecks to produce this protein in a heterologous expression system for further analysis need to be taken care of. In order to understand the structure function relationship of the ApoA-I of L. rohita, attempt to express the L. rohita ApoA-I gene in E. coli resulted in no expression. Therefore, a synthetic gene (without the signal sequence) was designed according to codon preference of E. coli.

It has been reported earlier that due to its amphiphilic property and instability, expression and purification of Apolipoprotein is not always easy as the protein tends to show auto-aggregation and degradation (Schmidt et al. 1997). Initial attempt to express the native ApoA-I gene of L. rohita in E. coli did not result in any expression (data not shown). Therefore, we chose to express a codon-biased L. rohita ApoA-I gene in E. coli. Analysis of expression showed that unlike the native ApoA-I gene of L. rohita, the codon-biased synthetic hexa-histidine-tagged LrApoA-I could be successfully and efficiently expressed in E. coli. The evident increase in the intensity of the rLrApoA-I protein band at the expected size in the cells harboring the codon-biased ApoA-I gene is much greater than the predicted theoretical increase ~ 10% to ~ 20% of total soluble protein after codon optimization which was in fact undetectable when expressed from native gene sequence (Plotkin and Kudla 2011). More than five-fold increase in yield of the human ApoA-I from optimized cDNA was obtained as compared with the original cDNA (Ryan et al. 2003). It appears that the L. rohita’s ApoA-I gene codon preference is much far apart from E. coli than human ApoA-I as the codon optimization of L. rohita ApoA-I resulted in very good expression whereas no expression was seen with the native gene. This dramatic increase in the expression of the rLrApoA-I resulted in higher expression because “rare” codons were replaced with abundant codons, and the codon adaptation index (CAI) was increased using the optimized sequence. Our results are in agreement with those of Khalili and coworkers who also reported a significant increase in the expression of recombinant and biologically active human heparin-binding epidermal growth factor (HB-EGF3) through codon optimization in an E. coli expression system (Khalili et al. 2015). Often over-expression of the recombinant protein leads to the formation of inclusion bodies. For a protein to be biologically active, it is important that the protein is expressed in soluble form. Therefore, recombinant expression of the codon-biased rLrApoA-I gene was optimized to obtain maximum expression of the soluble protein by carrying out induction at different temperatures, i.e., 15 °C, 25 °C, 37 °C, and 40 °C. Analysis of soluble and insoluble fractions of the cultures induced at different temperatures clearly showed maximum expression at 37 °C in soluble fraction, though some amount of protein was also seen in insoluble fraction. Even at 40 °C, though there was a decrease in the expression, the protein mostly remained in the soluble fraction. These data show that the protein did not form aggregate even at 40 °C. These observations are in line with the stable secondary structure of the protein which did not show any significant change in the secondary structure even when heated up to 45 °C for 30 min. Initial attempts to purify the protein using standard binding buffer (10 mM Tris-HCl buffer, pH 7.5) did not result in any binding of the rLrApoA-I with the resin, possibly due to the inaccessibility of the histidine tag in the soluble protein. Therefore, to make the hexa-histidine tag of the rLrApoA-I to the immobilized Co2+, 4 M urea was included in the soluble fraction of the cell lysate, so as to allow partial unfolding of the protein. The exposed portion of hexa-histidine tag was hence made available for binding with the resin and the soluble rLrApoA-I was thus recovered from the soluble fraction of the induced culture using single-step purification from heterologous expression system with a yield of ~ 75 mg of the purified rLrApoA-I from 1 L of culture at shake flask level. This yield of soluble rLrApoA-I of L. rohita is far greater than that reported by Smith and coworkers for human ApoA-II (~ 10 mg/L) (Smith et al. 2012). The yields obtained in the present study are comparable with those reported by Ryan and his associates wherein individual mutations of the codons for proline, arginine, leucine, and one glycine were incorporated in the cloned gene converted low abundant codon to high-frequency codons (Ryan et al. 2003). MALDI-TOF-MS analysis confirmed the authenticity of the rLrApoA-I of L. rohita as the mass analysis of the peptides generated from proteolytic cleavage matched with the amino acid sequence of the theoretically generated peptides of the L. rohita ApoA-I.

In order to understand structure function relationship, it was important to initially know the secondary structure of the rLrApoA-I and any alteration that changes in temperature may induce. Secondary structure prediction of the rLrApoA-I using CD spectroscopy exhibited 82.56% alpha helices, which is much greater than that reported for human ApoA-I (50–57% alpha helical content). Theoretical prediction of the secondary structure from the amino acid sequence of human ApoA-I indicated presence of amphipathic helices of repeating 11 or 22 amino acids separated by proline residues (Segrest et al. 1998). The predicted secondary structure has been found to be conserved among vertebrates and teleost (Bashtovyy et al. 2011). This evolutionarily conserved structure may be responsible for the conserved defensive function of ApoA-I (Concha et al. 2004).

