Benefits of Soleris^® over the Conventional Method for Enumeration of Microbial Load in Salacia Herbal Extract

Salacia, a genus is one of the medicinal plants group broadly used to treat many ailments viz. hypoglycemia, hypolipidemia, inflammation, and diabetic patients. Salacia is represented by 21 species in India, of which 15 species occur in Peninsular India with S. reticulata and S. oblonga as predominant species (Bagnazari et al. 2017). The root of Salacia spp. (S. chinensis, S. reticulata, S. oblonga) is one of the preferred drug sources for treating diabetes in the indigenous systems of medicine. Salacia species contain two active compounds salacinol and kotalanol, which are reported to exhibit α-glucosidase inhibitory activity (Akaki et al. 2014). Apart from its anti-diabetic potential, Homma and coworkers (2019) have attributed skin whitening capabilities to salacinol with low cytotoxicity.

A wide range of diseases affecting humans has been treated since ancient times with medicinal plants, as extracts, individually or as polyherbal preparations. A survey by the World Health Organization (WHO) has indicated that in the developing countries, nearly 70–80% of the population prefers to use non-conventional herbal medicines for their primary health care (Abba et al. 2009). The herbal medicines are safe, natural, relatively accessible, and cheaper than the synthetic drugs.

Factors that influence the contamination of medicinal herbs are several. These include certain environmental factors, mainly humidity, rainfall, storage conditions of crude and processed medicinal-plant materials, and handling and hygiene of people handling the herbs. Also, microorganisms derived from soil, air and water may contaminate the herbal raw materials leading to pathogenic effects to humans (Alonzo et al. 1994; de Freitas Araújo and Bauab 2012). The presence of such pathogens limits the use of medicinal plants and also exerts an important impact on the overall therapeutic quality of herbal drugs and preparations. Also, since the presence of such pathogenic microorganisms constitutes a potential hazard to human health, reducing their concentration from their source is critical.

Escherichia coli and Pseudomonas spp. are the typical pathogens seen in herbal extract powders. Fungal species such as Rhizhopus, Penicillium, Aspergillus are also reported (Hitokito et al. 1978; Kneifel et al. 2002). Although bacterial endospores and fungal spores are considered as the two dominating groups of contaminants seen on medicinal plants, the presence of pathogenic bacteria like B. cereus, Aeromonas hydrophila, Shigella spp., Enterobacter agglomerans, E. cloacae, Vibrio fluvialis, Pasteurella multocida, S. epidermidis, Acinetobacter iwoffii, Klebsiella spp., and B. subtilis have been reported in plant samples analyzed recently (Idu et al. 2011).

Assessment of medicinal plants’ microbial load such as Matricaria chamomilla, Achillea millefolium, Ocimum basilicum, Calendula officinalis, and Tilia cordata, Hypericum perforatum, and Salacia has been reported (Kumar et al. 2015; Oprea et al. 2015).

The most widely used technique for a total count of microorganisms in plant materials is a technique recommended by the WHO. In this methodology, 10 g of sample is recommended to be suspended in 90 ml of buffer sodium chloride-peptone of pH to 7.0. Suitable dilutions of the sample is plated on casein-soybean digest agar and incubated at 30–35°C. The total aerobic count is measured after 48 h. For yeast and molds, the technique employed is the sowing depth in Sabouraud dextrose plus a solution of 10% tartaric acid to obtain pH 3.0 to 3.5 and incubated at 20–25°C for five days (WHO 1998). The specification of the WHO for total aerobic microorganisms is not more than 10⁷ CFU/g, while the specification of the WHO for yeasts and molds is at most 10⁴ CFU/g. Both the Brazilian Pharmacopeia (ANVISA 2010) and the United States Pharmacopeia (USP 2005) have recommended specifications for oral use: 104 aerobic bacteria/g or ml and 102 fungi/g. The European Pharmacopoeia (EDQM 2007) also gives guidance on acceptance criteria (5.1.8 for herbal medicinal products for oral use and 5.1.4 for other herbal medicinal products). The limits of microbial contamination given in European Pharmacopoeia for herbal medicinal products to which boiling water is added before use are total aerobic bacteria (10⁷ CFU/g), fungi (10⁵ CFU/g) as against 10⁵ CFU/g), fungi (10³ CFU) when water is not added before use.

