ABSTRACT

Recent outbreaks traced to contaminated flour have created a need in the milling industry for a process that reduces pathogens in wheat while maintaining its functional properties. Vacuum steam treatment is a promising technology for treatment of low-moisture foods. Traditional thermal treatment methods can compromise wheat functionality due to high temperatures; thus, maintaining the functional quality of the wheat protein was critical for this research. The objective of this study was to evaluate the effect of vacuum steam treatment of hard red spring (HRS) wheat kernels on final flour quality and the overall efficacy of vacuum stream treatment for reducing pathogens on HRS wheat kernels. HRS wheat samples were treated with steam under vacuum at 65, 70, 75, and 85°C for 4 and 8 min. Significant changes in dough and baked product functionality were observed for treatments at ≥70°C. Treatment time had no significant effect on the qualities evaluated. After determining that vacuum steam treatment at 65°C best preserved product quality, HRS wheat was inoculated with Escherichia coli O121 and Salmonella Enteritidis PT 30 and processed at 65°C for 0, 2, 4, 6, or 8 min. The treatments achieved a maximum average reduction of 3.57 ± 0.33 log CFU/g for E. coli O121 and 3.21 ± 0.27 log CFU/g for Salmonella. Vacuum steam treatment could be an effective pathogen inactivation method for the flour milling industry.

HIGHLIGHTS
• Vacuum steam may be an effective treatment for decontamination of wheat before milling.

• Temperature is critical to maintaining wheat end-use quality.

• Volumes of baked bread loaves decreased significantly at processing temperatures ≥75°C.

• E. coli and Salmonella reductions of 3.6 and 3.2 log CFU/g, respectively, occurred at 65°C.

Historically, little concern has been expressed for the safety of wheat flour and its related food products because flour is low in moisture. Microbes were considered unable to survive in such an environment; however, current research indicates that some microbes can survive under low-moisture conditions for extended periods (3, 6, 18). Wheat flour products such as cookie dough, cake batter, or frozen dough are intended to be consumed only after a cooking step, yet consumers may eat raw dough despite warnings on food labels to avoid such practices. Recent outbreaks related to consumption of raw flour, primarily caused by Salmonella and Shiga toxin–producing Escherichia coli, have raised concerns about the safety of flour products (11).

Wheat kernels are exposed to many potential contaminants in the environment from animals, insects, soil, and wind and from poorly sanitized harvesting equipment, transportation vehicles, and storage containers. Current wheat milling practices do not utilize techniques that actively aim to reduce microbial populations (2, 18). The development of a process that can reduce foodborne pathogens while maintaining wheat quality is of great interest to wheat millers and processors. Traditional thermal treatment methods utilize high heat to inactivate pathogens and are undesirable in the flour industry because high temperatures can reduce the functionality of the grain. The gluten protein, imperative for the formation of dough and integral to the quality of bread products, is particularly susceptible to denaturation by high-temperature treatments.

One processing method for potential use in the milling industry is steam treatment applied under a vacuum. Steam is the desired medium for pathogenic reduction in the food industry because of the high level of heat transfer. Vacuum steam treatment utilizes steam under subatmospheric pressure, which creates lower temperatures. The low temperatures attainable with this method do not negatively affect the quality of the wheat but maintain the effective heat transfer attained with steam, allowing for a greater reduction of pathogens. In previous research, vacuum steam treatment was effective for reducing pathogens in low-moisture foods such as flaxseed, sunflower seeds, quinoa, and peppercorns (16). The process could be used on wheat kernels before milling to reduce pathogens commonly found on the outer bran layer of the kernel. The objectives of this study were to (i) establish vacuum steam processing parameters that preserve flour functionality and (ii) determine pathogen inactivation levels under vacuum steam processing conditions that maintain flour functionality.

## MATERIALS AND METHODS

### Wheat material

A blend of hard red spring (HRS) wheat varieties grown in 2017 were used for this experiment: 29.5% Linkert, 29.5% Glenn, 15.2% SY Soren, 9.8% Elgin-ND, 9.5% ND VitPro, and 6.5% SY Ingmar. The mixture was homogenized with a homogenizer (FPB-005, American Process Systems, Gurnee, IL) and cleaned for processing and milling on a dockage tester (Carter Day International, Minneapolis, MN) with a number 8 riddle.

### Vacuum steam processing for wheat quality analysis

Vacuum steam processing of samples used for quality testing and microbial analysis was performed separately but with the same operating procedures. The laboratory-scale vacuum steam treatment system used in this study has been described previously (14) and was capable of processing 1 kg of wheat kernels at a time. A preheating treatment was used before vacuum steam processing, as is common in industrial settings, which preheated the kernels to 40 ± 4°C with forced dry air in an oven (GCA/Precision Scientific, American Sterilizer Co., Erie, PA). A preliminary experiment (17) was performed to confirm that the preheating process did not compromise wheat quality because the present experiment was designed to study specifically the effects of the vacuum steam treatment.

