Control of industrially relevant microbial isolates by antimicrobial agents: Implications for sugar factories

Abstract

 

Microbial isolates from sugar crop processing facilities were tested for sensitivity to several industrial antimicrobial agents to determine optimal dosing. Hydritreat 2216 showed broad-spectrum activity against all bacterial isolates as well as Saccharomyces cerevisiae. Sodium hypochlorite showed broad-spectrum activity against all isolates, but at much higher effective concentrations. Hops BetaStab XL was effective against Gram-positive isolates. Magna Cide D minimum inhibitory concentration was lowest for S. cerevisiae and Zymomonas mobilis but was less effective against Gram-positive bacterial strains. Based on laboratory experiments, factory losses of sucrose from a single microbial species in the absence of antimicrobials could range from 0.13 to 0.52 kg of sucrose per tonne of cane. Additional improvements in sugar yield are anticipated from agents with broad-spectrum activity. A cost analysis was conducted considering sucrose savings due to antimicrobial application to provide estimates for break-even costs, which ranged from approximately $0.50 to $2.00/L for a given antimicrobial agent.

One-Sentence Summary

Application of antimicrobial agents at minimal inhibitory doses for microbes results in optimal inhibition of microbial growth and sucrose consumption.

Graphical Abstract

Graphical Abstract.

Introduction

Normal table sugar is produced from sugar beet and sugarcane. Processing sugarcane involves the initial extraction of sugar-containing juice (crusher or first juice) from sugarcane billets (short sections of stalk) followed by several stages of re-extraction of the solids (cane fiber) with water. This secondary extract is combined with the crusher juice to form mixed juice, which undergoes heating, pH adjustment, clarification, evaporation, and crystallization. The process is complex and varies slightly from factory to factory; however, it is assumed that microbial-derived sucrose loss occurs before the mixed juice is heated. Microorganisms are inoculated into sugar crop processing streams from microbial flora of sugarcane billets from the harvest field, cane pile, and associated soil. These microorganisms carry over into the juices during cane milling, resulting in sucrose losses and formation of bacterial exopolysaccharides (EPS) (Nel et al., 2019). In addition to sucrose losses, the bacterial EPS cause operational problems and lost productivity (Chen & Chou, 1992; Eggleston, 2002; Solomon, 2009). Collectively, these processing issues have an economic impact. Various estimates have been made for how much sugar is lost due to microbial impact. Tilbury et al. (1977) stated that typical sugar losses range from 1 to 2.5 kg of sucrose per metric ton (tonne) of cane ground and they reported that 62% of those losses were related to microbial activity (Tilbury et al., 1977) while others reported that 93% of sucrose losses in laboratory experiments were attributable to microbial activity (Eggleston, 2002). Furthermore, undetermined losses can represent approximately 0.4% of the incoming sucrose sugar (Chen & Chou, 1992).

The sugarcane juices and resulting biofilms present in raw sugar factories during processing, are polymicrobial in nature (Qi et al., 2023). Furthermore, the microbial composition and abundance may vary across time, factory, geographic location, and with different weather and soil conditions (Nel et al., 2019, 2020; Qi et al., 2023; Terrell et al., 2022). Given the dynamic nature of sugar crop processing streams, it can be difficult to assess the effectiveness of antimicrobial agents for mitigation of microbial activity across a wide sampling of different juices. Typical reports on the efficacy of antimicrobial agent applications are done on juices, with unknown microbial load and composition making it difficult to assess whether the most problematic microbes are susceptible to the treatment (Boone et al., 2017; Eggleston, 2002). Furthermore, during sugarcane processing, there is a very short actual residence time for treating juices, indicating that antimicrobial control agents must act quickly on microorganisms in juices (Boone et al., 2017, 2016).

We previously reported on the microbiome of sugarcane juices and biofilms from Louisiana raw sugarcane factories showing the most abundant bacterial genera typically found in juices and biofilms including Leuconostoc, Weissella, Lactobacillus, Zymomonas, and Pantoea and other members of the Enterobacteriaceae, and noted differences in relative abundance among the factories sampled (Qi et al., 2023). Additionally, the microbial community in juices likely changes according to truckload, field, storage conditions, and weather (Misra et al., 2020; Solomon, 2009).

