Introduction

Biological indicator system (BIS) are used globally to monitor the efficacy of sterilization processes. BIS is a microbiological test system providing a defined resistance to a specified sterilization process. The system consists of a known number of microorganisms, with a determinate resistance to the mode of sterilization, in or on a carrier and enclosed in a protective package plus a recovery medium that promotes the growth of any surviving test organism cells (Food and Drug Administration (FDA) 2007). BIS allows the qualitative evaluation, by visual identification, of a color change or turbidity of the substrate media caused by the growth of the inactivated bacteria. They are used primarily in the health care industry but also in many other commercial and industrial applications. It is often necessary to monitor the effectiveness of the processes used to sterilize equipment, such as medical and nonmedical devices, instruments, foods, medical or other microbial contaminated wastes, and other articles and materials (Gillis et al. 2010). The periodic utilization of bioindicators to monitor the sterilization processes used in the medical and industrial sectors has been a major requirement of regulatory agencies.

Conventional BIS has used dormant spores of some gram-positive bacteria such as Bacillus sp. (ISO 11138-4 2006). Spores of Bacillus atrophaeus are used as a biological indicator (BI) to assess antiseptic and sterilizing products and procedures, such as low-temperature steam, dry heat, plasma, UV radiation, electrothermal deactivation, and microwaves (Halfmann et al. 2007; Oliveira et al. 2010) and in the development of dry-heat sterilization cycles for spacecraft (Kempf et al. 2008). These spores also have biotechnological applications in water and wastewater treatments and systems (Szabo et al. 2007), as a surrogate for Bacillus anthracis in biodefense (Weber et al. 2003) and as a potential adjuvant for veterinary rabies vaccine (Oliveira-Nascimento et al. 2012).

It is known that all of the BI system components, such as spore characteristics, the carrier material on which the spores are deposited and the primary packaging and recovery media, have an impact on the BI performance (Gillis et al. 2010). The sporulation conditions significantly affect the spore properties, including germination capacity and resistance to heat treatment (Nguyen Thi Minh et al. 2008) and different composition of the recovery medium may not have the same degree of ability to promote the growth of injured spores, both factors interfering with the efficacy of the BI system.

Glycerol or 1,2,3-propanetriol, a clear, odorless, viscous liquid, is a by-product from the biodiesel industry. Biodiesel, a renewable fuel, is produced from vegetable oils or animal fats through transesterification, generating about 10 % (w/w) glycerol. World biodiesel production is increasing exponentially and was at 19 billion liters in 2010, a 12 % increase from 2009 (Shrank and Farahmand 2011). The excess of glycerol generated has a consequence in that its price is low and the substance may become an environmental problem since it cannot be disposed of in the environment (Da Silva et al. 2009). It is a growing global concern to reduce the future environmental problems caused by the accumulation of glycerol and to make biodiesel production more profitable. So the development of new value-added applications for glycerol is of worldwide interest. Glycerol is a good source of carbon and energy for the growth of several microorganisms, and it may be suitable for the biotechnological production of bioindicators.

Solid-state fermentation (SSF) using sand as support was first described by Sella et al. (2012) as a process that provided a high-yield of B. atrophaeus spores. Sand was considered an inert support that enables the direct use of fermented material in BI, making the final product less expensive.

The current work proposes the development of a new and economical bioprocess for the production of a B. atrophaeus sterilization BI system (spores on a carrier plus a recovery medium) by using a glycerol as the sole carbon source.

Materials and methods

This study was executed in three steps: (a) development and optimization of the recovery medium, (b) development and optimization of spore production by SSF, and (c) performance evaluation of the developed BIS by comparing sporulation and germination kinetics as well as resistance characteristics and process costs, with a commercial BI (Fig. 1).

Fig. 1
figure 1

Illustrative chart for the design and performance evaluation of glycerol-based bioindicator (CCD central composite design, PB Plackett–Burman design, BIS biological indicator system)