The 1-kg wheat sample was held in a metal basket (25 cm diameter) with a bed depth of approximately 2.5 cm. The basket was placed on a metal plate and covered with a glass bell to create a sealed vessel. Ports in the bottom of the metal plate were used for steam addition, vacuum application, and pressure and temperature measurements. Thermocouples (T-37X-T, ThermoWorks, American Fork, UT) were buried in the center of the bed of wheat to monitor and record processing temperatures. Thermocouple data loggers (HOBO, Onset, Bourne, MA) were used, and each thermocouple recorded processing temperatures at 15-s intervals. For each of the desired processing temperatures, a separate vacuum pressure was required to maintain steam conditions within the chamber. Temperatures of 65, 75, and 85°C required a vacuum to be maintained at 230 to 275, 356 to 421, and 536 to 627 mbar, respectively. The vacuum steam treatment unit was operated with maximum steam addition (rate dependent on the vacuum level) until the target temperature was obtained, after which the temperature was held within ±3°C for either 4 or 8 min under vacuum. After processing, kernel moisture increased up to 16%, which was measured using a moisture tester (GAC 2100, Dickey-John, Minneapolis, MN). Samples with moisture >14% were dried at room temperature overnight or until their moisture was <14% and placed in cold storage before milling.

### Wheat tempering and milling

Samples were tempered to 16% moisture 18 to 24 h before milling. Twenty minutes before milling, samples were tempered to 16.5% moisture. Samples were milled randomly over three consecutive days on a mill (MLU-202, Bühler Industries, Uzwil, Switzerland) according to American Association of Cereal Chemists (AACC)–approved method 26-21.02 (4). Milling conditions were maintained at 68% relative humidity and 22 ± 1°C. Milling extraction (as a percentage) was calculated by dividing the weight of straight grade flour produced from the weight of the initial grain sample.

### Flour analyses

Moisture, ash, and protein contents were analyzed with AACC methods 44-15.02, 08-01.0, and 46-30.01, respectively (4). Flour color was analyzed with the CIE 1976 L*a*b* scale with a colorimeter (CR-410 Chroma Meter, Konica Minolta, Ramsey, NJ) with a granular-materials attachment (CR-A50, Konica Minolta).

Apparent enzymatic activities of α-amylase, xylanase, and polyphenol oxidase (PPO) were quantified with kits (Megazyme International Ireland, Bray, Ireland). PPO was analyzed in whole kernel samples, and procedures were followed as detailed in AACC method 22-85.01 (1, 4). Absorbance was read at 475 nm in cuvettes with a spectrophotometer (DR/4000 U, Hach, Loveland, CO). Whole wheat flour samples were ground from kernels in a sample mill (Cyclone, UDY Corp., Fort Collins, CO) for xylanase and α-amylase activity testing. Xylanase activity was tested with a kit (T-XAX, Megazyme) with modifications as described previously (5). AACC method 22-05.01 was used for α-amylase measurement with amylazyme tablets from the kit (4).

To study the functional characteristics of the protein, the wet gluten (Glutomatic GM 2200 gluten washer, Perten Instruments, Springfield, IL) and gluten index (Gluten Index centrifuge 2015, Perten Instruments) were analyzed according to AACC method 38-12.02 (4). Starch damage and total starch procedures followed AACC methods 76-31.01 and 76-13.01, respectively, utilizing kits (Megazyme) (4). Farinograph analysis was performed according to AACC method 54.21.02 (4) with a 50-g bowl attachment. Tests were conducted on the Farinograph-E (Brabender, Duisburg, Germany) with software version 4.0.3. Farinograph analysis is used to understand the quality of the dough produced from a flour sample and the gluten functionality.

Bread was baked following a straight dough method with a 2-h fermentation, according to AACC method 10-10.03. Samples were baked over two consecutive days. Bread was evaluated for appearance, color, and crumb following AACC method 10-12.01 (4). Additional analyses for bread texture were conducted with a texture analyzer (TA-XT2i, Texture Technologies Corp., Scarsdale, NY) with a cylindrical acrylic probe (2.5 cm in diameter by 35 mm high).

### Bacterial strains used in this study

Two pathogens were studied in this experiment: E. coli O121:H19 (TW07932) from a sporadic case of human illness (STEC Center, Michigan State University, East Lansing) and Salmonella enterica Enteritidis PT 30 (ATCC BAA-1045). These pathogens are commonly used in evaluation of thermal treatments for low-moisture foods. Bacterial stocks were stored at −80°C in brain heart infusion (BHI) broth (Criterion, Hardy Diagnostics, Santa Maria, CA) with 15% glycerol.

### Inoculation of wheat

The inoculation protocol previously described by Shah et al. (16) was used with minor modifications. Bacterial freezer stocks were streaked on BHI agar plates for isolation and grown overnight at 37°C. For each strain, a colony was transferred to 5 mL of BHI broth and incubated at 37°C for 20 h. The overnight broth culture (250 μL) was plated uniformly onto BHI agar plates (100 by 15 mm) with a sterile spreader (Fisher Scientific, Waltham, MA), and plates were incubated at 37°C for 24 h. To achieve 8 log CFU/g, the bacterial lawns from 14 plates were collected with a sterile spreader and mixed into a sterile beaker containing 2.5 mL of sterile water. The bacterial suspension was poured onto 600 g of unconditioned HRS wheat kernels in Whirl-Pak bags (Nasco, Fort Atkinson, WI). The bag contents were mixed by hand for 3 to 5 min to obtain a homogenous distribution of bacteria.