Previous isolation studies from sugar crop factories have yielded numerous microbial isolates for further study and testing (Bruni et al., 2022; Qi & Bruni, 2023). While antimicrobial susceptibility testing is most often performed in the context of clinical microbiology or during food processing (Al-Kadmy et al., 2023; Gajic et al., 2022), we applied this technique to determine the minimal concentrations of oxidizer or antimicrobial agents often termed by the sugar industry as “biocides” that are most likely to inhibit the growth of microbial isolates typically found in juices or biofilms and therefore reduce sucrose losses. Moreover, several of the microbial strains in this study also produce some kind of EPS, which often cause various operational problems during raw sugar manufacturing (Solomon, 2009; Tallgren et al., 1999). In particular, the Leuconostoc and Weissella strains produce dextrans while the Gluconobacter and Zymomonas strains produce levan fructan EPS (Bruni et al., 2022, 2024; Silbir et al., 2014).

To our knowledge, this is a first-of-its-kind study for the sugarcane industry, in which we tested the hypothesis that antimicrobial agents can be used to effectively control microbial load and mitigate sucrose consumption when applied at an effective dose. There is an inherent novelty in testing industrially relevant, sugar factory-derived microbial isolates for susceptibility—instead of sugarcane juice with unknown microbial load and composition—to various antimicrobial agents for determination of minimum inhibitory concentration (MIC) values. Additionally, the effect of antimicrobial agents on sucrose consumption was measured and used to conduct break-even cost analysis.

Materials and Methods

Microbial Isolation and Identification

Microbial strains isolated from Louisiana sugarcane factories were anonymized and named for LA Sugar Mill (LASM#) for publication. Additionally, one biofilm isolate from a sugar beet factory biofilm was named for the stakeholder, Beet Sugar Development Foundation (BSDF), from which Leuconostoc BSDF2-3 originated (Qi et al., 2024). Strains used in this study are listed in Table 1. For isolation and identification of isolates not previously described, including LASM77 and -91, crusher juice samples from Louisiana raw sugar factories, which requested not to be named, were serially diluted with sterile ultrapure water and spread on nutrient agar plates (Research Products International, Mt. Prospect, IL, USA). For isolation of LASM131, crusher juice was serially diluted with sterile ultrapure water and spread on CJ+ agar (Bruni et al., 2022) plates, followed by one round of isolation on Tryptone Glucose Yeast (TGY) medium (Haynes et al., 1955), followed by three rounds of streaking on nutrient agar or DeMan, Rogosa and Sharpe (MRS) agar plates to obtain axenic cultures. All culturing was done aerobically at 28°C. After three rounds of streaking to obtain axenic cultures, an isolated colony was picked and resuspended in 50 µL of autoclaved ultrapure water to prepare a PCR template. The 16S rRNA gene was PCR-amplified with universal primers 27F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1492R (5′-GGTTACCTTGTTACGACTT-3′) using Extaq Hot Start DNA polymerase (Takara). PCR products were purified with Zymo Clean and Concentrator-5 kits (Zymo Research, Irvine, CA, USA) and shipped to Eurofins Genomics, LLC, for Sanger sequencing with primers 27F and 1492R. The chromatogram files were imported into Geneious Prime Version 2023.0.1, which was used to trim the primer-binding regions, assemble the forward and reverse reads, and BLASTn to query the consensus sequences against the NCBI 16S RefSeq database to identify the nearest related organism.

Table 1.

Strains Used in This Study

Strain #	Microorganism	Source Description or Notes	Reference
LASM5	Leuconostoc suionicum	Sugarcane factory mixed juice	Qi and Bruni (2023)
LASM6	Leuconostoc suionicum	Sugarcane factory mixed juice	Qi and Bruni (2023)
LASM7	Leuconostoc suionicum	Sugarcane factory mixed juice	Qi and Bruni (2023)
LASM16	Leuconostoc lactis	Sugarcane factory mixed juice	Qi and Bruni (2023)
BSDF2-3	Leuconostoc sp.	Sugar beet factory biofilm	Qi et al. (2024)
NRRL B-1064	Weissella confusa	McClesky, C.S., LSU, isolated from sugarcane, USA	Collins et al. (1993); Holzapfel and Kandler (1969)
LASM91	Chryseobacterium aureum	Sugarcane factory crusher juice	This study
LASM77	Chryseobacterium bernardetii	Sugarcane factory crusher juice	This study
LASM22	Pantoea dispersa	Sugarcane factory crusher juice	Qi and Bruni (2023)
LASM131	Pantoea conspicua	Sugarcane factory crusher juice	This study
LASM12	Gluconobacter japonicus	Sugarcane factory crusher juice	Qi and Bruni (2023)
LASM56	Saccharomyces cerevisiae	Sugarcane factory pre-diffuser biofilm	Bruni et al. (2022)
NRRL B-806	Zymomonas mobilis	Also known as ATCC 10988; type strain	Silbir et al. (2014)