In Ghana, around 65% of the population depends on herbal medicine and hence estimation of the accurate microbial load becomes essential in such cases to avoid health issues to herbal users (Agyeman-Duah et al. 2017). Herbal medicines in liquid form also have also been shown to have microbial contamination, and hence determination of quantitation of microbial load becomes critical (de Sousa Lima et al. 2020). In this paper, we describe the ease of detecting microbial load in Salacia extract powder using Soleris^® instrument and strongly believe the applicability of this method to all herbal extracts.

The Soleris^® system consists of an incubator, ready-to-use vials, and system software that are 21 CFR complaints. The system’s flexibility is appreciable since 1 to 512 samples can be tested simultaneously, and up to four instruments can be connected to a single PC, allowing any combination of 32 and 128 sample units. The Soleris^® instrument with a temperature-controlled incubator is equipped with a photodiode-based optical detection system that allows the growth of microorganisms when inoculated in incubation vials containing selective growth media, supplements, and substrates specific to the microbial species to be detected. The carbon dioxide produced metabolically by the growing microbes results in a color change as metabolic processes happen and it causes a change in pH and denotes a positive detection time (DT). The samples with a higher level of microbial contamination would show a faster DT, and hence this instrument gives a correlation between microbes present and DTs reliably.

The inoculated Soleris^® vials are placed into the selected drawer location and experiments are initiated. The Soleris^® software indicated positive test results in less than 24 hrs. Determinations producing no detection within 24 h were considered negative. In the case of positive results, the growth curves were evaluated, and the visual validation of medium color change was also carried out Fig. 1. In the case of positive results, confirmation was done using conventional methods. For yeast and mold, the completed system results are available after 72 h, although positive samples were indicated after three rises in CO₂ levels at the preset threshold, usually within 48 h, whereas the yeast and mold detection plates required an incubation of four days and five or more days respectively.

Fig. 1.

The sensor in the Soleris^® Non-Fermenting-Total Viable Count (NF-TVC) vial system utilizes detection of carbon dioxide, a universal bacterial metabolite, rather than detection of acid production, expanding the inclusivity of the vial to include non-fermenting organisms with a sensitivity of 1 CFU (Alles et al. 2009; Mozola et al. 2013). Optical readings from Soleris^® test vials are graphed with time (h) in X-axis vs. optical units in the Y-axis. Table I gives the values of total viable count achieved through Soleris^® and its comparison with the bacterial CFU enumerated by conventional method plating.

Fig. 2a shows the presence of bacteria for all dilutions. It is inferred through the observation of the bacterial growth curve above the baseline with the count of ~1×10³ CFU/g, a value that corresponds well with the reports of Kumar et al. (2015). The dilution which does not detect any microbes shows a curve touching the baseline while a blue curve indicates the dilutions where bacteria is detected. On the other hand, the red curve indicates the highest amount of organisms present in that particular or lowest dilution. The conventional method for the SR011903 sample yielded 1.09 × 105 CFU/g, as evident from Table II. The conventional method for the SR011903 sample yielded 1.09 × 10⁵ CFU/g as evident from Table II. The microbial count of the Salacia extract after steam sterilization (SS), ETO sterilization (ET), and sterilization by gamma irradiation (GR) showed 50%, 0%, and 75% reduction respectively in comparison to the untreated control, and the values matched well when estimated by the conventional method (Table II). The CFU ranges indicated in Table II were estimated using the Soleris^® system and from the traditional plate counts done in parallel. To compare the sensitivity of the Soleris^® method over the conventional method, we opted for calculation of the relative level of detection (RLOD) between the two methods as described (Mărgăritescu and Wilrich 2013). If the RLOD value was below 1, it indicated that the Soleris^® method was more sensitive than the conventional plating method while the value was between 1 and 1.5, indicated both the methods to have a similar sensitivity. The RLOD value of above 1.5 indicated that the plating method was more sensitive. It is clear from Table II that the RLOD values of Soleris^® in almost all the cases were sensitive over the traditional method of estimation by plating. The growth curve of sample after SS, ET and GR is represented as Figs. 2b, 2c, and 2d, respectively.