### Assessing the homogeneity of Salmonella and E. coli inoculated onto wheat kernels

To assess the homogeneity of bacteria inoculated onto unconditioned wheat kernels, eight 25-g samples were randomly plated in duplicate at the time of inoculation (0 h) and 24, 48, and 72 h postinoculation. To plate samples, inoculated wheat was weighed in a Whirl-Pak bag, and Butterfield dilution buffer was added in appropriate amounts. Samples were homogenized (Masticator, IUL Instruments, Barcelona, Spain) for 90 s, and serial dilutions were plated in duplicate onto tryptic soy agar with 0.1% ferric ammonium citrate (J. T. Baker, Phillipsburg, NJ) and 0.06% sodium thiosulfate (VWR, Radnor, PA) for Salmonella and onto selective agar plates (HiCrome ECC, Sigma-Aldrich, St. Louis, MO) for E. coli O121. The plates were incubated at 37°C for 24 to 48 h, and black colonies indicative of Salmonella and purple colonies indicative of E. coli were enumerated (Q-Count reader, Advanced Instruments, Norwood, MA). Four additional randomly selected 25-g samples were processed for enumeration to confirm homogeneity for each experiment at the time of inoculation. A standard deviation of <0.5 log CFU/g was deemed an acceptable range to indicate homogeneity of the inoculum for each inoculated batch of wheat.

### Water activity equilibration and storage of inoculated wheat

The water activity (aw) of the wheat kernels before and after inoculation was measured with an aw meter (4TE, Aqualab, Pullman, WA). The average aw values of the wheat kernels before and immediately after inoculation were 0.53 ± 0.03 and 0.69 ± 0.05, respectively. To adjust the aw to the original concentration, lithium chloride (anhydrous, 99%, −20 mesh; Alfa Aesar, Ward Hill, MA) was used. The inoculated wheat was transferred to a sterile stainless steel tray (30.5 by 23 cm), which was placed in a closed cooler chamber (24 by 16 in. [61 by 41 cm]; Coleman Co., Kingfisher, OK). Lithium chloride (15 to 30 g) was weighed into plastic trays (Fisher Scientific) and saturated with water, and these trays were placed adjacent to the stainless steel trays in the closed chamber. The aw of the inoculated wheat was 0.59 ± 0.02 after 24 h and 0.5 ± 0.02 within 48 h.

### Vacuum steam processing for microbial reduction analysis

The vacuum steam procedure for microbial inactivation was largely similar to the procedure for analysis of wheat quality. However, microbial reduction experiments were performed at only 65°C for 0, 2, 4, 6, and 8 min. To simulate the 1-kg sample size used for wheat quality analysis, 925 g of uninoculated HRS wheat was added to a metal autoclave basket lined with window screen and preheated to 40 ± 3°C, which took approximately 35 min. After preheating, the basket was removed from the oven, and three cotton bags (5 by 8 cm; Uline, Pleasant Prairie, WI) each containing 25 g of inoculated wheat were buried within the bed of grain. For samples with a treatment time of 0 min, the vacuum steam treatment unit was allowed to heat until thermocouples in the bed of wheat read 65°C, after which the sample was immediately removed from the vessel. Each pathogen was inoculated and processed separately.

After processing, the bags containing the inoculated wheat were immediately removed from the processing vessel, and the wheat was transferred into sterile Whirl-Pak bags (33 by 18 cm). Samples were diluted, homogenized, and plated as described above. Inoculated wheat that was not exposed to thermal treatment was plated as a control to determine initial pathogen levels. Plates were incubated at 37°C for 24 h, and colonies were enumerated (Q-Count reader).

### Statistical analysis

For the wheat quality experiment, all vacuum steam processing conditions were repeated in triplicate. Four control samples were analyzed, and mean values are presented. Mean separation and least significant difference tests were used to determine significant differences (P < 0.05) between treatments with a one-way analysis of variance (ANOVA) in SAS for Windows, version 9.4 (TS level IM4; SAS Institute, Cary, NC). A two-way ANOVA was also used to analyze the interaction of temperature and time in treatments. Correlation analysis was performed using the CORR procedure in SAS.