Sourcing of Antimicrobial Reagents

Sodium hypochlorite and thyme oil were purchased from Sigma-Aldrich (St. Louis, MO, USA). Hydritreat 2216 (22% peracetic acid) was generously provided as a sample from Hydrite Chemical Co. (Brookfield, WI, USA). Hops BetaStab XL (9% hops acid emulsion) (Di Lodovico et al., 2020; Pollach et al., 2002; Ting & Goldstein, 1996) was generously provided by Betatec (Washington, DC, USA) and Magna Cide D, a proprietary carbamate formulation, was generously provided by Protech USA (Thibodaux, LA, USA).

Susceptibility Testing of Antimicrobial Agents by the Microdilution Technique

Susceptibility testing of microbial strains to antimicrobial agents was adapted from the microdilution technique (Jorgensen & Ferraro, 2009). Antimicrobial agents were suspended in Tryptone Sucrose Yeast (TSY) medium (50 g/L of sucrose) (Haynes et al., 1955) and twofold serially diluted in 96-well plates in TSY medium so that each well contained 100 μL. Once 100 μL of inoculum was added to each 96-well plate, the final volume was 200 μL and this resulted in a working concentration range from 0 to 1,000 parts per million (ppm) of antimicrobial agent. Precultures of microbial strains were started in TSY for Gram-negative bacteria and for Saccharomyces cerevisiae while Gram-positive bacterial cultures were grown in MRS medium (DeMan et al., 2003) to minimize EPS production and allow for pipetting nonviscous precultures. Optical densities (OD₆₀₀) of the microbial precultures were measured in an Eppendorf BioPhotometer (Hamburg, Germany) and normalized to 0.1 in TSY medium. Each 100 μL of inoculum was added to triplicate wells in 96-well plates already containing serially diluted antimicrobial agents so that the starting OD₆₀₀ of each strain in the experiment was 0.05. Strains used in this study are listed in Table 1. Microwell plates were sealed with sterile breathable Aeraseal membranes (Excel Scientific Inc., Victorville, CA, USA) to prevent spilling between wells and incubated at 28°C, 250 rpm, for 18 hr. Eighteen hours was selected as a reproducible endpoint in this conventional research method, at which time OD₆₀₀ was measured in a Sunrise plate reader (TECAN, Männedorf, Switzerland) with Magellan software, Version 7.5. Lack of growth was used to estimate the MIC for each antimicrobial/microorganism combination. Each experiment was performed three times.

Measurement of Percent Sucrose Saved by Application of Approximate MIC in Tube Cultures

In order to examine whether application of biocide or oxidizer at approximate MIC values reduced sucrose losses caused by representative microbial strains, selected isolates were grown in tube cultures instead of microwell plates to facilitate measurement of sugars. These strains were grown as precultures for each set of antimicrobial susceptibility experiments as described earlier and inoculated into 10 mL of TSY medium in 50-mL conical tubes at a starting OD₆₀₀ of 0.05. The initial sucrose concentration in the experiments was 58 ± 2.2 g/L. The active ingredient ppm of each antimicrobial agent was decided from Table 2 MIC values. Tube cultures were incubated at 28°C, 250 rpm, for 18 hr. Control experiments with 0 ppm biocide were conducted at the same time, to ensure that any minor variations between precultures were taken into account. Cultures receiving 0 ppm biocide were diluted 1:10 in TSY medium before reading OD₆₀₀. In addition to the inoculated controls, uninoculated medium controls were included to verify potential loss or conversion of sugar due to sterilization, evaporation, and abiotic conversion of the sugars. Each experimental condition was conducted in triplicate.

Table 2.

