All Periplaneta americana samples were collected from the dormitories and the outdoor surroundings of Guangdong Pharmaceutical University (Guangdong, China), which were identified morphologically and molecularly. Periplaneta americanas were fasted for 3 days and then sacrificed following anesthesia with 100% CO₂ for 2 h. They were then washed with distilled water. Sterilization was conducted with 75% ethanol and 3.5% sodium hypochlorite/water solution followed by 3 washes with 1X sterile phosphate-buffered saline (PBS). The entire gut of the Periplaneta americana was extracted from the abdomen and then ground into a slurry. The bacterial fluid was diluted into 0.01, 0.005 and 0.00025 µg/ml by gradient, coated on the agar-solidified Luria-Bertani (LB) medium (Amresco, LLC, Solon, OH, USA), which contained of 75 µg/ml potassium dichromate, and then incubated in the dark at 28°C for a week (10). The morphology and surface characteristics of the bacterial colonies were examined using a scanning electron microscope. Finally, the red-pigment-producing bacterial strain was named, WA12-1-18.

The strain WA12-1-18 was cultured in LB medium broth for 2 days with shaking at 37°C and harvested by centrifugation (12,000 × g at 20°C for 15 min). DNA was extracted and amplified by polymerase chain reaction (PCR) using the genomic DNA as a template and the bacterial universal primers, 27 forward (5′-GAGTTTGATCACTGGCTCAG-3′) and 1492 reverse (5′-TACGGCTACCTTGTTACGACTT-3′). The PCR mixture (25 µl) contained 1 µl of template DNA, 2.5 µl of 10X Taq DNA polymerase buffer, 1.5 µl primers (each) and 18.5 µl ddH₂O. The PCR mixture was then incubated with the following thermocycling conditions: 94°C for 3 min, followed by 35 cycles of 94°C for 0.5 min, 58°C for 0.5 min and 72°C for 1.5 min, followed by a final extension performed at 72°C for 10 min (11). The amplification products were purified using a DNA purification kit (Qiagen, Inc. Valencia, CA, USA), and sequencing was performed by Invitrogen (Thermo Fisher Scientific, Inc., Waltham, MA, USA). Phylogenetic trees were then constructed from evolutionary distances using the neighbor-joining method (12).

A single colony was inoculated in 10 ml of LB medium and incubated for 12 h at 37°C, and then transferred into 200 ml LB medium and incubated for a further 3 days at 37°C. Then, 250 µl bacterial cultures were spread onto LB solid plates and incubated in the dark at a low temperature (23°C) for 72 h (13). The lawn plates were harvested and dissolved in ethanol and centrifuged at 10,000 × g for 20 min at 4°C. The upper layer was collected, dried and re-dissolved in ethylacetate, the pigment was purified by silica gel column chromatography (28×12 cm) with petroleum ether and ethylacetate/n-hexane (40:1, v:v), as described previously (14,15). At the end of these procedures, the pigment fractions were pooled and dried.