Full size image

Bacterial strain, spores suspension, and glycerol source

B. atrophaeus ATCC 9372 (batch: 1403349) was obtained from standard strains supplied by Instituto Nacional de Controle de Qualidade em Saúde (INCQS/MS, Brazil). Spores suspensions were utilized as seed and prepared with a standard agar sporulation method. The inoculum was prepared from a suspension of vegetative cells grown for 18 h at 36 °C in tryptic soy broth (TSB), and the cultures were grown in Roux flasks containing 400.0 mL of sporulation medium—8.0 g/L yeast extract, 4.0 g/L nutrient broth, 0.05 g/L MnSO4·4H2O, 0.05 g/L CaCl2·6H2O, and 30.0 g/L agar. Sporulation was carried out at 36 °C for 7 days. Spores were harvested and suspended in cold and sterile 0.02 M calcium acetate solution adjusted to pH 9.7 with 1.4 g/L calcium hydroxide solution. The suspension was filtered through cotton and gauze tissue and centrifuged three times at 1,048×g for 20 min at 4 °C. The spores’ suspensions were subjected to a heat treatment (80 °C, 10 min), which was lethal to vegetative cells but not to spores, and were stored at 4 °C.

Three different brands of glycerol were used: (a) glycerol P.A. (Nuclear, Diadema, Brazil), ≥99.5 % purity; (b) glycerol for electrophoresis (Sigma-Aldrich, St. Louis, MO), ≥99.0 % purity; and (c) glycerol commercial grade (Rioquimica, São José do Rio Preto, Brazil), ≥95 % purity.

Development and optimization of the BI recovery medium

Selection of the nitrogen source

Five different media with 10.0 g/L glycerol, 0.02 g/L bromothymol blue, and different organic and inorganic nitrogen sources were formulated: (1) 5.6 g/L tryptone, (2) 4.8 g/L casein peptone, (3) 4.0 g/L ammonium sulfate, (4) 2.0 g/L ammonium sulfate, and 2.0 g/L ammonium nitrate, and (5) 4.0 g/L ammonium nitrate. The pH was adjusted with NaOH 0.1 N to 7.0, and the media were autoclaved at 121 °C for 15 min. These media were loaded in 3.0-mL aliquot into borosilicate glass vials, stoppered with silicon rubber, and sealed with an aluminum seal. These studied recovery media were submitted to a viability assay and dry-heat resistance tests with commercial BI spore on strips (CPPI, Piraquara, Brazil).

Experimental designs for optimization

  1. (a)

    A 22 full-factorial central composite design (CCD) with four axial points (α = 1.682) and two replications at the center points was employed to search for the optimal nitrogen source and glycerol concentrations. The amounts of the main components were as follows: 5.0–15.0 g/L glycerol and 2.0–10.0 g/L tryptone. The values of bromothymol blue (0.002 g %) was fixed, and the pH was adjusted before sterilization to 7.0 (design and results are shown in Table 1).

    Table 1 A 22 full-factorial CCD with uncoded values for glycerol and tryptone and experimental results for the first optimization of B. atrophaeus biological indicator recovery medium
    Full size table
  2. (b)

    A second 22 full-factorial CCD design was employed to evaluate the soluble starch and CaCl2 influence and their optimum concentrations. The amounts of the evaluated components were as follows: 0.5–2.0 g/L soluble starch and 0.1–0.3 g/L CaCl2. The values of glycerol, tryptone, and bromothymol blue were fixed at the optimum, and the pH was adjusted before sterilization to 7.0.

Development and optimization of spore production

Fermentation process

SSF was carried out in 250-mL Erlenmeyer flasks containing 50.0 g of support each. Washed commercial grade sand, washed once in tap water, twice in distilled water, dried in trays for 24 h at 90 °C in an air oven and sieved to obtain a ∼1.0-mm particle size (mesh, 14–20) was used as inert support. The support was impregnated with 13.0 mL of the tested medium, according to the sand’s water-holding capacity. The pH was adjusted after sterilization with 0.1 M sterile NaOH to 7.0. Flasks were autoclaved at 121 °C for 15 min. The inoculum size was 4 % (v/v substrate). The inoculum was produced in 30 mL of (TSB) medium, inoculated with 100 μL of 109 CFU/mL spores suspension, and incubated for 18 h at 36 °C. Sporulation was carried out at 36 °C for 7 days. The fermented material was dried in trays for 24 h at 65 °C in an air oven. Spore detachment for counting was performed as previously described (Sella et al. 2012). SSF yield was available as colony-forming units per gram of dry matter and spores resistance was determined by the D 160 °C value (decimal reduction time).

Experimental designs for optimization

  1. (a)

    A Plackett–Burman design (Plackett and Burman 1946) was used to evaluate the effects of medium components on the spores yield and heat resistance. The variable factors—glycerol, tryptone, KH2PO4, MnSO4, CaCl2, and MgSO4—were examined at two coefficient levels, low (−) and high (+), using a set of orthogonal contrasts. Central points were investigated in triplicate to identify any experimental anomalies (design and results are demonstrated in Table 2).