Fig. 2. a)

Fig. 2. b)

Fig. 2. c)

Fig. 2. d)

The Salacia’s microbial load is known to increase upon storage (Kumar et al. 2015), and gamma irradiation has been reported to reduce the microbial load by almost 2 logs. Similar results have been shown by Gupta et al. (2011) for plants such as Terminalia chebula, Curcuma longa, Syzygium aromaticum, and Mentha piperita.

Our results shown in Table II do indicate differences in efficiencies of microbial killing for the Salacia extract with different agents. While gamma irradiation and steam sterilization showed promising microbial load reduction, the ETO treatment did not show any reduction in the microbial load. This could be attributed to the type of microorganism present in Salacia extract and the insufficient ETO dose required to efficiently kill resident microbes. Interestingly, differences in efficiency of killing of microorganisms contaminating the herbal extracts has been shown to depend on the types of contaminating organisms, their chemical and physical structure, antiseptic properties of different plants, and their ability to recover from the radiation injury (Gupta et al. 2011), hence, the dose required for efficient microbial killing could be different for different plants.

It is evident from the current observations that none of the traditional methods are efficient enough to reduce the microbial load in Salacia, and hence, our observations of reductions in microbial counts in sample RDP/SR/134 by almost 2 logs through the use of an antibacterial agent (proprietary, patent-pending) appear promising and can be taken up for trials in manufacturing scales.

The sterilization methods using radiations are not readily adaptable for manufacturing plant processes since it requires special equipment and methods. The other methods, such as ethylene oxide, are banned in European Union countries due to the generation of carcinogenic substances such as ethylene glycol, 2-chloroethanol. Also, steam sterilization is not often used for herbal materials since the treated materials become clumped after steam treatment (Brodowska et al. 2014). A new alternative Electron-beam (E-beam) technology, is being used in the food industry to decontaminate food materials (Silindir and Özer 2009). Since installations of such specialized equipment would require investments, alternative methods that are cost-effective to sterilize materials, such as what we describe here, appear to be an attractive proposition.

It is also clear from Table II that the commercially available Salacia extract (sample ID RDP/SR/136) showed a TVC of > 1,000 CFU/g. The growth curve seen for TVC for samples RDP/SR/068, RDP/SR/134, and RDP/SR/136 is shown in Figs. 2e, 2f and 2g, respectively. Sample RDP/SR/068 showed microbial detection in all the dilutions (10^–3 and 10^–2) which accounts for ~ 1,600 CFU/g obtained by plating one of the dilutions while sample RDP/SR/134 showed no bacterial growth in any of the dilutions (10^–3 and 10^–5) tested, and showed merely 20 CFU/g from 10^–1 dilution reflecting the efficacy of the antibacterial agent in lysing the microbes contaminating the Salacia extract. RDP/SR/136 sample showed nearly 1,500 CFU/g as seen with the 10^–2 dilution in NF-TVC Soleris^® vials denoted by the red curve (Fig. 2g).