Antimicrobial Susceptibility Profile of Microbial Isolates

Strain #	Microorganism	Sodium	Hydritreat 2216	Hops BetaStab	Magna Cide D	Thyme Oil
		Hypochlorite (ppm)	(ppm)	XL (ppm)	(ppm)	(ppm)
Gram-positive
LASM5	L. suionicum	250	63	16	500	500
LASM6	L. suionicum	250	125	16	63	500
LASM7	L. suionicum	250	125	16	>1,000a	250
LASM16	L. lactis	250	125	16	>1,000a	250
BSDF2-3	Leuconostoc sp.	250	63	16	>1,000a	500
NRRL B-1064	W. confusa	125	63	16	32	250
Gram-negative
LASM91	C. aureum	250	125	–	63	125
LASM77	C. bernardetii	250	125	–	63	250
LASM22	P. dispersa	250	125	–	63	250
LASM131	P. conspicua	250	125	–	32	250
LASM12	G. japonicus	250	125	–	500	250
NRRL B-806	Z. mobilis	250	63	–	16	125
Yeast
LASM56	S. cerevisiae	250	125	–	16	125

Note. Minimum inhibitory concentrations (MICs) of microbial isolates were determined from three independent experiments by microdilution assay. Parts per million (ppm) were calculated based on antimicrobial agent active ingredients.

^aMIC >1,000 ppm.

Analysis of sucrose, glucose, and fructose levels was performed by high-pressure liquid chromatography as previously described (Qi et al., 2023). Example chromatograms are provided in Supplementary Figures S1-S3. The laboratory sucrose consumption in the absence of a biocide was calculated by taking the difference between sucrose concentration at 18 hr with biocide and sucrose concentration at 18 hr without biocide.

In order to estimate the sucrose saved in a factory setting in the presence of an antimicrobial agent, a simplifying assumption was made that the observed sucrose loss in the laboratory was linear with time over the 18-hr experiment. While microbial growth curves are typically sigmoidal, our linear assumption also takes into account that these microbes are known to secrete various enzymes such as invertase, dextransucrase, or levansucrase enzymes that continually catalyze cleavage of sucrose into invert sugars glucose and fructose that are consumed as a sugar source and/or used for EPS formation. Since the antimicrobial contact and sucrose consumption times at a factory would be in the order of 10 min (0.167 hr) (Boone et al., 2017, 2016), the linear assumption was extrapolated to calculate an estimate for sucrose consumption at 10 min. Unaccounted sugar loss recorded by factories on a daily basis, attributable to microbes in part, has been reported to be on the order of 1–2 kg/tonne (Chen & Chou, 1992).

Thus, the estimated sucrose loss (in g/L) at a factory would be 0.926% of the laboratory value. This estimated factory sucrose loss (in g/L) was multiplied by the mixed juice flow rate of a typical sugarcane factory (561,000 L/hr) and divided by the cane crush rate (590 tonne/hr) (American Sugar Cane League, 2023) to provide sugar loss, in the absence of biocide, in units of kg sucrose/tonne crushed cane. Also, using the typical sugar production rate of a sugarcane factory of 64 tonne/hr of sucrose, the increase in sucrose yield (in percent) in the presence of an antimicrobial agent could also be calculated by multiplying the estimated sucrose loss (in g/L) with the mixed juice flow rate divided by the sugar production rate.

Technoeconomic Analysis

A cost analysis is presented in order to provide estimates for the break-even price of antimicrobial application, based on additional revenue realized by avoiding sucrose losses during antimicrobial application. Using material balances provided from sugarcane processing facilities in Louisiana, key metrics were determined to account for sucrose throughout the raw sugar manufacturing process. These underlying assumptions are delineated in the “Technoeconomic Analysis for Antimicrobial Agent Application” subsection. Using estimates for sucrose savings from laboratory antimicrobial analyses (as earlier) enables computation of increased revenue from additional sucrose production. This added revenue (determined based on the price of raw sugar) is set as the break-even price for antimicrobial application. Using MIC data, the final cost on a mass and/or volumetric basis for a given antimicrobial is estimated on a revenue-neutral basis.

Statistical Analysis

Mean values as well as standard deviation error bars are displayed in several figures. Statistically significant differences were calculated using the recommended Fisher–Hayter procedure in the case of equal group variances as determined by the Bartlett test, and by the Games-Howell procedure when group variances were found unequal (Klasson, 2024; Lim & Loh, 1996). Equal letters above bars indicate that means are statistically the same at a level of significance of .05.