    Table 2 Plackett–Burman design matrix with uncoded values and experimental results to investigate the effects of the sporulation medium composition on Bacillus atrophaeus spore yield and heat resistance
    Full size table
  2. (b)

    A 23 full-factorial CCD was used to optimize the three main components of the sporulation media: glycerol, tryptone, and calcium chloride. Independent factors were studied at five different levels with two replications at the center points. The amounts of the main components were as follows: 10.0–70.0 g/L glycerol, 5.0–35.0 g/L tryptone, and 0.01–0.20 g/L CaCl2.

BIS performance determination

The effectiveness of the developed formulation was checked by evaluating three different BIS batches using different brands of the raw material.

Recovery medium

Performance was determined by spores growth viability and dry-heat resistance (D 160 °C value) after inoculation with B. atrophaeus commercial spore strips (CPPI, Piraquara, Brazil).

Fermented medium with spores

Biological indicators were prepared by diluting the dried fermented product with sterilized dry sand to obtain ∼106 CFU spores/g and filling 7.0 mL borosilicate glass vials with 1.0 g of material. The loaded vials were closed with silicon rubber stoppers and sealed with aluminum seals. After the dry-heat resistance test, the vials were cultured with 3.0 mL of commercial recovery medium (CPPI, Piraquara, Brazil). Germination and growth of the spores were observed over 24 up to 48 h incubation time.

Biological indicator system

Dry-heat and ethylene oxide resistance was tested for developed BIS, and a commercial BI system (CPPI, Piraquara, Brazil) was used as a control.

Viability

For the viability assay, germination and growth, the recovery medium was inoculated with spores (10 units) that was incubated at 36 °C and observed over 24 to 48 h incubation time. Positive results were indicated by changes in the color (green to yellow) and turbidity of the recovery medium.

Dry-heat resistance test

The D-value is defined as the time taken, in a specified set of conditions, for the spore population to decrease by 90 % or 1 log (Food and Drug Administration (FDA) 2007). The D value was determined by fraction-negative analysis, using the limited Spearman–Kaber method (United States Pharmacopeia XXXI (USP) 2008). Dry-heat conditions were as follows: 160 °C for 8, 12, 16, 20, 24, 28, 32, 36, and 40 min in a tabletop air oven. Temperature was monitored using a K-type thermocouple 1.5 × 400 mm coupled to a digital thermometer SALCASTERM mod1200-2 (SALCAS, São Paulo, Brazil). A survivor curve was constructed for BIS resistance determination based on the spores’ direct enumeration. The survival/kill times were determined by the following formulae: survival time (log of population − 2) × D value and kill time = (log of certified population + 4) × D value.

Ethylene oxide resistance test

SSF glycerol-based dry fermented were diluted with sterile inert silica sand quartz to achieve 106 CFU/g and 1.0 g was conditioned into a gas-permeable package and sealed. Ethylene oxide (EtO) sterilization was performed using 650 mg/L at 55 °C, 55 % humidity, with a partial cycle time of 15 min. D EtO values (decimal reduction time) were calculated based on logarithmic curve inactivation, after determining the number of surviving spores. A commercial EtO BIS (BROWNE, Leicester, UK) was used as control and had the following labeled characteristics: D EtO = 3.4 min and initial number of spores N 0 = 1.8 × 106 CFU/unit.

Sporulation kinetics

SSF was carried out under the optimum conditions. Standard agar sporulation was done as previously described in “Bacterial strain, spores suspension and glycerol source.” Growth was monitored by total cell and spore counts (total cells − spores = vegetative cells). Total cells and spore counts were done by serial decimal dilutions in distilled sterile water and 50 μL of each dilution was inoculated on a tryptone soy agar plate surface in duplicate. Plates were incubated overnight at 36 °C. For spore counts the sample was submitted to heat treatment (80 − 85 °C, 10 min) to eliminate vegetative cells. Time course results of spores formation was compared with sporulation on agar. Glycerol consumption profiles on SSF were analyzed by high performance liquid chromatography in a Varian HPLC using a Shodex KS 801 column, with refraction index detector, mobile phase purified water, 0.5 mL/min, 70 °C. The samples were centrifuged at 10,000 × g and filtered with 0.22 μm PVDF membranes, diameter 13 mm (Millipore, Billerica, MA). Glycerol was identified by comparing retention times with a standard. For pH determination, 1.0 g sample was diluted with 9.0 mL distilled water, blended for 2 min, and then centrifuged. The supernatant liquid was used for the assay.