To better study nonlinear trends in the data, the logistic power (Sigmoid model) was calculated in Curve Expert Professional (version 2.6.3, https://www.curveexpert.net/products/curveexpert-professional/):
$$\def\upalpha{\unicode[Times]{x3B1}}$$$$\def\upbeta{\unicode[Times]{x3B2}}$$$$\def\upgamma{\unicode[Times]{x3B3}}$$$$\def\updelta{\unicode[Times]{x3B4}}$$$$\def\upvarepsilon{\unicode[Times]{x3B5}}$$$$\def\upzeta{\unicode[Times]{x3B6}}$$$$\def\upeta{\unicode[Times]{x3B7}}$$$$\def\uptheta{\unicode[Times]{x3B8}}$$$$\def\upiota{\unicode[Times]{x3B9}}$$$$\def\upkappa{\unicode[Times]{x3BA}}$$$$\def\uplambda{\unicode[Times]{x3BB}}$$$$\def\upmu{\unicode[Times]{x3BC}}$$$$\def\upnu{\unicode[Times]{x3BD}}$$$$\def\upxi{\unicode[Times]{x3BE}}$$$$\def\upomicron{\unicode[Times]{x3BF}}$$$$\def\uppi{\unicode[Times]{x3C0}}$$$$\def\uprho{\unicode[Times]{x3C1}}$$$$\def\upsigma{\unicode[Times]{x3C3}}$$$$\def\uptau{\unicode[Times]{x3C4}}$$$$\def\upupsilon{\unicode[Times]{x3C5}}$$$$\def\upphi{\unicode[Times]{x3C6}}$$$$\def\upchi{\unicode[Times]{x3C7}}$$$$\def\uppsy{\unicode[Times]{x3C8}}$$$$\def\upomega{\unicode[Times]{x3C9}}$$$$\def\bialpha{\boldsymbol{\alpha}}$$$$\def\bibeta{\boldsymbol{\beta}}$$$$\def\bigamma{\boldsymbol{\gamma}}$$$$\def\bidelta{\boldsymbol{\delta}}$$$$\def\bivarepsilon{\boldsymbol{\varepsilon}}$$$$\def\bizeta{\boldsymbol{\zeta}}$$$$\def\bieta{\boldsymbol{\eta}}$$$$\def\bitheta{\boldsymbol{\theta}}$$$$\def\biiota{\\boldsymbol{\iota}}$$$$\def\bikappa{\boldsymbol{\kappa}}$$$$\def\bilambda{\boldsymbol{\lambda}}$$$$\def\\bimu{\boldsymbol{\mu}}$$$$\def\binu{\boldsymbol{\nu}}$$$$\def\bixi{\boldsymbol{\xi}}$$$$\def\biomicron{\boldsymbol{\micron}}$$$$\def\bipi{\boldsymbol{\pi}}$$$$\def\birho{\boldsymbol{\rho}}$$$$\def\bisigma{\boldsymbol{\sigma}}$$$$\def\bitau{\boldsymbol{\\tau}}$$$$\def\biupsilon{\boldsymbol{\upsilon}}$$$$\def\biphi{\boldsymbol{\phi}}$$$$\def\bichi{\boldsymbol{\chi}}$$$$\def\bipsy{\boldsymbol{\psy}}$$$$\def\biomega{\boldsymbol{\omega}}$$$$\def\bupalpha{\bf{\alpha}}$$$$\def\bupbeta{\bf{\beta}}$$$$\def\bupgamma{\bf{\gamma}}$$$$\def\bupdelta{\bf{\delta}}$$$$\def\bupvarepsilon{\bf{\varepsilon}}$$$$\def\bupzeta{\bf{\zeta}}$$$$\def\bupeta{\bf{\eta}}$$$$\def\buptheta{\bf{\theta}}$$$$\def\bupiota{\bf{\iota}}$$$$\def\bupkappa{\bf{\kappa}}$$$$\def\\buplambda{\bf{\lambda}}$$$$\def\bupmu{\bf{\mu}}$$$$\def\bupnu{\bf{\nu}}$$$$\def\bupxi{\bf{\xi}}$$$$\def\bupomicron{\bf{\micron}}$$$$\def\buppi{\bf{\pi}}$$$$\def\buprho{\bf{\rho}}$$$$\def\bupsigma{\bf{\sigma}}$$$$\def\buptau{\bf{\tau}}$$$$\def\bupupsilon{\bf{\upsilon}}$$$$\def\bupphi{\bf{\phi}}$$$$\def\bupchi{\bf{\chi}}$$$$\def\buppsy{\bf{\psy}}$$$$\def\bupomega{\bf{\omega}}$$$$\def\bGamma{\bf{\Gamma}}$$$$\def\bDelta{\bf{\Delta}}$$$$\def\bTheta{\bf{\Theta}}$$$$\def\bLambda{\bf{\Lambda}}$$$$\def\bXi{\bf{\Xi}}$$$$\def\bPi{\bf{\Pi}}$$$$\def\bSigma{\bf{\Sigma}}$$$$\def\bPhi{\bf{\Phi}}$$$$\def\bPsi{\bf{\Psi}}$$$$\def\bOmega{\bf{\Omega}}$$$$y = {a \over {\left[ {1 + {{\left( {{x \over b}} \right)}^c}} \right]}}$$

For the microbial reduction experiment, three biological and three technical replicates were processed and plated in duplicate. Duplicate plate counts for each technical replicate were log transformed and averaged in Excel (Microsoft, Redmond, WA). One replicate of each treatment was completed on separate days. Each pathogen was analyzed individually. Linear regression was analyzed with Sigma Plot 12 (Systat Software, Chicago, IL). The D65°C for each pathogen was calculated from the slope of the regression.

## RESULTS

### Flour analyses

Milling yield did not change between treatments, averaging 68.7%. The lowest average extraction rate was 68.7%, and the highest was 69.3%. The moisture of the flour was not significantly different between samples (Table 1). Ash was not significantly different from the control for any treatments, except for samples treated at 65°C for 8 min, which were 0.1% higher (Table 1).

TABLE 1

Proximate composition of refined flour after milling kernels treated with vacuum steam (VS)

Protein in the flour was significantly reduced in samples that received vacuum steam treatment at temperatures ≥75°C. Despite the significance in the reduction in protein, the overall decrease of approximately 0.2% protein is unlikely to affect the functionality of the flour.

The total starch analysis revealed no significant differences between treatments at 74.3 to 78.2% (Table 1). Analysis of starch damage revealed no significant differences (P < 0.05) from the control for all treatments at 7.7 to 8.2% (Table 1).