Results and Discussion

Determination of MIC Values

In this study we tested the susceptibility of some of the most common microbes found in sugarcane factory juices and biofilms (Qi et al., 2023). While some studies report using the Kirby-Bauer disk diffusion assay with standard medium called Mueller Hinton agar, we found some of the strains in this study do not grow well on this medium. Furthermore, several of the strains produce EPS that can run across the plate making measurements of diffusion halos difficult. Additionally, the various antimicrobial agents tested in this study may vary in ability to diffuse through agar, so this method is semiquantitative at best and is not suited to determine MICs. Therefore, in order to test the range of sugar crop factory isolates including Gram-positive and Gram-negative bacteria as well as yeast, a microdilution assay with sucrose-containing medium that approximately mimics factory juices, on which all isolates readily grow, was required. Most of these isolates we obtained from Louisiana sugarcane factories, while Weissella confusa was isolated previously from sugarcane by C.S. McClesky at Louisiana State University (Collins et al., 1993; Holzapfel & Kandler, 1969). Microbial strains used in this study are listed in Table 1. Leuconostoc sp. strain BSDF2-3 was included since it was isolated from biofilm in a sugar beet factory to determine whether a biofilm-forming strain might have a different susceptibility phenotype.

The combined results of susceptibility testing by the microdilution technique are summarized in Table 2 with predicted MIC that we observed from three independent experiments when ppm concentrations were calculated based on the active ingredient concentration of each reagent tested. The MIC values of sodium hypochlorite for Leuconostoc strains isolated from sugarcane juices, LASM5, LASM6, LASM7, and LASM16 as well as the sugar beet biofilm isolate, Leuconostoc sp. BSDF2-3, were 250 ppm. Hydritreat 2216 with peracetic acid was more effective at lower concentrations and showed broad-spectrum activity with an MIC of 63–125 ppm for all Leuconostoc and Weissella strains. The MIC of Hydritreat 2216 was also 125 ppm for the Gram-negative bacteria including Chryseobacterium, Pantoea, and Gluconobacter, and the yeast Saccharomyces cerevisiae (LAM56), while the MIC for Zymomonas mobilis was 63 ppm. Magna Cide D had widely variable MIC values for all strains tested ranging from 16 ppm for S. cerevisiae and Z. mobilis to 500 ppm and above for some Leuconostoc strains. For instance, the MIC for L. suionicum LASM6 was 63 ppm, while it was much higher for the other Leuconostoc strains, at 500 ppm and above (e.g., >1,000 ppm for three isolates). Overall, Magna Cide D tended to be generally more effective at controlling Gram-negative bacteria and yeast than at Gram-positive bacteria such as Leuconostoc.

Hops BetaStab XL is known to be effective against Gram-positive, but not Gram-negative bacteria (Hein & Pollach, 1996). Indeed, preliminary tests showed no growth inhibition of any Gram-negative bacteria or S. cerevisiae as previously reported by others (Behr & Vogel, 2010; Fahle et al., 2022; Teuber & Schmalreck, 1973). Therefore, susceptibility testing for MIC values of Hops BetaStab XL was focused on Leuconostoc and Weissella strains, which indicated an MIC of 16 ppm for all Gram-positive strains tested (Table 2).

Antibacterial activity of plant-based essential oils has been reviewed previously (Tariq et al., 2019). Others have reported on various plant extracts that function as effective antimicrobial agents against bacterial species problematic for raw sugar production from sugar beets (Yousef et al., 2023). In particular, thyme oil has been reported to have broad-spectrum activity against Gram-positive and Gram-negative bacteria comparable with gentamicin controls (Borugă et al., 2014; Prabuseenivasan et al., 2006; Semeniuc et al., 2017). Additionally, plant-based essential oils may represent a sustainable antimicrobial option for producing organic sugar products that are typically sold at higher cost. Thyme oil MIC values for Leuconostoc isolates ranged from 250 to 500 ppm. Thyme oil was slightly more effective against Gram-negative bacteria with MIC values of 125–250 ppm observed for C. aureum, C. bernardetii, P. dispersa, P. conspicua, and Gluconobacter japonicus strains. Zymomonas mobilis as well as the yeast S. cerevisiae MIC was also measured at 125 ppm. MIC values for Gram-positive strains ranged from 250 to 500 ppm (Table 2). These results showing the effectiveness of thyme oil as well as other reports (Yousef et al., 2023; Yousefi et al., 2020) suggest potential for other plant-based oils as sustainable antimicrobial agents and warrant further investigation.