BI recovery kinetics

Unheated BI (spores in vials), and those that were subjected to sublethal heating (160 °C for 15 min),were inoculated into developed recovery medium and cultured for 24 h at 36 °C. Measurements of colony forming units were used to monitor the time course of growth in recovery medium. The 160 °C dry-heat exposure time of 15 min was calculated based on the optimized medium D value obtained for unheated spores as the time to kill ∼50 % of the initial population (106 spores). Glycerol consumption profiles were observed by high performance liquid chromatography as described.

General conditions and statistical analysis

The assays were carried out in triplicate. Microbiological assays were conducted under aseptic conditions in a Good Manufacturing Practices certified laboratory ISO 5 clean room. Experimental designs and analyses were carried out using Statistic 8.0 software (StatSoft, Tulsa, OK) and the SGWIN program (Stat Graphic Plus for Windows version 5.0, Statistical Graphics, Cheshire, CT). Analysis of variance was used to evaluate the model at 95.0 % of significance level. The quality of the polynomial model equation was judged statistically using the coefficient of determination (R 2), and its statistical significance was determined by an F test. The validation of the optimization of medium composition was carried out by the analysis of three tests using the optimized conditions and the results analyzed statistically. Commercial B. atrophaeus Biological Indicators (IB Calor Seco, CPPI, Piraquara, Brazil) were used as controls in all assays and for comparative results. The commercial BI samples had the following labeled characteristics: D 160 °C = 6.5 ± 0.2 min and initial number of spores N 0 = 3.6 ± 0.8 106 CFU/unit.

Results

Development and optimization of the BIS recovery medium

For glycerol-based medium development the results of nitrogen sources selection demonstrated that only organic sources have a positive effect over heat injured spores recovery after 48 h incubation time. Besides, all media had promoted the germination and outgrow of unheated spores. Glycerol media with ammonium sulfate and ammonium nitrate promoted later heat damaged spores growth (only after 7 days of incubation time). The spores D 160 °C values observed with the addition of organic nitrogen sources were 5.8 ± 0.1 min for glycerol plus tryptone medium, 5.4 ± 0.2 min for glycerol plus casein peptone medium and 6.5 ± 0.1 min for the control medium (Online resource 1 in the Electronic supplementary material (ESM)).

The optimum glycerol concentration in the recovery medium (the main carbon source) and nitrogen source concentration (tryptone) were determined. The results are shown in Table 1. The value of R 2 was 91.99 % suggesting the strong significance of the model. The best result obtained was D 160 °C = 6.9 ± 0.1 min for the concentrations of 12.0 g/L of glycerol and 10.0 g/L of tryptone. The relationship between the experimental variables and the heat resistance of spores was plotted on a three-dimensional surface graph (Fig. 2a) and fitted to a second-order polynomial equation describing the surface:

Fig. 2
figure 2

Estimated response surface plots showing the effects of tryptone and glycerol (a) and CaCl2 and soluble starch (b) on the recovery of B. atrophaeus spores, as determined by D 160 °C

Full size image
$$ \begin{array}{*{20}{c}} {{{D}_{{160^\circ {\text{C}}}}}{\text{value}} = - 8.63 + 1.35\,{\text{glycerol}} + 1.50\,{\text{tryptone}}} \\ { - 0.06\,{\text{glycero}}{{{\text{l}}}^{2}}} \\ { + 0.001\,{\text{glycerol}}\,{\text{tryptone}}} \\ { - 0.07\,{\text{trypton}}{{{\text{e}}}^{2}}} \\ \end{array} $$
(1)

Where the optimum calculated D 160 °C value was 6.9 min.

The starch and CaCl2 optimization results showed a significant effect of starch (p value = 0.0001) and no significant effects of CaCl2 presence in the developed medium (p value = 0.9678) (Online resource 2 in the ESM). The contour response surface plot illustrate the behavior of the D 160 °C value (Fig. 2b). Application of the response surface methodology-RSM gave the following regression equation, which is an empirical relationship between the D 160 °C-value and the tested variables:

$$ {D_{{{16}0{ }^\circ {\text{C}}}}} = {5}.{83} + {1}.0{2}\,{\text{starch}}-0.{29}\,{\text{starc}}{{\text{h}}^{{2}}} $$
(2)

Where the optimum calculated D 160 °C value was 6.7 min for 2.0 g/L soluble starch without CaCl2 addition.