Flour color was analyzed with the CIE L*a*b* scale (Table 2). No significant difference was observed between treatments for L* or a* values, indicating no differences in lightness or greenness of the sample. Treated flour b* values were significantly lower than those of the control, indicating decreased yellow color of the flour; however, no observable differences in yellowness were found after visual inspection of the flours.

TABLE 2

Color of refined flour obtained from kernels treated with vacuum steam (VS) on the CIE L*a*b* scalea

Apparent PPO, α-amylase, and xylanase activity was measured because heating during vacuum steam processing could have either promoted enzymatic activity or denatured the enzymes. PPO activity was low in the control sample and was not significantly different among treatments (Table 3). High α-amylase activity is generally undesirable because too much hydrolyzed starch can result in a sticky dough and poor baking characteristics, such as low bread loaf volume. Apparent α-amylase activity decreased significantly from the control for all treatments, but no significant difference in apparent α-amylase activity was found among temperature treatments (65, 75, and 85°C) or treatment times at each temperature. Xylanase activity, typically undesirable because of its contribution to the syruping phenomenon in refrigerated doughs, decreased significantly as treatment temperature increased, with reductions of approximately 25, 50, and 96% from the control for treatments at 65, 75, and 85°C, respectively (Table 3).

TABLE 3

Apparent enzyme activity in whole wheat flour prepared from wheat kernels treated with vacuum steam (VS)a

### Dough functionality

Wet gluten represents the quantity of gluten in the sample, and the gluten index indicates the quality of the gluten in the sample. No significant difference in wet gluten or the gluten index was found between the 65°C treatment and the control sample; however, treatments at 75 and 85°C resulted in decreases in both parameters, and at 85°C no gluten was obtained from the flour. These results suggest that denaturation of the gluten protein occurs above 65°C, with complete denaturation at 85°C. Regression analysis confirmed the rapid reduction in gluten quantity and quality observed at 75°C (Fig. 1).

FIGURE 1

Negative sigmoidal relationship between both (A) wet gluten and (B) the gluten index with increasing temperature of vacuum steam treatment. Data fitted using the logistic power model.

FIGURE 1

Negative sigmoidal relationship between both (A) wet gluten and (B) the gluten index with increasing temperature of vacuum steam treatment. Data fitted using the logistic power model.

In the farinograph analysis, absorption is the amount of water the flour needs to form optimum dough consistency; often, higher water absorption is desired for industrial uses. Analysis of samples treated with vacuum steam revealed an overall trend of decreasing absorption of 1 to 2% from the control at treatment temperatures up to 75°C (Table 4). However, at 85°C, the trend reversed, and water absorption of the flour increased to values similar to those for the control, though overall changes were minor.

TABLE 4

Dough quality of flour obtained from kernels treated with vacuum steam (VS) as measured with the farinographa

Stability time (time during which the dough remains at the desired consistency) is another important parameter for industrial uses because the dough should remain tolerant of overmixing. Typically, longer dough stability is desired and indicates a stronger gluten network. The control flour sample had a stability time of 28.2 min, indicating a strong gluten network (Table 4). At processing temperatures of 75 and 85°C, the stability time decreased an average of >20 min. The stability of the samples treated at 85°C was 1.2 min (Table 4).

The peak time is the time it takes the dough sample to reach its peak viscosity before the dough begins to break down and soften. The average peak time of the control was 9.2 min and decreased for the 65°C treatments to 6.5 and 5.5 min after 4 and 8 min, respectively. Vacuum steam treatments of 75 and 85°C resulted in a significant reduction in peak time compared with both the control and 65°C treatments.

The mixing tolerance index (MTI) is used to inform how much the dough will soften during mixing. Low MTI values are desirable for industrial uses, indicating less softening. The MTI of the treated flours increased significantly as treatment temperature increased above 65°C (Table 4). The control flour had an MTI of 8.0 Brabender units. At 65°C, the MTI was not significantly different; however, at 75°C, a rapid increase in MTI was observed. At 85°C, the MTI increased significantly again. Results from all parameters of farinograph analysis suggest a significant decrease in dough quality at temperatures of ≥75°C, which further supported results obtained with the wet gluten analysis.

### Baking analysis

End-use quality of the flour was determined by bread baking analysis. Oven rise (Table 5) is the increase in height of the loaves of bread during baking. The control sample averaged 3.90 cm rise in the oven, similar to the oven rise observed for 65°C treatments (Table 5). Vacuum steam treatment at 75°C resulted in a significantly reduced oven rise of approximately 1 cm, and 85°C samples collapsed slightly by 0.30 and 0.10 cm in the oven. These results suggest that gluten proteins were altered at vacuum steam temperatures of ≥75°C, thus compromising the ability of the dough to hold the expanding gas.

TABLE 5

Baking quality of bread samples prepared with flour obtained from kernels treated with vacuum steam (VS)a

Loaf volume is the size of the loaf after baking. The control sample had a loaf volume of 1,011 cm3. Loaf volume decreased significantly as the vacuum steam treatment temperature increased (Table 5). Treatment at 65°C for 4 min yielded a loaf volume similar to that of the control, but treatment for 8 min resulted in a significant reduction in loaf volume to 880 cm3. Samples treated at 75 and 85°C had just over one-half and one-third of the loaf volume of the control. At the same temperatures at which a decrease in loaf volume was observed, the specific volume increased because the bread became more compact and denser.

