Dye extract of calyces of Hibiscus sabdariffa has photodynamic antibacterial activity: A prospect for sunlight‐driven fresh produce sanitation

Abstract Photodynamic sanitation of fresh produce could help reduce spoilage and disease transmissions where conventional methods of sanitation are not available, and sunlight is available for free. In this study, we evaluated the photostability and photodynamic antibacterial activity of the dye extracts of calyces of Hibiscus sabdariffa. The dye extracts were very photostable in water but bleached in acetate‐HCl buffer (pH 4.6), phosphate buffer saline (pH 7.2), and tris base‐HCl buffer (pH 8.6). The photostability correlated with the photodynamic antibacterial activity of the dye extracts. Both the methanol and water dye extracts at the concentration of 0.0625 mg/ml caused complete inactivation of Bacillus subtilis (reductions of 8.5 log CFU/ml) within 2 min either with the visible light exposure at 10 mW/cm2 or in the dark without the light exposure. Reductions of 4.8 log CFU/ml and 2.2 log CFU/ml of Escherichia coli were observed when 1 mg/ml of methanol and water dye extracts were used, respectively, in water with the light exposure at 10 mW/cm2 for 20 min. Discussions are included about the ease of the dye extractions of the calyces of H. sabdariffa even in water without the need of energy for heating and the suitability of the dye extracts for the fresh produce sanitation. Dye extract of calyces of H. sabdariffa has photodynamic and nonphotodynamic antibacterial activity which could be exploited for the development of a low‐tech sunlight‐driven fresh produce sanitation system that is cheap, sustainable, and environmentally friendly.

However, the main impacts of foodborne diseases are seen in the developing countries because of unsafe water used for cultivating, cleaning, and processing of food, absence of food storage infrastructures, and inadequate/poorly enforced regulatory standards (Chigor et al., 2012;Chigor, Umoh, & Smith, 2010;WHO, 2015). There is little postharvest processing and sanitation of fresh produce in most of the sub-Saharan Africa and this contributed to about 55% and 72% of fresh produce grown to perish before they can be consumed (Adeoye, Odeleye, Babalola, & Afolayan, 2009;Idah, Ajisegiri, & Yisa, 2007;Olayemi, Adegbola, Bamishaiye, & Daura, 2010). Also, ordinarily, when weather and temperature are considered, the rate of fresh produce spoilage in sub-Saharan Africa would be 3-4 times than in the Europe. The daily average temperature in sub-Saharan Africa throughout the year is 30-35°C-this is an optimum temperature range for the growth and proliferation of most microorganisms including fresh produce spoilage and pathogenic bacteria.
Photodynamic effect is the generation of singlet-oxygen and other reactive oxygen species (ROS) by interaction of a photosensitizer, light of appropriate wavelength and molecular oxygen (Costa et al., 2013;Komagoe, Kato, Inoue, & Katsu, 2011;Lukšiene, 2005;Tavares et al., 2011). Photodynamic sanitation of foods is a process in which the generated singlet-oxygen and other ROS inactivate pathogens and reduces microbial load of the food to a level considered safe by the public health standards Lukšiene, 2005;Luksiene & Brovko, 2013). In developing countries including sub-Saharan Africa, there are high prospects of successfully establishing cheap, efficient, and sustainable low-tech photodynamic postharvest sanitation of fresh produce because food-grade photosensitizers can be sourced locally from the plant materials and wastes, and abundant sunlight can be used to irradiate the photosensitizers (Luksienė & Zukauskas, 2009;Majiya et al., 2019).
Hibiscus sabdariffa (also commonly called Roselle) is an annual tropical and subtropical bushy shrub with red or pale green stems and red or pale yellow inflated edible calyces. It is native to west Africa but now grown in other tropical and subtropical countries especially Africa and Asia including Nigeria, Niger, Ghana, Angola, Sudan, Egypt, China, Thailand, Sri Lanka, Malaysia, West India, and also in the Mexico (Abou-Arab, Abu-Salem, & Abou-Arab, 2011;Da-Costa-Rocha, Bonnlaender, Sievers, Pischel, & Heinrich, 2014;Ilori & Odukoya, 2005). The red pigmentation of the calyces is due to the high concentration of anthocyanins. Other chemical constituents of the calyces are organic acids, flavonoids, and other polyphenols. The facts that H. sabdariffa is not a staple crop, safe for usage as food and medicine, available in abundance, easy extraction of its dye, and easily grown in developing countries (Abou-Arab et al., 2011;Da-Costa-Rocha et al., 2014) make it a potential good candidate as a source of dye extract that can be used as a food-grade photosensitizer for the development of a low-tech sunlight-driven sanitation system for foods including fresh produce in those regions of the world that lacks the conventional systems of fresh produce sanitation.
In this work, we evaluated the dye extracts from the calyces of H. sabdariffa for potential usage as a food-grade photosensitizer. We investigated the photodynamic inactivation of model bacteria using the dye extracts.