In order to check the model adequacy, three experiments were performed with the optimized medium: 12.0 g/L glycerol, 10 g/L tryptone, 2.0 g/L soluble starch, and 0.025 g/L bromothymol blue, and their results were compared with those of the reference medium. The optimized medium attained a D 160 °C value of 6.6 ± 0.1 min, which is relatively close to the theoretically predicted value and the industrialized control recovery medium (6.7 ± 0.1 min).

Development and optimization of spores’ production

The Plackett–Burman assays were conducted to study influence of the composition of the sporulation medium. The results demonstrated spore yield variation from 102 up to 106 CFU/g and the dry-heat resistance (D 160 °C value) variation within 3.0 min (Table 2). The ANOVA analysis demonstrated the statistical significance of the components of the sporulation medium on spore yield and dry-heat resistance by observing the mean square and estimating the experimental error. For spores yield, only potassium and calcium showed significant influence, negative (p value = 0.0001 and effect = −7.53) and positive (p value = 0.0335 and effect = 2.64), respectively. In the case of heat resistance, four factors presented p values of <0.05 (potassium, calcium, glycerol, and tryptone) indicating that they produce significant effects on the spores dry-heat resistance. Potassium and glycerol demonstrated a negative influence (Online resources 3 and 4 in the ESM). In this way, the first one was removed from the medium and the glycerol concentration was reduced in the next experiments.

For sporulation medium evaluation, it is important to consider quantitative aspects such as spores yield and qualitative aspects such as heat-resistance. A CCD was employed to determine the optimal levels of glycerol, tryptone, and CaCl2 and the effect of their interactions (Online resource 5 in the ESM). The ANOVA analysis of results indicated that among the independent variables, CaCl2 had no significant effects (p value = 0.0714 for spore yields and 0.7937 for heat resistance) and negative coefficients, suggesting that its absence increased spore production and resistance. In this model, R 2 was 91.83 and 95.79 % for spores yield and heat resistance, respectively. The results were plotted (Fig. 3) and fitted with the following second-order polynomial:

Fig. 3
figure 3

Estimated response surface plots showing the effects of tryptone and glycerol on spores yield (a) and dry-heat resistance (b) of B. atrophaeus spores, as determined by CFU/g dry matter and D 160 °C, respectively

Full size image
$$ {\text{Spores}}\left( {{\text{CFU}}/{\text{g}}\,{\text{dry}}\,{\text{matter}}} \right) = {3}.{3} \times {1}{0^{{7}}} + {6}.{7} \times {1}{0^{{5}}}{\text{glycerol}}-{1}.{6} \times {1}{0^{{4}}}{\text{tryptone}}-{1}.{9} \times {1}{0^{{4}}}{\text{glycero}}{{\text{l}}^{{2}}} + {5}.{3} \times {1}{0^{{3}}}{\text{glycerol}}\,{\text{tryptone}}-{7}.{7} \times {1}{0^{{{3} }}}{\text{trypton}}{{\text{e}}^{{2}}} $$
(3)

and,

$$ {D_{{{16}0^\circ {\text{C}}}}} = {4}.{7} + 0.0{3}\,{\text{glycerol}}-0.0{17}\,{\text{tryptone}}-0.000{87}\,{\text{glycero}}{{\text{l}}^{{2}}} + 0.000{14}\,{\text{glycerol}}\,{\text{tryptone}} + 0.000{12}\,{\text{trypton}}{{\text{e}}^{{2}}} $$
(4)

The models predicted that the optimal values for the test factors of variables were glycerol = 25.0 g/L and tryptone = 12.0 g/L. The maximum predicted value of spore yield was 3.8 × 107 CFU/dry matter and for heat-resistance D 160 °C = 4.9 min.

Finally, three tests were carried out under the optimized conditions to validate the model. The experimentally derived D 160 °C value was 4.8 ± 0.2 min and N 0 = 2.3 ± 1.2 × 108 CFU/g dry matter. The results of the validation of the regression equations obtained with the response surface methodology-RSM demonstrated no statistically significant variation among the three batches produced with different brands of glycerol (p value = 0.7075).