As described previously (16), prodigiosin was analyzed via HPLC at the absorption wavelength 534 nm using Phenomenex Luna C-182 semipreparative column (250×10 mm; sample, 5 µl; Phenomenex, Torrance, CA, USA) with the Bio-Rad HPLC system (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The separation was performed using water (solvent A) and acetonitrile/methanol (1:1; solvent B) mobile phases, and a gradient elution program at 3 ml/min with the following parameters: 0–25 min 15–100% B (linear gradient), 25–35 min 100% B and 35–40 min 15% B to re-equilibrate the column. Fractions containing the targeted compounds were combined and concentrated by solvent evaporation. UV absorption of prodigiosin in methanol at pH 5 and pH 9 was scanned using a spectrophotometer at a range of 200–800 nm. In the FTIR study, FTIR spectra were made on a Nicolet Nexus 670 FT-IR infrared spectrometer. All spectra were taken using the Attenuated Total Reflection method (17). Briefly, prodigiosin was mixed with potassium bromide in a clean glass pestle and compressed to produce a pellet. Scanning was performed following base line correction by setting a wave number range of 400–4,000/cm⁻¹. In LC-MS analysis, the pigment was dissolved in acetonitrile, and 10 µl was analyzed in API-4000 Q trap LC-MS/MS. The operating conditions were as follows: 4,500 ionization voltage, 350°C ionization temperature, ionization mode ESI positive, a MS C18 (50×4.6 mm) column was used, 90:10 (acetonitrile:water) mobile phase, and the injection volume was 10 µl with a flow rate of 0.5 ml/min; performed as described previously (18). In addition, prodigiosin dissolved in d-chloroform was analyzed by hydrogen (¹H)-NMR and carbon-13 (¹³C)-NMR to identify the structure of the purified product. NMR characterization (¹H NMR or ¹³C NMR) was performed using the pulse program zg30 with a spectral width of 8,278.146 Hz using Bruker Advance 400 MHz NMR spectrophotometer (Bruker Corporation, Ettlingen, Germany) with 5 mm probe in CDCl₃ solvent.

The HeLa cell line (Experimental Animal Center of Sun Yat-sen University, Guangzhou, China) was maintained in Dulbecco's modified Eagle's medium. An MTT assay was used to measure the levels of cell proliferation. A total of 5×10⁴ cells were treated with different concentrations (0, 0.5, 1.0 and 2.0 µg/ml) of prodigiosin for 24, 48 and 72 h at 37°C, followed by incubation in MTT (0.1 mg/ml) at 37°C for 4 h and dissolution in dimethyl sulfoxide at room temperature for 10 min. Microplate reader (at 570 nm) was used to measure the absorbance of each well. Three independent experiments were performed (19–21).

In order to view the nucleus, 5×10⁴ HeLa cells were fixed in methanol, incubated with DAPI for 10 min at 37°C (Nanjing KeyGen Biotech Co., Ltd., Nanjing, China; cat. no. KGA215-10), and analyzed using a fluorescence microscope (22). HeLa cells (5×10⁵) were treated with different concentrations of prodigiosin (0, 0.5, 1.0 and 2.0 µg/ml) for 24 h in 6-well plate. The HeLa cells were then harvested, washed with PBS and fixed with 2.5% glutaraldehyde at 4°C for 12 h. Following washing with PBS, HeLa cells were dehydrated with ethanol and acetone, sequentially fixed with 1% osmium tetroxide (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) for 2 h at 37°C, followed by embedding in epoxy resin for 12 h at 37°C. Ultra-thin slices (80 nm) were then produced, stained with 1.0% uranylacetate for 5 min at room temperature and viewed on a transmission electron microscope (23,24).

Since the emission wavelength of prodigiosin was similar to that of propidium iodide, the levels of apoptosis were detected using Annexin V. HeLa cells (5×10⁵) were treated with prodigiosin in a concentration- (0, 0.5, 1.0 and 2.0 µg/ml) and time (24, 36 and 48 h)-dependent manner at 37°C. Following the different exposure times, cells were washed with PBS and resuspended in 150 ml of binding buffer, which contained 5 ml of Annexin V-fluorescein isothiocyanate (Biolegend Inc., San Diego, CA, USA; cat. no. 640912) and were incubated at room temperature for 15 min. The cells were analyzed using a FACScan flow cytometer and the associated software (Becton, Dickinson and Company, Franklin Lakes, NJ, USA).