Fig. 2. e)

Fig. 2. f)

Fig. 2. g)

It is evident from Table III that the Soleris^® can detect yeast and molds when present within 24 h while it takes more than 5 days when tested by conventional methods. It is also clear that when the extract was pre-treated with the proprietary antifungal agent, the material (RDP/SR/068) showed > 99% reduction in the count of yeasts and molds (Table IV). The Soleris^® Direct Yeast and Mold (DYM 109-C) method offers huge time savings for labs engaged in microbial identification in herbal powders. While the yeast and mold results are accessible within 48 h by Soleris^®, the conventional methods take up to 5 days. None of the samples tested here showed a growth curve for yeast and molds with the least dilutions except for SR011903 and RDP/SR/136, represented by Figs. 3a and 3b, respectively. Supplementary figures represent the growth curve of other samples, namely SR/011903, RDP/SR/070/SS01, RDP/SR/070/ES02, RDP/SR/070/GR03, RDP/SR/068 and RDP/SR/134, Figs. S1a, S1b, S1c, S1d, S1e, and S1f, respectively. It is interesting to emphasize here that the total yeast and mold counts of the Salacia extract enumerated using Soleris^® and conventional method of ~ 290 to 380 CFU/g matches well with the reported values of yeast and molds in Salacia by other workers (Kumar et al. 2015) demonstrating the sensitivity and accuracy of the Soleris^® method of microbial enumeration in the herbal extract of Salacia as tested here.

Fig. 3. a)

Fig. 3. b)

As far as the total coliforms counts were concerned, we found similar results of Solaris^® against the conventional method, indicating the specificity of the tests employed. None of the samples showed any count for total coliforms both by conventional method and by Soleris^® test as evident from Table SIa and SIb (supplementary tables). The growth curve pattern on Soleris^® also did not show any growth-related optical unit readings (data not shown).

Health surveys conducted in several countries have demonstrated the use of herbal medicines as a mainstream practice among the elderly population, and hence the risk of microbial contamination in herbal drugs might have an adverse effect on the health of such population (de Medeiros et al. 2013; Famewo et al. 2016). Several pathogenic bacteria have been detected in some of the herbal medicines, which is also a severe concern for such medicines’ quality issues (Abba et al. 2009; de Medeiros et al. 2013). Hence, it becomes imperative to estimate microbial load in such medicinal herbs and extracts so that such material consumption is safe without any side effects.

The killing of microorganisms contaminating the herbal extracts is dictated by the dose rate of the antimicrobial agent employed and also by the physical state of the growth of the microbial cells, and other external factors, such as oxygen, water, other chemical agents, and temperature (Halls 1992). Hence, our current data on achieving a microbial reduction of > 99% in Salacia is attractive, and obtaining such data from a sensitive instrument like Soleris^® assumes critical importance. The Non-Fermenting Total Viable Count (NF-TVC) and the Direct Yeast and Mold (DYM) test methods have been used for various food matrixes, and these have been validated as alternatives to conventional plating techniques (Limberg et al. 2016). The advantages of these growth-based automatic measurements in contrast to conventional plate-counting methods include precision, accuracy, reproducibility, speed, and cost (Blivet 2014; Curda and Svir’akov’a 2014).

Microbial contamination in herbal extracts is one of the issues addressed by the new FDA regulations, and faster and more streamlined microbiological tests are therefore required by the herbal industry to meet the new challenges. Soleris^® TVC method has been successfully used for rapid and accurate detection of microorganisms in a variety of food commodities, and its technology is based on monitoring pH change or CO₂ production as a means of microbial growth activity in media vials. The performance of the Soleris^® TVC method in detecting microbes in herbal extracts is the first report to date. The sensitivity of the Soleris^® method was found to be comparable to that of the reference procedures for the detection of microorganisms. The Soleris^® test from Neogen is approved from the AOAC Research Institute as a Performance Tested Method certification 071203 for detecting microorganisms in as little as 4 h, and only takes 24 h to a negative result. The Soleris^® vial is substituted for the agar plate. The sensitivity of the equipment is 1 CFU/g. The significant reduction in the generation of biomedical lab waste concerning Solaris^® compared to the conventional methods is really attractive. Since Solaris^® system reads the vials photometrically by looking into the color change relative to the starting color, it would be the choice of method for all researchers in academics and industries alike who are engaged in determining the microbial contamination in food and other natural products.