Texture analysis of the slices of bread after baking revealed a significant increase in firmness from flours treated at 85°C for both 4 and 8 min (Table 5). This increased firmness results from the dense crumb structure, as indicated by the increasing specific volume. Although statistical analysis did not reveal significant differences between the control and the 65, 70, and 75°C treatments, sensory analysis of the loaves of bread could have revealed differences.

### Microbial reduction

Analysis of all quality parameters indicated that wheat grain processed at 65°C could be made into flour of acceptable quality, and that the treatment time of 4 or 8 min did not significantly affect this quality. Thus, microbial reduction studies were completed at 65°C for 0, 2, 4, 6, and 8 min. The pathogens E. coli O121 and Salmonella were used in this study because they have been associated with multiple outbreaks and recalls involving flour products in recent years (11). Figure 2 shows the relationship between increasing processing time and microbial reduction in the wheat sample. The time for the wheat sample to reach the target temperature of 65 ± 3°C (come-up time) was 12 min for Salmonella testing and 10 min for E. coli O121 testing, which is comparable to the average come-up time observed during quality analysis (9.7 min), indicating that samples underwent similar processing conditions.

FIGURE 2

Inactivation of Salmonella (A) and E. coli O121 (B) on wheat kernels treated with vacuum steam. Boxes represent the log N/N0 of each pathogen, and circles with whiskers represent the temperature profile. Data for each of the three replicates are plotted, with the regression line indicated by a solid line. Data for the microbial populations were averaged for the three technical replicates.

FIGURE 2

Inactivation of Salmonella (A) and E. coli O121 (B) on wheat kernels treated with vacuum steam. Boxes represent the log N/N0 of each pathogen, and circles with whiskers represent the temperature profile. Data for each of the three replicates are plotted, with the regression line indicated by a solid line. Data for the microbial populations were averaged for the three technical replicates.

For E. coli O121, the untreated inoculated grain had an average pathogen level of 7.98 ± 0.07 CFU/g. During the come-up time when the wheat increased in temperature from 40 to 65°C, the E. coli O121 level was reduced significantly by 2.46 ± 0.34 CFU/g. E. coli O121 levels continued to decrease with increased processing time. The D65°C was 6.7 min, as determined by the E. coli O121 levels once the target temperature of 65°C was reached (Fig. 2A). The highest overall average reduction in E. coli O121 on the wheat kernels was 3.57 ± 0.33 log CFU/g and was achieved after treatment at 65°C for 8 min.

Wheat kernels inoculated with Salmonella had an initial level of 8.96 ± 0.28 log CFU/g. During the come-up time, a reduction of 2.56 ± 0.39 log CFU/g was observed. The D65°C was 14.8 min, as determined from the Salmonella levels once the target temperature of 65°C was reached. After processing for the longest period (8 min), an average reduction of 3.21 ± 0.27 log CFU/g was achieved. Overall, Salmonella was more resistant than E. coli O121 to the heat treatments utilized in this study.

## DISCUSSION

### Wheat quality

In our experiments, vacuum steam treatment of wheat kernels did not result in changes to the milling yield, findings similar to those of a 1978 study in which evaluation of the effect of heating on wheat quality revealed no significant difference in milling yield (7). Throughout our farinograph, wet gluten, and baking analyses, a significant reduction in gluten quality was observed at processing temperatures ≥75°C. Similar findings of gluten denaturation during heating were found by Schofield et al. (15). Their experiments after heat treating isolated wet gluten samples indicated that significant changes to the glutenin fraction of the gluten occurred at 55 to 75°C. The authors speculates that increased temperatures promoted the unfolding of the three-dimensional structure of the proteins, after which disulfide-sulfhydryl interchange reactions locked unfolded proteins in their denatured state (15). The importance of glutenin for optimal dough quality has been studied extensively (21, 22). The heat sensitivity of the glutenin protein at <75°C, as demonstrated by Schofield et al. (15), coupled with the importance of glutenin in dough quality present a unique challenge for heat treatment methods.

Multiple tests analyzing the quality of wheat samples treated with vacuum steam revealed very few significant differences between treatment times of 4 and 8 min at the same temperature. Schofield et al. (15) obtained a similar result and hypothesized that the loss of gluten functionality occurred within 2 or 3 min, after which prolonged heating at the same temperature did not cause further functionality loss. This finding could be of importance when considering microbial reduction because prolonged heat treatment may not cause additional damage to the quality of the grain but may allow further reduction of microbial populations. Processors of wheat products could adopt a low temperature–long time process for the reduction of pathogens while maintaining gluten functionality.

An important distinction between the present study and many other studies of the effect of heating on gluten is the application of heat before versus after gluten formation. In the present study, heat was applied to whole wheat kernels when gliadin and glutenin proteins had not interacted to form a gluten complex. In contrast, the effect of heat on wet gluten or isolated gluten portions of the flour in other studies (15, 20, 23) occurred after gliadin and glutenin interaction had formed the gluten matrix, which may affect the arrangement of bonds or the mechanism of disassociation of the gluten network.

### Microbial analysis

Comparison of the present microbial reduction study to studies of other wheat products is difficult because of differences in microbial populations, inoculation procedures, and the types of heat treatments, including the use of dry heat versus steam, pressure, and treatment times and temperatures.