Assessment of Sucrose Losses Following Application of Antimicrobial Agents to Individual Microbial Strains Grown as Tube Cultures

Sucrose consumption was quantified from laboratory culture experiments in sucrose-containing medium in the presence and absence of each antimicrobial agent at an approximate MIC (Table 2) and quantified by HPLC. Example HPLC chromatograms have been included in the Supplementary Material. Antimicrobial agents and MIC values tested include Hydritreat 2216 (125 ppm), sodium hypochlorite (250 ppm), thyme oil (250 ppm), Magna Cide D (125 ppm), and Hops BetaStab XL (16 ppm). Since MIC values for Magna Cide D were determined to be over a broad range, including higher concentrations that are unlikely to be achieved in factories, 125 ppm was selected for this reagent. Each antimicrobial concentration tested was calculated based on the active ingredient ppm. Detailed sugar analysis showing consumption of sucrose and production of invert sugars glucose and fructose in laboratory experiments with single strain cultures treated with either 0 ppm or indicated antimicrobial are shown in Fig. 1. Several observations can be noted in Fig. 1. Overall, sugar “consumption” in the presence of an antimicrobial at its MIC was not noticed except in a few cases. To be fair, since the MIC of G. japonicus LASM12 to Magna Cide D was measured at 500 ppm, it did consume sugar in the presence of 125 ppm Magna Cide D and accumulated glucose (Fig. 1). While there was only minor consumption of sugars by Leuconostoc sp. BSDF2-3, there was an increased accumulation of fructose in the presence of the Magna Cide D. In the absence of a biocide, cultures with microbial isolates L. suionicum LASM7, Leuconostoc sp. BSDF2-3, and L. lactis LASM16 showed accumulation of only fructose, indicating formation of glucose-based polysaccharides (i.e., dextran). Sugar analysis in experiments with the yeast S. cerevisiae LASM56 showed both glucose and fructose accumulation, which is typical for S. cerevisiae cultures where invertase is secreted to invert sucrose into glucose and fructose. The G. japonicus LASM12 isolate produced high levels of glucose as expected for a fructan polysaccharide-forming strain (Bruni et al., 2022, 2024).

Fig. 1.

Sugar concentrations after 18 hr of growth in the presence and absence of an antimicrobial agent. Results from abiotic controls are also included. Due to the number of tube cultures run in triplicate under each condition, each panel represents one collective experiment that was conducted for each antimicrobial. The standard deviation error bars are based on total sugar concentrations. Samples LASM12 with Hydritreat 2216 and LaSM12 with Magna Cide D were run in duplicate.

The laboratory sucrose “consumption” was calculated as described in the “Measurement of Percent Sucrose Saved by Application of Approximate MIC in Tube Cultures” subection as the difference between sucrose levels in the presence of a biocide minus the sucrose levels in the absence of biocide. The results are shown in Fig. 2, and each panel shows a different microbial agent with several microbial isolates. Also included in Fig. 2 are the estimated factory sucrose losses when not adding the antimicrobial and estimated potential increases in sucrose yield if the antimicrobial is used (calculated as described in the “Measurement of Percent Sucrose Saved by Application of Approximate MIC in Tube Cultures” subection). In Fig. 2, in general, the amount of laboratory sucrose consumption is indicative of the microbial isolate’s capability to consume (or convert) sucrose in the absence of an antimicrobial; the smaller the striped bar, the less sucrose is consumed or converted. For example, with Hydritreat 2216 (at 125 ppm), the laboratory sugar consumption from not adding this antimicrobial agent over 18 hr was between 28 to 47 g/L for the strains tested. This translates to an estimated factory sucrose loss of 0.27–0.45 kg of sucrose/tonne of cane crushed (Fig. 2). If the sucrose was not lost due to the presence of Hydritreat 2216, the sucrose yield for the factory would increase by 0.22%–0.38% (Fig. 2).

Fig. 2.

Sucrose consumption in laboratory tube experiments (wide green, striped bars) without antimicrobial addition for individual strains, estimated factory sucrose loss without antimicrobial (dashed red bars), and estimated improved sucrose yield with antimicrobial (blue dotted bars). Standard deviation error bars are based on sucrose levels.

Sodium hypochlorite, at 250 ppm, showed very similar results as with Hydritreat 2216 (at 125 ppm), preventing sugar consumption. In the absence of sodium hypochlorite, the microbial strain consumed a very similar amount of sucrose, as expected. This produces very similar estimated factory sucrose loss and increased yield, as in the case of Hydritreat 2216. The plant-based thyme oil, at 250 ppm, performed well as a biocide at the concentration selected and prevented sucrose and overall sugar consumption as well as Hydritreat 2216 and sodium hypochlorite. Magna Cide D had only a modest effect on sucrose consumption. To be fair, the Magna Cide D concentration used in this experiment (125 ppm) was well below the MIC for LASM7, BSDF2-3, and LASM12, while it was above the MIC for LASM56. However, since the determined MIC values for all strains were broadly distributed over a range from 16 to >1,000 ppm, and the higher end of which is unrealistic for factories, a Magna Cide D concentration was chosen from closer to the middle of this range for the experiment. Particularly noticeable is consumption and conversion of sucrose by LASM12 (G. japonicus) in the presence of 125 ppm Magna Cide D (Fig. 1).