Temperature-induced unfolding kinetics of the purified rLrApoA-I of L. rohita was assessed by fluorescence spectroscopy. Protein sequence of mature Apolipoprotein A-I contains one phenylalanine residue, eight tyrosine residues, and one tryptophan residues with intrinsic fluorescence properties. However, for tertiary structure characterization, we measured Tyr and Trp fluorescence because their quantum yield was high enough to give a good fluorescence signal. Fluorescence spectra for both the tryptophan and tyrosine show a decrease in fluorescence with an increase in temperature from 20 to 90 °C, with a red shift in the absorbance maximum. The decrease in the fluorescence with increase in temperature thus appears to be due to some loss of tertiary structure, and not due to the quenching of the intrinsic fluorescence with an increase in temperature. The native folded state of rLrApoA-I contained tyrosine and tryptophan residues within the core of the protein, whereas in a partially folded or unfolded state, they become exposed to solvent. In a hydrophobic environment, tyrosine and tryptophan resulted in high quantum and hence high fluorescence intensity. Thermal denaturation resulted in unfolding of the protein thus exposing the 8 Tyr and 1 Trp residues to a hydrophilic environment reflected by decreased fluorescence intensity. Thus, while the secondary structure of the rLrApoA-I remained stable until 45 °C, there was a temperature-dependent loss of tertiary structure.

ApoA-I from mammals has been demonstrated to have a variety of biological activities such as its role in cholesterol transport, lipid metabolism, and anti-atherogenic properties (Lewis and Rader 2005). Protective and immunomodulatory role of ApoA-I in mammals, lower vertebrates, and as well as in some teleost fishes has also been reported (Zamanian-Daryoush and DiDonato 2015; Mohapatra et al. 2016). Native ApoA-I isolated from different fish species such as Cyprinus carpio (carp), Oncorhynchus mykiss (rainbow trout), Morone saxatilis (striped bass), and Ictalurus punctatus (channel catfish) has been reported to possess lytic and antimicrobial activity in vitro against Gram-positive and Gram-negative bacteria (Concha et al. 2004; Johnston et al. 2008; Pridgeon and Klesius 2013). Though ApoA-I from different organisms do not share high degree of identity (Mohapatra et al. 2016), the antimicrobial function of ApoA-I is retained through evolution.This led us to assess if the recombinant ApoA-I of L. rohita expressed in E. coli retained its antimicrobial properties. Potential antibacterial effect of the rLrApoA-I against a number of important Gram-negative bacterial (both freshwater and marine) pathogens that infect both fish species and mammals was evaluated. Presence of rLrApoA-I resulted in significant reduction of bacterial growth of the selected bacteria, namely A. hydrophila, A. liquefaciens, A. sobria, A. culicicola, V. harveyi, V. parahaemolyticus, and E. tarda, both in liquid culture using broth dilution assay as well as on agar plates using disc diffusion assay. Thus, the rLrApoA-I of L. rohita expressed in heterologous expression system exhibited broad spectrum antibacterial activity. ApoA-I from other species including teleosts have been reported to inhibit growth of few bacteria (Beck et al. 2013), whereas in the present study, we report antibacterial activity of the rLrApoA-I of L. rohita against a number of bacteria that result in major economic losses to both freshwater and marine aquaculture. For L. rohita ApoA-I also, MIC concentration of ApoA-I only for A. hydrophila and E. tarda has been reported without any detailed analysis on their growth kinetics etc. The observed MIC of rLrApoA-I in the present study is in agreement with the reported MIC (100 μg/ml) of recombinant ApoA-1 from channel catfish, stripped bass, and Hong Kong grouper fish against a number of both Gram-positive and Gram-negative bacteria including A. hydrophila, V. harveyi, and E. tarda (Johnston et al. 2008; Pridgeon and Klesius 2013; Qu et al. 2014; Mohapatra et al. 2016). Growth inhibitory effect of ApoA-I from amphioxus (Branchiostoma belcheri) against A. hydrophila and V. vulnificus at 1.47 μM and 2.15 μM, respectively has recently been reported (Wang et al. 2019). The recombinant ApoA-I from channel catfish showed MIC against A. hydrophila as high as 6 μM (Pridgeon and Klesius 2013). Thus, the observed MIC of rLrApoA-I falls within the range reported for ApoA-I from other species. The antibacterial activity of ApoA-I can be attributed to its amino acid composition and its structure. Human and mouse ApoA-I are composed of an N-terminal helix bundle and independently folded C-terminal domain (Koyama et al. 2009). The N- and C-terminal helical regions in the ApoA-I molecule contribute to the strong lipid-binding properties of this protein as well as the conformational stability in solution (Wang et al. 2007; Palgunachari et al. 1996). Ionic interactions also play an important role in binding interaction of ApoA-I and negatively charged membrane components of outer and inner bacterial membrane composed of LPS and phosphatidylglycerol (PG) respectively. Human ApoA-I is able to bind lipopolysaccharides present in the outer membrane of Gram-negative bacterium and has been shown to have a strong preference for bilayer vesicles consisting of phosphatidylglycerol over phosphatidylcholine (Beck et al. 2013).The C-terminal residues 220–241 play a major role in binding of ApoA-I to phospholipids and also interact with LPS (Palgunachari et al. 1996; Laccotripe et al. 1997; Henning et al. 2011).The role of lysine residues of ApoA-I and the contribution of the C-terminal domain in the interaction of ApoA-I with bacterial membrane components have also been investigated (Ajees et al. 2006; Jenssen et al. 2006; Kaconis et al. 2011). Studies suggest that alpha-helical and amphipathic content of the protein is responsible for the lytic and pore formation–mediated antimicrobial properties (Perez-Morga et al. 2005). ApoA-I recognizes Yersinia enterocolitica through O-antigen present on the surface and the activity of ApoA-I-mediated pathogen interaction is directly proportional to the amount of O-antigen (Biedzka-Sarek et al. 2011). Human ApoE has been reported to effectively neutralize potential lethal dose of LPS in bacterial membrane (van Oosten et al. 2001). Thus, high cationic residue content, large hydrophobic regions, and the presence of lysine residue at 223 positions the L. rohita ApoA-I might be playing a significant role in protection against Gram-negative bacterial pathogens.