Similar to the present vacuum steam research, Hu et al. (12) studied microbial reduction on wheat kernels before milling. These researchers applied saturated steam at multiple temperatures (110, 140, 170, and 200°C) for 0 to 80 s to both tempered and nontempered grain and measured the reductions in total bacteria, Bacillus spp., and mold on uninoculated grain. Temperatures of the wheat grain were not reported, although testing was performed on 200 g of wheat spread in a thin layer for each replicate, with steam application at a higher velocity, so temperatures probably increased quickly. Maximum reduction in total wheat bacterial counts (3.37 log CFU/g) were observed after treatment with saturated steam at 200°C for 80 s on tempered grain. Most of the microbial reduction occurred within the first 40 s of processing. A major difference in the vacuum steam treatment in the present study is the longer come-up time to reach the target temperature in comparison with the application of saturated steam at higher temperatures. We observed a significant reduction in pathogen levels during the come-up time (10 to 12 min), and the use of shorter times will be a focus of future research.

The majority of published research relating to the microbial safety of wheat flour has been focused on inoculation followed by thermal treatment of the wheat flour rather than on treatment of the wheat kernels before milling. For direct thermal treatment of flour, different heating times have been reported to achieve similar levels of pathogen reduction, as observed in the present study. Heating small volumes of inoculated wheat flour in 0.5-mL tubes in a digital dry bath at 65°C for 60 min achieved an average 3.32-log reduction of enterohemorrhagic E. coli (EHEC) (8). Wheat flour inoculated with E. coli O157:H7 heated in 1-g portions in plastic bags placed in a water bath at 65°C for 5 min resulted in an ∼3.5-log reduction (10). Although in both of these studies similar reductions of EHEC were found, the amount of flour that was heat treated was small (0.3 to 1 g), and the methods of heating were different, likely leading to the different treatment times needed to achieve this level of reduction. The D65°C for EHEC strains in flour reported by Forghani et al. (8) were 8.01 to 12.20 min, similar to our findings. When larger volumes of wheat flour have been heat treated (e.g., 3 kg heated to 75°C with radio waves), a D75°C of 17.65 ± 1.58 min was reported for Salmonella PT 30 (13).

In recent studies, differences in thermal tolerance have been reported between Salmonella and EHEC, in particular E. coli O121. Similar to our findings, D-values for Salmonella PT 30 were greater than those for E. coli O121 in wheat flour (19). Forghani et al. (9) also reported lower δ values for E. coli O121 and other EHEC serotypes than for Salmonella PT 30 in wheat flour. Although these researchers evaluated the efficacy of heating for decreasing microbial populations, the ways in which these treatments affect the functionality and quality of the flour itself are unknown.

Vacuum steam processing can maintain the quality and functional characteristics of flour produced from treated HRS wheat kernels while achieving a significant reduction in microbial populations. With laboratory-scale vacuum steam treatment equipment, HRS wheat of acceptable quality was obtained at vacuum steam temperatures up to 65°C; however, higher temperatures inhibited the functionality of the dough because of apparent denaturation of the gluten protein. The most useful indicators of flour quality were the wet gluten, farinograph, and baking analyses, which suggested that gluten denaturation begins at 75°C, with nearly complete denaturation of the gluten protein during vacuum steam treatment at 85°C. Vacuum steam treatment was effective for reducing the apparent enzyme activity of PPO and xylanase, which would be desirable for end-use applications of the flour. Treatment times of 4 and 8 min at the same temperature had the same effect on the quality of the wheat, indicating that temperature was the important parameter during processing.

Implementation of the vacuum steam technology in the grain industry will likely require that treatments maintain wheat functionality; thus, microbial reduction studies were performed at only 65°C at which wheat quality was maintained. After 8 min at 65°C, log reductions of 3.57 ± 0.33 for E. coli O121 and 3.21 for Salmonella Enteritidis PT 30 were observed. Vacuum steam treatment is an effective pathogen inactivation method that could be used in the flour milling industry.

## ACKNOWLEDGMENTS

This work was supported in part by grant USDA-NIFA-2016-69003-24851 (to T. M. Bergholz), funding from the State Board of Agricultural Research and Education (SBARE-Wheat), and the North Dakota Wheat Commission.