| Calyces of H. sabdariffa
The dried pinkish red calyces of H. sabdariffa were purchased from the Kure Market, Minna, Niger State, Nigeria. The calyces were used as purchased for the extraction of the dye with no prior preparation and processing.

| Solvents and other consumables
All the chemicals, solvents, and other consumable used were of analytical grade. All inorganic chemicals and salts were purchased from Sigma-Aldrich. Methanol and acetic acid were purchased from Acros Organics. LB broth and agar for culturing bacteria were purchased from Fisher Scientific: Janssen Pharmaceuticalaan. The distilled water used in this work was provided by the Thermo Scientific.

| Light source and irradiation conditions
The light source used for the photostability and Photodynamic inactivation (PDI) experiments was a Slide Projector-Zett Royal II afs (Remote Control Leica Projection GmbH Type 508, Zett Gerate), which provides a cool bright light. A 420-nm light filter was inserted into the projector to cutoff the UV light. Fluence rate of irradiations were measured using a Digital Radiometer (Solar meter, model 9.6 Red light, Solar Tech Inc.). Visible light was used throughout the study and the fluence rates (radiant exposure) were 5-10 mW/cm 2 at 20-22°C under aerobic conditions. The solvents and buffers used for the photophysical characterization of the dye extracts and PDI were distilled water (dH 2 O), methanol, Phosphate Buffer Saline (PBS; pH 7.2, 6.6, 7.6; 10 mM Na 2 HPO4, 1.8 mM KH 2 PO4, 137 mM NaCl, and 2.7 mM KCl), Acetate-HCl (AH) Buffer (pH 3.6, 4.6, 5.6; 0.1 M CH 3 COOH and 0.1 M CH 3 COOK), and Tris Base-HCl (TBH) Buffer (pH 8.6; 0.1 M Tris and 0.1 M HCL).

| Bacterial strains
The bacteria strains-E. coli Nissle 1917 (a model for Gram-negative bacteria) and B. subtilis DB 104 (a model for Gram-positive bacteria)-were from the laboratory of Prof. Ulrich Dobrindt, University of Munster, Germany. Methanol,0.1% (v/v)  The dye extracts were then centrifuged at 11, 500 g for 5 min. After the centrifugation, the supernatants were decanted into fresh clean round-bottom flasks (100 ml). The methanol dye extracts were then concentrated and dried using Rotavapor (BUCHI, Switzerland). The distilled water dye extracts were first frozen using liquid nitrogen and then lyophilized (Christ, Germany) to concentrate and dry. The dried samples were weighed, and the yields of the dye extracts were determined. The samples were then separately dispensed into smaller labeled bottles and stored in the dark under refrigeration for subsequent use.