It is of interest to note that though a loss in secondary structure of the rLrApoA-I was seen at and above 55 °C, no loss in the antibacterial activity of the rLrApoA-I against V. harveyi, A. culicicola, and A. hydrophila (against which the protein exhibited maximum zone of inhibition) was observed even when the protein was incubated up to 30 min at temperatures up to 85 °C. Comparative analysis of the CD spectrum of the rLrApoA-I sample which was heated to 85 °C and allowed to cool down to room temperature with the CD spectrum of the rLrApoA-I heated at 85 °C only showed that there was no difference in the CD spectra of the two samples, thus ruling out the possibility of rapid renaturation during the time of the antibacterial activity assay. Our results are in agreement with the earlier report on recombinant human ApoA-I where also no renaturation was observed after cooling of the sample which was heated for only 10 min (Kim et al. 2000). This indicates that the loss in secondary and tertiary structure is not essential for antibacterial activity. Antimicrobial property of few proteins such as napin has been attributed to the permeabilization of the pathogen’s membrane caused by electrostatic contacts between basic residues of the protein and acidic phospholipids present on the membrane of the pathogen (Rico et al. 1996). Amino acid sequence analysis of the codon-biased rLrApoA-I using Peptide hydrophobicity/hydrophilicity Analysis Tool (Peptide 2.0 Inc., available online) shows that it is highly rich in hydrophobic (41%) and basic amino acid residues (19.25%). Thus, it is likely that penetration and permeabilization of the bacterial membrane by the hydrophobic and basic amino acid residues of the rLrApoA-I itself result in the antibacterial activity and that the integrity of secondary and tertiary structure is not absolutely necessary for it antibacterial activity.

In conclusion, we have presented strategies to produce biologically active recombinant ApoA-I of Labeo rohita in large amounts, which showed broad spectrum antibacterial activity against a number of Gram-negative bacteria of both marine and freshwater environment. The structure-function analysis revealed that the protein retained its antibacterial activity even after the loss of secondary and tertiary structure. Retention of antibacterial property even when protein is subjected to higher temperatures is of significance as it overcomes the loss of biological activity for most proteins during transport. Since the ApoA-I of L. rohita has been shown to confer an immunomodulatory role also, the observed antibacterial activity together with its immunomodulatory role is beneficial for its use in fish cultures. However, application of any therapeutic, in this case rLrApoA-I, in aquaculture by injection is not practical as it is tedious and time consuming. Administration of rLrApoA-I to fish by immersion could also lead to its degradation. Therefore, oral administration using pH-sensitive formulation or edible nano-formulations could possibly be employed for its use in aquaculture as reported earlier (Kulkarni et al. 2014; Costa et al. 2016; Shaalan et al. 2016; Mathews et al. 2018). Further studies using site-directed mutagenesis would help identify amino acid residues of the ApoA-I critical for its antibacterial activity and may give insight into mechanism of rLrApoA-I action. This may lead to generation of potential therapeutic antibacterial peptides for application in aquaculture.