## REFERENCES

REFERENCES
1.
Anderson,
J. V.,
and
Morris
C. F.
2001
.
An improved whole-seed assay for screening wheat germplasm for polyphenol oxidase activity
.
Crop Sci
.
41
:
1697
1705
.
2.
Berghofer,
L. K.,
Hocking
A. D.,
Miskelly
D.,
and
Jansson
E.
2003
.
Microbiology of wheat and flour milling in Australia
.
Int. J. Food Microbiol
.
85
:
137
149
.
3.
Beuchat,
L. R.,
Komitopoulou
E.,
Beckers
H.,
Betts
R. P.,
Bourdichon
F.,
Fanning
S.,
Joosten
H. M.,
and
Kuile
B. H. T.
2013
.
Low–water activity foods: increased concern as vehicles of foodborne pathogens
.
J. Food Prot
.
76
:
150
172
.
4.
Cereals & Grains Association
.
2009
.
AACC approved methods of analysis, 11th ed
.
Available at: http://methods.aaccnet.org/. Accessed 3 October 2019.
5.
Courtin,
C. M.,
Gys
W.,
Gebruers
K.,
and
Delcour
J. A.
2005
.
Evidence for the involvement of arabinoxylan and xylanases in refrigerated dough syruping
.
J. Agric. Food Chem
.
53
:
7623
7629
.
6.
Crowe,
S. J.,
Bottichio
L.,
L. N.,
Whitney
B. M.,
Corral
N.,
Melius
B.,
Arends
K. D.,
Donovan
D.,
Stone
J.,
and
Allen
K.
2017
.
Shiga toxin–producing E. coli infections associated with flour
.
N. Engl. J. Med
.
377
:
2036
2043
.
7.
Dermott,
T.,
and
Evans
D.
1978
.
An evaluation of fluidized-bed heating as a means of disinfesting wheat
.
J. Stored Prod. Res
.
14
:
1
12
.
8.
Forghani,
F.,
den Bakker
M.,
Futral
A. N.,
and
Diez-Gonzalez
F.
2018
.
Long-term survival and thermal death kinetics of enterohemorrhagic Escherichia coli serogroups O26, O103, O111 and O157 in wheat flour
.
Appl. Environ. Microbiol
.
84
:
e00283
-
18
.
9.
Forghani,
F.,
den Bakker
M.,
Liao
J.-Y.,
Payton
A. S.,
Futral
A. N.,
and
Diez-Gonzalez
F.
2019
.
Salmonella and enterohemorrhagic Escherichia coli serogroups O45, O121, O145 in wheat flour: effects of long-term storage and thermal treatments
.
Front. Microbiol
.
10
:
1
12
.
10.
Greene,
S. E.
2012
.
Thermal inactivation of Escherichia coli O157:H7 and Salmonella Agona in wheat flour
.
M.S. thesis.
Virginia Polytechnic Institute and State University
,
Blacksburg
.
11.
Harris,
J. L.,
and
S.
2018
.
Flour and cereal grain products: foodborne illness outbreaks and product recalls
.
12.
Hu,
Y.,
Nie
W.,
Hu
X.,
and
Li
Z.
2016
.
Microbial decontamination of wheat grain with superheated steam
.
Food Control
62
:
264
269
.
13.
Liu,
S.,
Ozturk
S.,
Xu
J.,
Kong
F.,
Gray
P.,
Zhu
M.-J.,
Sablani
S. S.,
and
Tang
J.
2018
.
Microbial validation of radio frequency pasteurization of wheat flour by inoculated pack studies
.
J. Food Eng
.
217
:
68
74
.
14.
McEvoy,
J. M.,
Doherty
A. M.,
Sheridan
J. J.,
Blair
I. S.,
and
McDowell
D. A.
2001
.
Use of steam condensing at subatmospheric pressures to reduce Escherichia coli O157:H7 numbers on bovine hide
.
J. Food Prot
.
64
:
1655
1660
.
15.
Schofield,
J.,
Bottomley
R.,
Timms
M.,
and
Booth
M.
1983
.
The effect of heat on wheat gluten and the involvement of sulphydryl-disulphide interchange reactions
.
J. Cereal Sci
.
1
:
241
253
.
16.
Shah,
M. K.,
Asa
G.,
Sherwood
J.,
Graber
K.,
and
Bergholz
T. M.
2017
.
Efficacy of vacuum steam pasteurization for inactivation of Salmonella PT 30, Escherichia coli O157:H7 and Enterococcus faecium on low moisture foods
.
Int. J. Food Microbiol
.
244
:
111
118
.
17.
Snelling,
J.
2019
.
Vacuum steam pasteurization of hard red spring wheat
.
M.S. thesis.
North Dakota State University
,
Fargo
.
18.
Sperber,
W. H.
2007
.
Role of microbiological guidelines in the production and commercial use of milled cereal grains: a practical approach for the 21st century
.
J. Food Prot
.
70
:
1041
1053
.
19.
Suehr,
Q. J.,
Anderson
N. M.,
and
Keller
S. E.
2019
.
Desiccation and thermal resistance of Escherichia coli O121 in wheat flour
.
J. Food Prot
.
82
:
1308
1313
.
20.
Weegels,
P.,
De Groot
A.,
Verhoek
J.,
and
Hamer
R.
1994
.
Effects on gluten of heating at different moisture contents. II. Changes in physico-chemical properties and secondary structure
.
J. Cereal Sci
.
19
:
39
47
.
21.
Weegels,
P.,
Hamer
R.,
and
Schofield
J.
1997
.
Depolymerisation and re-polymerisation of wheat glutenin during dough processing. II. Changes in composition
.
J. Cereal Sci
.
25
:
155
163
.
22.
Weegels,
P.,
Van de Pijpekamp
A.,
Graveland
A.,
Hamer
R.,
and
Schofield
J.
1996
.
Depolymerisation and re-polymerisation of wheat glutenin during dough processing. I. Relationships between glutenin macropolymer content and quality parameters
.
J. Cereal Sci
.
23
:
103
111
.
23.
Weegels,
P.,
Verhoek
J.,
De Groot
A.,
and
Hamer
R.
1994
.
Effects on gluten of heating at different moisture contents. I. Changes in functional properties
.
J. Cereal Sci
.
19
:
31
38
.
This is an open access article