Western blotting was performed as previously described (25). The human caspase-3 antibody (cat. no. 9662) was purchased from Bioworld Technology, Inc. (St. Louis Park, MN, USA), and the human B-cell lymphoma 2 (Bcl-2; cat. no. 2872), Bcl-2-associated X (Bax; cat. no. 2774) antibodies and β-actin (cat. no. 8457; dilution of all primary antibodies; 1:1,000) were purchased from Bosterbio Biological Technology (Pleasanton, CA, USA). The membranes were washed with Tris-buffered saline with 0.1% Tween-20 (TBST) 3 times and incubated with secondary horseradish peroxidase-conjugated antibodies (Anti-rabbit Immunoglobulin G; cat. no. 7074; Thermo Fisher Scientific, Inc.; dilution, 1:5,000) at 25°C for 2 h. Membranes were then washed with TBST 3 more times and immune reactive protein bands were detected using an enhanced chemiluminescence kit (Beijing Transgen Biotech Co., Ltd., Beijing, China). Image Quant TL 7.0 software (GE Healthcare, Chicago, IL, USA) was employed to quantify protein expression levels. For RT-qPCR, following the supplier's protocol, TRIzol™ reagent (Promega Corporation, Madison, WI, USA) was used to extract the total RNA of HeLa cells that were treated with different concentrations (0, 0.5, 1.0 and 2.0 µg/ml) of prodigiosin, as aforementioned. RNA was quantitated by optical density measurement using a spectrophotometer. The RT-for-PCR Kit (Promega Corporation) was used for RT-PCR. Briefly, 1 µg total RNA with 1 µl 20 µM oligo(dT)18 primers was heated at 70°C for 2 min, then quenched rapidly on ice. The RT reaction was followed by the addition of 4 µl 5X Reaction buffer (250 mM Tris-HCl, pH 8.3, 375 mM KCl, 15 mM MgCl₂), 1 µl dNTP mixture (10 mM each dATP, dGTP, dCTP and dTTP), 20 U recombinant RNase inhibitor and 200 U MMLV reverse transcriptase in a 20 µl reaction volume. Samples were incubated at 42°C for 1 h in an air incubator, followed by inactivation of the reverse transcriptase at 95°C for 5 min. Then, 80 µl nuclease-free water was added to a final volume of 100 µl. RT-qPCR was performed using primer pairs for Bcl-2, Bax and GAPDH (Table I) with the SYBR-Green PCR Master mix (Promega Corporation) with the following thermocycling conditions: 1 cycle at 94°C for 4 min, followed by 20 cycles at 94°C for 30 sec and 68°C for 1.5 min, and finally 1 cycle at 68°C for 5 min. The ABI PRISM 7000 sequence detection system (Applied Biosystems; Thermo Fisher Scientific, Inc.) and the 2^−∆∆Cq (26) method were used for quantitative analysis.

SPSS 19.0 statistical software (IBM Corp., Armonk, MA, USA) was used to evaluate the results. One-way analysis of variance was used for comparisons between multiple groups. For post hoc analyses, heterogeneity of variance was accepted if P>0.05, and the Least Significant Difference method was used. Otherwise, Dunnett's T3 method was used. Homogeneity of variance of the samples to be compared was tested using a Levene's test. The results were expressed as the mean ± standard deviation. P<0.05 was considered to indicate a statistically significant difference.

Observation of morphological characteristics revealed that the colony of strain surface was smooth, convex and shiny with a vivid red color (Fig. 1A). A straight pole shape with blunt ends and uneven surface waves were observed via micromorphological analysis (Fig. 1B). The 16S r DNA sequences of the red colony (comprising 1,562 nucleotides) were obtained and submitted to EzBioCloud (www.ezbiocloud.net/identify). The sequence exhibited the highest similarity (99.59%) to that of Serratia marcescens. A phylogenetic tree was constructed based on the 16S r DNA PCR products as well as on the coding gene sequences of the isolate and its nearest relatives (Fig. 1C). The strain was identified as Serratia marcescens and named WA12-1-18.