| Absorption spectra
The absorption spectra of the dye extracts (1 mg/ml) of the calyces of H. sabdariffa were determined in dH 2 O, AH buffer (pH 3.6, 4.6, 5.6 and 6.6), PBS (pH 7.6), and TBH buffer (8.6) using Agilent 8453 UV-visible spectrophotometer (Agilent Technologies) and the wavelength range used was from 300 to 800 nm.

| pH stability in distilled water
The pH of different concentrations of the dye extract (CM) of the calyces of H. sabdariffa and the pH stability were determined and followed for 1 month using a digital pH meter (Seven Compact, Mettler Toledo, USA). The pH of the samples was read and recorded at the beginning of every week until the 4th week.

| Photostability and bleaching
Photostability and bleaching of the dye extract (CM) of the calyces of H. sabdariffa were determined in distilled water, Acetate Buffer (pH 4.6), PBS (pH 7.2), and Tris Base Buffer (pH 8.6) using Agilent 8453 UV-visible spectrophotometer (Agilent Technologies). The absorptions (300-800 nm) of the samples were read intermittently every 2 min of the irradiation at 5 mW/cm 2 until 20th minutes.

| Mass spectroscopy of the dye extracts of the calyces of H. sabdariffa
Mass spectroscopy of all the dye extracts was determined in methanol and water using MicroTof-ESI (Bruker Daltonics) and Orbitrap LTQ XL-ESI (Thermo Fisher Scientific) in order to assess the effect of the temperature conditions and solvents used for the extractions on the compositions of the extracts. Additionally, the fractionization and mass spectroscopy of the dye extract (CM) was done using 6110 HPLC-MS-ESI (Agilent Technologies). In all cases, both positive and negative ionization modes of the mass spectroscopy were acquired. two control experiments were simultaneously setup; dark experiments which were carried out in the presence of dye extracts but without illumination, while "light only" experiments were exposed to light but in the absence of the dye extracts. All experiments were repeated three times. After PDI experiments, the solutions were plated for bacterial enumeration to determine the rate and extent of inactivation. Also, live/dead staining (BacLight Bacterial Viability Kit, Invitrogen) was used according to the manufacturer's instructions to determine the viability of the bacteria after the PDI. Fluorescence images were captured using 40× magnification objective with a Nikon Eclipse Ci microscope (Nikon Corporation Tokyo).

| Effects of the temperature and solvents on the yields and pH of the dye extracts of the calyces of H. sabdariffa
The yields of dye extractions of the calyces of H. sabdariffa by different temperatures and solvents are shown in Table 1. When solvents are considered, more higher yields were observed in water dye extractions of the calyces ( Table 1). As for the temperatures of the extractions, hot temperature gave higher yields compared with the cold temperature (Table 1). Although all the dye extracts irrespective of the solvents and temperatures of extractions are very soluble in water, the pH varies slightly in dH 2 O at the same concentration (Table 1). The methanol dye extracts have lower pH when compared with the water dye extracts (Table 1).

| Effect of buffers/solutions on the absorption spectra and colors of the dye extracts
The absorption spectra (Figure 1) of the dye extracts (1 mg/ml) of the calyces of H. sabdariffa were determined in distilled water (dH 2 O), Acetate-HCl (AH) buffer (pH 3.6, 4.6, and 5.6), Phosphate Buffer Saline (PBS; pH 6.6 and 7.6), and Tris base-HCl (TBH) buffer (8.6). The peak absorption for all the samples in dH 2 O was at 520 nm ( Figure 1a).
However, the intensity of the absorptions differs, and samples extracted under cold conditions seem to have higher intensity compared with the samples extracted at hot and boiling conditions ( Figure 1a).
Also, although the absorption peaks for all samples were at 520 nm, the spectrum pattern of the sample extracted under boiling condition (BW) slightly differs with other samples (Figure 1a). The absorption spectra of the dye extract (CM) in different buffers and pH have different absorption peaks and colors (Figure 1b,c). The absorption peaks tend to move toward infrared as the pH increases ( Figure 1b). The peaks for pH 3.6, 4.6, 5.6, 6.6, 7.6, and 8.6 were at 530, 538, 536, 559, 583, and 592 nm, respectively (Figure 1b). The changes in colors with different pH are characteristics of the dye anthocyanins. The pH stability of different concentrations of dye extract (CM) in dH 2 O for 1 month is shown in Table 2. The dye extract is very stable in water (Table 2).