Performance evaluation of the developed BI system’

There are international references regulations recommending BI requirements (ISO 11138 2006; Food and Drug Administration (FDA) 2007; United States Pharmacopeia XXXI (USP) 2008). Besides the recommended parameters (viable spore population assay, ∼106 CFU/unit and resistance characteristics: D 160 °C value, ≥3.0 min and survival time, ≥12.0 min and D EtO value, ≥3.0 min and survival time, ≥15.0 min), the sporulation and germination kinetics were determined. The survivor curve was also plotted to elucidate the process characteristics and to evaluate the effect of the sterilization process on the developed BIS.

Sporulation kinetics

A time course of spore production using the optimized medium composition was performed and compared with the sporulation on standard agar medium (Fig. 4). It was observed that maximal spore yield 2.9 × 108 CFU/g dry matter was achieved after 120 h of SSF; 2.5 × 1010 CFU/g dry matter of spores were reached at 144 h for standard agar medium. A pH increase during all incubation time was observed only for agar sporulation. The HPLC analysis indicated that the strain is able to consume it completely in the first 72 h (Fig. 4b).

Fig. 4
figure 4

Sporulation kinetics of B. atrophaeus: a solid-state fermentation on optimized glycerol-based sporulation medium (squares), on standard agar medium (triangles), and pH measurement (dashed line), (b) optimized solid-state fermentation process showed spores (squares) and germinative cells (circles) and glycerol (triangles) and proteins (diamonds) concentration

Full size image

Germination kinetics

The kinetics of spore germination and cellular growth is presented in Fig. 5. In all conditions, there was a rapid cell growth from 0 to 3 h after the start of incubation and then a slower growth until 15–18 h. Visual detection of color change and turbidity caused by microbial growth and/or cumulative metabolic activity in the glycerol-based recovery medium was observed after 18 h for nondamaged spores (4.1 × 107 CFU/mL) and 24 h for injured spores inoculum (3.7 × 107 CFU/mL). For the control medium, the corresponding times were 12 and 15 h, respectively.

Fig. 5
figure 5

Growth, glycerol consumption, and pH kinetics of B. atrophaeus spores in the biological indicator (BI) recovery medium before (a, c) and after (b, d) heat damage at 160 °C for 15 min. Bacteria were on optimized glycerol + tryptone BI recovery medium (squares) or commercial BI recovery medium as a reference (triangles). In (c) and (d), glycerol is indicated by dotted lines. The arrows indicate visual detection of color change and turbidity due to microbial growth

Full size image

BIS sterilization resistance

Three lots of newly developed BI were produced and tested, and all three gave satisfactory results: N 0 = 1.2 ± 0.6 × 106 CFU/unit and viability = 100 % growth in a 24-h incubation time. The determination of the D 160 °C value was performed by the limited Spearman–Kaber (LSK) and survival curve methods (Online resource 6 in the ESM), and the obtained results were D 160 °C = 3.1 ± 0.1 min, survival time = 12.2 min, and kill time = 32.3 min and D 160 °C = 4.6 ± 0.1 min, survival time = 18.4 min, and kill time = 47.4 min, respectively. Both results are in accord with the recommended parameters (D 160 °C value, ≥3.0 min). Control commercial BIS results were D 160 °C = 6.5 ± 0.2 min, survival time = 29.9 min, and kill time = 68.9 min for LSK method and D 160 °C = 6.1 ± 0.1 min, survival time = 26.2 min, and kill time = 62.8 min for survival curve method.

The oxide ethylene resistance was demonstrated by D EtO = 5.6 ± 0.6 min, survival time = 21.2 min, and kill time = 64.5 min results. Control commercial BIS results were D EtO = 3.1 ± 0.1 min, survival time = 12.8 min, and kill time = 32.8 min.

Cost analysis

Preliminary comparative production costs between conventional BIS and the developed glycerol BIS are demonstrated in Table 3. The minimum costs reduction was 41.8 % for the quality control tests and the maximum was for the feedstock (72.8 %). The process cycle time was reduced by 44.8 % with corresponding reduction in direct labor and use of equipment.