The red coating of the solid medium was collected following culture for 72 h at 23°C. The solubility test demonstrated that the dark-red powder had good solubility in polar solvent (MeOH and EtOAc) but rarely dissolved in water (Fig. 2A and B). The yield of prodigiosin was high, reaching 2.77 g/l, and the purity of this secondary metabolite produced by Serratia marcescens was 98.25% determined by HPLC (Fig. 2C). UV spectrophotometry revealed that the maximum absorptions were at 534 nm (pH=5.0) and 466 nm (pH=9.0; Fig. 2D). The FTIR peak assignments showed that the characteristic wave numbers (cm⁻¹) were at 3,450 (N-H str aromatic), 2,924 and 2,853 (C-H₂ str aliphatic), 1,729 (C-O str ether), and 1,631 and 1,581 (C=C, C=N str aromatic; Fig. 2E). The molecular mass determined by LC-MS was 324.207(m/z 324.207; Fig. 2F). The chemical shifts of prodigiosin were further confirmed by NMR spectroscopy (Fig. 2G and H): ¹H-NMR (CDCl3, 300 MHz, ppm) δ 0.98 (3H, t, H11″), 1.29 (2H, m,H9″), 1.45 (2H, m, H10″), 1.59 (2H, m, H8″), 2.38 (2H, t, H7″), 2.47 (3H, s, H6″), 3.31 (1H,m, H3), 4.29 (1H, d, H3′), 4.88 (3H, s, OCH3), 5.10 (1H, brd, H3″), 5.34 (1H, m, H4), 6.37 (1H, brs, H6′),7.18 (1H, m, H2), 7.62 (1H, brs, H1), 7.71 (1H, brs, H1′); ¹³C-NMR (CDCl3, 75 MHz, ppm) δ 12.03 (C6″), 14.59 (C11″), 23.78(C10″), 23.90 (C7″), 26.79 (C8″), 30.94 (C9″), 37.84 (C3′), 59.57 (OCH3), 66.83 (C3), 95.46 (C6′), 108.79 (C4), 112.46 (C5′), 117.04 (C5), 123.67 (C2″), 128.35 (C2), 130.03 (C3″), 131.02(C4″), 132.50(C5″), 154.30 (C2′), and 169.22 (C4′). All characteristics corresponded to those of prodigiosin (C₂₀H₂₅N₃O).

Cell growth of prodigiosin-treated HeLa cells was measured by the MTT assay at different time points over 72 h of the treatment period. The results shown in Table II demonstrated a time- and dose-dependent decline in number of viable cells within 72 h of prodigiosin-treatment. Using the dose-response curves, the half-maximal inhibitory concentration (IC₅₀) values of prodigiosin in HeLa were 2.1 µg/ml at 24 h, 1.2 µg/ml at 48 h and 0.5 µg/ml at 72 h, respectively.

The results of the DAPI staining assay revealed that the number of apoptotic cells increased with the prodigiosin concentration following 24 h (Fig. 3A). The morphology of control HeLa cells at 0 µg/ml prodigiosin was used as the normal morphology (Fig. 3B). TEM revealed rich cell surface microvilli, normal nucleus and abundant cytoplasm; the endoplasmic reticulum can also be seen clearly. However, apoptosis was induced when HeLa cells were treated with 2 µg/ml prodigiosin (Fig. 3B), presenting no cell surface microvilli, folded nuclear membrane, increased nuclear heterochromatin and condensed chromatin. These phenomena suggested that the apoptotic process was induced by prodigiosin treatment.

To gain further understanding of the mechanism underlying the inhibition of proliferation, the present study investigated the effect on apoptosis by flow cytometry. HeLa cells were treated with 0, 0.5, 1.0 and 2.0 µg/ml of prodigiosin for 24, 36 and 48 h, respectively. Annexin V stained cells were tested by flow cytometry for quantitation. There were Annexin V-negative and Annexin V-positive on the dot-plots, as well as the stages of apoptosis located on the upper quadrant of the dot-plot (data not shown). The results shown in Table III demonstrated that there were dose- and time-associated increases in the apoptosis of HeLa cells following prodigiosin treatment.

To further elucidate the mechanism of apoptosis induced by prodigiosin, the present study evaluated the protein and mRNA levels of key regulators via western blotting and RT-qPCR. The results demonstrated that prodigiosin treatment could downregulate Bcl-2 levels, and concomitantly upregulate Bax and caspase-3 levels when compared with the control (0 µg/ml; Fig. 4A and B).