| Effects of buffers/solutions on the photostability and bleaching of the dye extracts
The absorption spectra (

| Mass spectroscopy of the dye extracts of the calyces of H. sabdariffa
The mass spectroscopy of the dye extracts was determined in order to assess the effects of extraction temperatures and solvents on the compositions of the extracts. The first three peaks (for positive and negative ionization modes) with the highest intensities for each of the samples are shown in Table 3 (see Figure S1A-G). The peaks indicated that the temperature and solvent of extraction did not affect the compositions of the dye extracts (Table 3). However, boiling at 100°C might have affected the compositions slightly (Table 3). The mass spectroscopy of the fractions of the dye extract (CM) showed it is a complex compound with more than 40 compounds (see Figure   S2A and B).

| Effect of the concentration and time of illumination on PDI using the dye extracts
The PDI of E.coli and B. subtilis using different concentrations of the

| Extraction and photophysical characteristics of the dye extracts of calyces of H. sabdariffa: implications for applications
In this work, we extracted the dye from the calyces of H. sabdariffa using different solutions; distilled water, methanol and acidified  (Table 1). The water dye extractions had higher yield than the methanol and acidified methanol extractions (Table 1). However, at the same concentration (1 mg/ml) the water dye extracts were a little bit less acidic (pH 3.0) when compared with the methanol (pH 2.85) and acidified methanol (pH 2.76) extracts (Table 1).

F I G U R E 5
Fluorescence microscopy images of live/dead staining of the PDI samples of E. coli when 1 mg/ml of the dye extract-CM was used in water. CM, cold methanol dye extract. L + CM is the light experiment (dye extract with light illumination at 10 mW/cm 2 for 20 min); D + CM is the dark control experiments (dye extract without light illumination for 20 min); and L is "light only" control experiment (no dye extracts but illuminated with light at 10 mW/cm 2 for 20 min)

F I G U R E 6
Fluorescence microscopy images of live/dead staining of the PDI samples of B. subtilis when 1 mg/ml of the dye extract (CM) was used in water. CM, cold methanol dye extract. L + CM is the light experiment (dye extract with light illumination at 10 mW/cm 2 for 20 min); D + CM is the dark control experiments (dye extract without light illumination for 20 min); and L is "light only" control experiment (no dye extracts but illuminated with light at 10 mW/cm 2 for 20 min) These pH (2.76-3.0) can inactivate many of the bacteria especially the Gram-positive bacteria. The pH of the acidified methanol dye extract we observed agrees with the earlier reported (Abou-Arab et al., 2011). Using the same solvent but different temperatures for the dye extracts does not significantly affect the % yields and the pH. All these indicated the ease at which the dyes of calyces can be extracted even in cold water with no additional need for energy for heating thereby making it suitable for the aim of this project.
The absorption and mass spectra (except for the dye extract-BW) indicated that irrespective of the solvents and temperatures of extractions, the dye extracts are the same (Figure 1a and Table 3). The absorption peak for the dye extracts is the same which is 520 nm.
The dye extract of the calyces of H. sabdariffa is very photostable and pH stable in water ( Figure 2a and Table 2). This means that the dye extract has fulfilled one of the cardinal desired characteristics of a photosensitizer for the lack of photodegradation and bleaching in water (DeRosa & Crutchley, 2002). However, the dye extracts bleaches in other buffers (acetate-HCl buffer pH 4.6, PBS pH 7.2, and tris base-HCl pH 8.6) used (Figure 2b-d). The bleaching may be pH and/or buffer type/compositions dependent. This may limit the application of the dye extracts as a photosensitizer to only environmental applications such as fresh produce sanitation where fresh water would be used as wash waters containing the dye extracts to photodynamically inactivate pathogenic and spoilage microorganisms. The dye extract may not be suitable for the PDI in biological samples and solutions/buffers of high salts contents. However, the non-PDI inactivation activity of the dye extracts could still be available or retained and could be exploited for the biological samples and solutions/buffers.