Table 3 Comparative preliminary cost of production to obtain 50,000 Bacillus atrophaeus sterilization Biological Indicator System (BIS) units (corresponding to one batch) between the developed glycerol-based BIS and the conventional BIS
Full size table

Discussion

Recovery conditions have been shown to influence the heat resistance of Bacillus spores. It has been recognized that the recovery of damaged bacterial spores depends on the medium in which the spores are cultured (Lòpez et al. 1997). Commercial recovery growth media are typically composed of water, protein extracts or digests, carbohydrates, minerals, and a pH indicator to create optimal conditions for injured spores’ outgrowth, so the alternative glycerol-based recovery medium composition should provide the same conditions, as demonstrated by organic nitrogen source addition to the medium. Since the killing of Bacillus spores by dry heat is accompanied by the accumulation of DNA damage (Setlow 2007), a mild thermal treatment may affect some enzymes synthesis. Elmerich and Aubert (1971) reported that Bacillus, which lacks glutamate dehydrogenase, glutamate synthase, and either glutamine synthetase, were unable to consume ammonium as nitrogen source. DNA repair systems involved in the repair of dry-heat damage to spore’s DNA may explain the later assimilation of the inorganic nitrogen sources. Calcium, because of its capacity to induce acid-soluble spore proteins formation, helps in regenerating the DNA of damaged cells (Setlow 2007). Starch, as an absorbent of free acids and some inhibitors, is reported as a favorable component supply for a recovery medium (Penna et al. 2000). The improvement of the germination and outgrow capacity of the glycerol-based recovery medium by the soluble starch addition is in agreement with Brown and Gaze (1988), López et al. (1997), and Penna et al. (2000). The nonsignificant effects of calcium that were observed in the present study, contradicting Sasaki et al. (2000), may have been caused by the additional calcium, which is present in the tested spore strips. Dry-heat resistance results obtained from the developed BI recovery medium are in agreement with the official standards for dry-heat sterilization BI, which recommend a D 160 °C value of ≥3.0 min (Food and Drug Administration (FDA) 2007; ISO-11138 2006; United States Pharmacopeia XXXI (USP) 2008). This fact demonstrates the glycerol-based recovery medium could replace the commercial one.

An optimum sporulation medium has to provide the maximization of spores yield associated with thermal resistance improvement. Under nutrient sources limitation and high cells density B. atrophaeus can sporulate by forming a highly resistant endospore, which is later released by lyses of the mother cell (Piggot and Losick 2002). The sporulation conditions such as heat resistance and germination capacity determine the characteristics of the formed spores (Rose et al. 2007). In addition, it was cited that divalent ions induce sporulation and increase the thermal resistance: calcium as key element for the development of resistance by inducing specific genes involved in the synthesis of the spore coat and potassium and manganese by stimulating the uptake of potassium (Cazemier et al. 2001; Hornstra et al. 2009). The negative effect of potassium, manganese, and magnesium on spores yield and resistance are in disaccord with Shi and Zhu (2006), which produced Bacillus subtilis in a distillery effluent and Cazemier et al. (2001) for spores’ resistance, which were produced in nutrient agar supplemented with ions. However, the observed calcium influence is in agreement with Hornstra et al. (2009) who produced spores of B. subtilis 168 cells in minimal defined medium with an increasing amount of Ca2+. It was demonstrated that ion traces, which are present in tryptone are enough to provide B. atrophaeus heat resistant spores’ formation. The results from the obtained model for spores’ production are in agreement with the predicted value, since a regulatory standard (United States Pharmacopeia XXXI (USP) 2008) allows a 50 up to 300 % variation on spore count, and it influences the D value calculation. No similar studies were published with which the obtained results could be compared. The optimized process of spore production resulted in high heat-resistant spores’ level that can be used for B. atrophaeus spores production for other applications, e.g., water and wastewater treatments.

Sporulation kinetics’ demonstrated that SSF promoted a drastic pH decrease in the first 24 h corresponding to the exponential growth period, due to organic acids released during initial glycerol metabolization (similar to the carbohydrate metabolization pathway), followed by an increase until 120 h, corresponding to the maximum spore yield. This increase in pH is justified by the reassimilation of the acid products and increase in protein turnover (caused by the extensive degradation and resynthesis of the proteins from the mother cell) (de Vries et al. 2005; Mandelstam and Waites 1968). A similar sporulation profile was described by Monteiro et al. (2005) for B. subtilis cultivated in a commercial sporulation medium. The glycerol consumption kinetics confirmed the B. atrophaeus suitability to employ the glycerol as the sole carbon source. The high number of vegetative cells after 72 up to120 h incubation time demonstrates that not all cells produced spores; but there is the possibility of spore germination. According to Veening et al. (2006), Bacillus sp. uses a bet-hedging strategy, whereby in the same culture only some cells enter into the sporulation process, whereas others use alternative metabolites to continue growth, and it leads to two distinguishable cell types: sporulating and nonsporulating cells. The number of vegetative cells declined after 144 h, but this was not accompanied by an increment in the number of spores, indicating cell death, which may explain the decrease in pH during the same period. Sporulating cells are able to kill the cells that have not yet initiated this process, and this phenomenon is usually responsible for the reduction in cell density of a sporulating culture (Gonzalez-Pastor et al. 2003). The continuous decrease of protein concentration may be caused by its consumption, although it was not possible to differentiate the presence of proteins in the medium derived from produced enzymes and/or cell lyses. Sporulation kinetics results also allowed the sporulation process to be restricted in 48 h.