Because of sucrose consumption and conversion in the presence of Magna Cide D by LASM12, the benefit is less and the estimated sucrose loss and improved sucrose yield are less than with the other biocides (Fig. 2).

Note that, in the case of Hops BetaStab XL, the factory strains S. cerevisiae LASM56 and G. japonicus LASM12 were not used as this antimicrobial is ineffective against yeast and Gram-negative bacteria. Instead, L. lactis LASM16 and W. confusa NRRL B-1064 strains were used.

Technoeconomic Analysis for Antimicrobial Agent Application

Based on improved sucrose yield from the application of antimicrobial agents, the break-even cost for these agents can be estimated from the additional revenue realized by higher product yields. From the computations of estimated factory losses (i.e., the difference in raw sugar production without the indicated antimicrobial agents compared to the sugar production achieved with the indicated antimicrobial agent applied), the associated revenue from sugar production can be determined based on the price of raw sugar. This additional revenue is equal to the break-even cost for applying a given antimicrobial agent. Then, based on operational parameters for an average sugar factory in Louisiana, an estimate for the break-even antimicrobial price ($/kg or $/L) can be calculated. The average cane grinding rate is taken as 590 tonne/hr with a conservative estimation of 1.1 tonnes of mixed juice per tonne of cane (Birkett, 1977; Ram Prakash & Munsamy, 2008; Smith et al., 2015). The price of raw sugar (No. 16, US) is taken to be $0.40/lb ($0.88 per kg) (FRED, 2024). The density of mixed juice is taken to be 1.05 kg/L (Chen & Chou, 1992); this value is also used for antimicrobial agent density as a simplifying assumption. Antimicrobial concentrations used in the experimental work are: 125 ppm for Hydritreat 2216 (22% active ingredient) and Magna Cide D (proprietary); 250 ppm for sodium hypochlorite (5% active ingredient) and thyme oil (100% active ingredient); and 16 ppm for Hops BetaStab XL (9% active ingredient). The effective concentration of the antimicrobial agent is calculated by dividing the experimental concentration by the active ingredient concentration. A complete calculation example for one strain and biocide is given in Table 3 and the break-even cost for each strain and biocide combination is given in Table 4.

Table 3.

Example Calculation of an Estimate for the Break-Even Cost for Control of Leuconostoc suionicum LASM7 by the Hydritreat 2216 Antimicrobial Agent

Parameter	Value	Units
Strain	Leuconostoc suionicum LASM7	n/a
Antimicrobial agent	Hydritreat 2216	n/a
Active ingredient conc.	22	%
Estimated fact. loss (Fig. 1)	0.269	kg–sugar/tonne–cane
Cane grinding rate	590	tonne–cane/hr
Sugar price	0.88	$/kg (U.S. dollar)
Break-even cost	3,354.29	$/day
Juice and biocide density	1.05	kg/L
Mixed juice flow rate	15 576	tonne–juice/day
Antimicrobial effective conc.	568	ppm
Biocide flow rate	8 429	L–biocide/day
Break-even biocide price	0.40	$/L
Break-even biocide price	0.42	$/kg

Table 4.

Estimates for Break-Even Costs per Liter and per Kilogram for Each Strain Treated With Each Antimicrobial Agent Tested (as per the Calculation Procedure in Table 3)

Strain	Antimicrobial	Price (/L)	Price (/kg)
LASM7	Hydritreat 2216 125 ppm	$0.40	$0.42
BSDF2-3		$0.60	$0.63
LASM56		$0.52	$0.54
LASM12		$0.67	$0.70
LASM7	Sodium hypochlorite 250 ppm	$0.05	$0.05
BSDF2-3		$0.07	$0.07
LASM56		$0.06	$0.06
LASM12		$0.09	$0.09
LASM7	Thyme oil 250 ppm	$0.59	$0.62
BSDF2-3		$1.15	$1.21
LASM56		$1.32	$1.38
LASM12		$1.68	$1.76
LASM7	Magna Cide D 125 ppm	$0.34	$0.36
BSDF2-3		$0.78	$0.82
LASM56		$1.14	$1.19
LASM12		$0.45	$0.47
LASM7	Hops BetaStab XL 16 ppm	$0.62	$0.65
BSDF2-3		$1.82	$1.91
LASM16		$1.95	$2.05
NRRL B-1064		$2.47	$2.59