| Dye extracts of the calyces of H. sabdariffa cause photodynamic and nonphotodynamic inactivation of E. coli and B. subtilis
The water and methanol dye extracts of the calyces of H. sabdariffa used in this work showed both photodynamic and nonphotodynamic inactivation of bacteria models (Figures 3 and 4). Singlet-oxygen and other ROS oxidize and cause irreversible damage to proteins, lipids, nucleic acid, and other cellular components of microorganisms and ultimately inactivate them (Costa et al., 2013).   (Figures 3 and 4). B. subtilis used in this study was particularly sensitive to acetate-HCl buffer (pH 4.6) and the buffer with its low pH might have induced some molecules within the bacterium to act as photosensitizers even without the addition of the external photosensitizers to generate ROS which inactivated it (Figure 3a,b).
Like most antimicrobial agents, the PDI and non-PDI activity of the dye extracts of calyces of H. sabdariffa shown in this work are dye dose/concentration and illumination time dependent (Figure 4).

Previous reports have shown that the PDI is proportionally related
to the light dose/fluence rate (Costa et al., 2010;Majiya, 2017). The light intensity used for the PDI in this work is 10 mW/cm 2 which is just about 1% of the bright sunlight under clear sky conditions in sub-Saharan Africa. In this work, under the same conditions, more B. subtilis-a Gram-positive bacteria was inactivated compared with the E. coli-a Gram negative. This is a general trend for most of the antibacterial agents including the PDI. Gram-positive bacteria are easily inactivated by singlet-oxygen oxidation as compared with gram-negative bacteria due to differences in the structure of their cell membrane (Alves et al., 2009;Bourré et al., 2010;Carvalho et al., 2007;Costa et al., 2012;Komagoe et al., 2011;Maisch, Spannberger, Regensburger, Felgenträger, & Bäumler, 2012). Gram-negative bacteria have an additional outer membrane apart from the cytoplasmic (inner) membrane, giving them extra protection against antimicrobial agents including singlet-oxygen and other ROS produced during photosensitization.

| CON CLUS ION
In summary, we observed that the photostability, photodynamic, and nonphotodynamic antibacterial activity of the water and methanol dye extracts of the calyces of H. sabdariffa were dependent on the solutions/buffers, dye dose/concentration, and illumination time. The dye extracts were very photostable in distilled water. However, the dye extracts bleached and photodegraded in acetate-HCl buffer (pH 4.6), PBS (pH 7.2), and tris base-HCl buffer (pH 8.6). This will limit the PDI use of the dyes to only environmental applications such as fresh produce sanitation in fresh water. The PDI of bacteria was clearly observed in distilled water because the dye extracts were photostable in the water. To the best of our knowledge, this study will be the first time that photodynamic antibacterial activity of the dye extracts of calyces of H. sabdariffa is reported. Rate and extent of the PDI and non-PDI was higher in B. subtilis compared with the E. coli.
This study can lead way to the development of a low-tech sunlight-driven fresh produce sanitation system that could be used in developing countries especially the sub-Saharan African to reduce foodborne diseases/outbreaks and fresh produce spoilage/ wastage.

This work was supported by TWAS-DFG Cooperation Visits
Programme 2019 (DFG-GA2362/3-1 to HM). The authors would like to thank members of the Galstyan and Dobrindt laboratories at the University of Munster, Germany.

CO N FLI C T O F I NTE R E S T
None.

E TH I C A L CO N S I D ER ATI O N
Ethics approval was not required for this research.