When evaluating a BI system for detection of its efficacy it is critical to understand that stressed microorganisms (post sublethal sterilization exposition) do not have the same responses to growth as the nonheat-damaged microorganisms. It is not unusual for heat-damaged microorganisms to exhibit an inhibition of growth or a retarded growth (Wesceie et al. 2009). The spores’ germination kinetics demonstrated that recovery medium allowed the conditions for injured spores germination and outgrowth. The visual color change detects acid metabolites produced during growth of the B. atrophaeus. This growth produces a pH change in the medium, which causes a change of the color of the medium from green to yellow. It requires an average of cell density greater than 106 cells/mL (Gillis et al. 2010). The results indicated that the medium had heat-damaged spores’ recovery capacity within a 24-h incubation time.

The nonglycerol consummation in germination and outgrowth was not expected. According to Burke and Tempest (1990), the glycerol kinase activity may be decreased when the extracellular glycerol concentration was close to a cell-saturating level, suggesting that glycerol kinase synthesis might be repressed by the excess of glycerol from the recovery medium plus the glycerol from fermentation. However, the germination might occur in response to environmental signals, including amino acids and cell wall peptidoglycan muropeptides (degradation products of the peptidoglycan cell wall) as described by Shah et al. (2008). The germination of a single spore might lead to the germination of most of the population (Setlow 2008). In this case, the starch present in the recovery medium may have been used as the initial carbon source since B. atrophaeus is an amylase producer. The microorganism multiplication might be supported by the spore-derived proteins released during the early germination, when 20 % of the spore proteins are degraded, providing the source of amino acids for the biosynthesis of new proteins and nucleotides (Setlow and Primus 1975). Since no glycerol was consumed by B. atrophaeus after spore germination, glycerol might have another particular mechanism to improve the dry-heat resistance, and more work should be performed in order to further elucidate the positive variation in D value with the presence of glycerol.

The newly developed BIS performance results meets international standards and regulations. The difference observed in results of D 160 °C value (1.5 min) was not an artifact of the method as suggested by the United States Pharmacopeia XXXI (USP) (2008). It is clear that the LSK method evaluated the BI system (spores plus recovery medium), while the survival curve method only evaluated the spores performance that were similar with that obtained after sporulation optimization, which was tested with a commercial recovery medium. This indicates that the recovery medium could be optimized for a specific use (with spores in the dried glycerol-based fermented). The obtained ethylene oxide resistance results are also in accord with the recommended parameters. The developed BIS showed a greater resistance to ethylene oxide than commercial BI; it may be reported as a positive factor to improve EtO process monitoring. Since the regular cycle time is 180 min, a BI with a kill time of approximately 60 min better demonstrates the sterilization security. Experimental evidence and scientific modeling may allow the modification of EtO sterilization process time or parameters. In this case, the developed BI, suitability for use in this specific sterilization process, should be established through new studies.

The preliminary cost analysis showed the process efficacy. The use of highly available raw material associated to the efficient use of energy and water and no emission of wastes, through SSF utilization reduces cost and makes the process environmentally friendly.

The results of this study show that glycerol could be used to produce a high-quality, low-cost BIS that meets international standards and regulations. The glycerol-based newly developed BIS (Online resource 7 in the ESM) could replace the commercial one. The use of glycerol as a carbon source for sporulation and as a recovery medium presents new opportunities for the development of a BIS product. The advantages of such applications include the profitable use of a biodiesel byproduct, and a decrease in the production cost of BI systems. Studies of the direct use of crude glycerol remains to be carried out and its purification cost must be considered.