The calculated break-even costs range from a low value of $0.05/L based on experimental data of sodium hypochlorite with LASM7 and a high value of $2.47/L for Hops BetaStab XL with NRRL B-1064. The interpretation of break-even cost is such that if the real antimicrobial agent price is less than or equal to the break-even cost, then it may be advisable to apply the given agent at the “MIC level” tested. In this regard, a higher break-even price implies that a given antimicrobial agent is more effective. Sodium hypochlorite break-even prices are the lowest primarily because of the low active ingredient concentration and high MIC, resulting in a significantly higher effective antimicrobial concentration for sugar loss prevention. Another consideration is that although break-even price is taken as a revenue-neutral point (where excess sugar production revenue equals antimicrobial agent cost), it is still likely advisable to apply antimicrobials at break-even price because of their indirect benefits. Examples of these indirect benefits include lowered costs associated with EPS management and overall better mill sanitation, resulting in potentially lower overall operational costs.

In general, most estimated break-even prices for antimicrobials are in the range of $0.50–$2.00/L (or kg) for a given biocide and strain combination. Because the technoeconomic analysis presented here is based on conditions relevant to Louisiana and in part on experimental data across several strain and antimicrobial combinations, some level of variability in break-even cost estimates is captured in situ. In practice (and especially at larger scale, i.e., that of an operating sugar mill), there can be significant spatial and temporal variations in microbial loads and their associated sucrose consumption and EPS production rates. More precise and practical estimates of break-even costs can be determined through real industrial-scale trials, similar to those presented in the Australian study (Shi et al., 2022). The feasibility of this will be explored in ongoing/future work.

Conclusions

MIC values of industrially relevant and plant-based antimicrobial agents were experimentally determined for several sugar crop-derived microbial isolates. Some antimicrobials showed broad-spectrum activity, while some were clearly more effective at controlling Gram-positive or Gram-negative microorganisms. Since Gram-positive bacteria including Leuconostoc and Weissella are typically some of the most abundant in sugarcane juices (Qi et al., 2023), practically speaking it would make sense to use the antimicrobial agents most effective at controlling these types of microbes. Furthermore, the yeast S. cerevisiae is also a robust sucrose consumer (Bruni et al., 2022) and should be taken into consideration for an optimal antimicrobial strategy. Antimicrobial agents were further tested on selected isolates to assess effectiveness at reducing sucrose consumption in laboratory experiments. Hydritreat 2216, sodium hypochlorite, thyme oil, and BetaStab XL were effective and prevented sucrose consumption and conversion at their respective MIC values. These results show that application of antimicrobials at MIC levels is effective at controlling microbial growth and minimizing sucrose consumption. Furthermore, when laboratory results were extrapolated to a factory setting, it was estimated that as much as a 0.43% increase in sucrose yield could be achieved in the presence of an effective antimicrobial agent. These products are widely used in sugarcane or sugar beet processing in the United States, except for thyme oil (which is generally recognized as safe (GRAS)). Downstream removal is typically not required because either the antimicrobials break down naturally as a consequence of the sugar manufacturing process or residual levels are so low that it is not problematic.

Determination of MIC values for several sugar industry-relevant microbial isolates as well as resulting sucrose savings provides a framework for estimating cost basis and break-even prices for application of various antimicrobial measures.

Funding

This work was supported by the U.S. Department of Agriculture (USDA)–Agricultural Research Service (ARS), under project number 6054-41000-114-000-D. This research was also supported in part by the American Sugarcane League under project number 23-84 to E.T. and G.O.B. and the Beet Sugar Development Foundation under project number 501 to G.O.B.

Conflicts of Interest

The authors declare no conflict of interest. The authors are employed by the funding organization. However, the funders had no role in the design of the study, in the collection, analyses, or interpretation of data or in the writing of the manuscript but approved the decision to publish the results.

Data Availability

All data are incorporated into the article and its online Supplementary Material. Raw data can be made available upon reasonable request to the corresponding author.