Post-high-pressure temperature and time — Overlooked parameters in high pressure treatment of bacterial spores

High pressure (HP) processing has high potential for bacterial spore inactivation with minimal thermal input. To advance HP germination and subsequent inactivation of spores, this study explored the physiological state of HPtreated spores using flow cytometry (FCM). Bacillus subtilis spores were treated at 550 MPa and 60 C (very HP (vHP)) in buffer, incubated after the HP treatment, and stained for FCM analysis with SYTO16 indicating germination and propidium iodide (PI) indicating membrane damage. FCM subpopulations were analyzed depending on the HP dwell time (≤20 min), post-HP temperature (ice, 37 C, 60 C) and time (≤4 h), germination-relevant cortex-lytic enzymes (CLEs) and small-acid-soluble-proteins-(SASP)-degrading enzymes by using deletion strains. The effect of post-HP temperatures (ice, 37 C) was additionally studied for moderate HP (150 MPa, 38 C, 10 min). Post-HP incubation conditions strongly influenced the prevalence of five observed FCM subpopulations. PostHP incubation on ice did not or only slowly shifted SYTO16-positive spores to higher SYTO16 levels. At 37 C post-HP, this shift accelerated, and a shift to high PI intensities occurred depending on the HP dwell time. At 60 C post-HP, the main shift was from SYTO16-positive to PI-positive subpopulations. The enzymes CwlJ and SleB, which are CLEs, seemed both necessary for PI or SYTO16 uptake, and to have different sensitivities to 550 MPa and 60 C. Different extents of SASP degradation might explain the existence of two SYTO16-positive subpopulations. Shifts to higher SYTO16 intensities during post-HP incubation on ice or at 37 C might rely on the activity and recovery of CLEs, SASP-degrading enzymes or their associated proteins from reversible HPinduced structural changes. These enzymes seemingly become active only during decompression or after vHP treatments (550 MPa, 60 C). Based on our results, we provide a refined model of HP germination-inactivation of B. subtilis spores and an optimized FCM method for quantification of the safety-relevant subpopulation, i.e., vHP (550 MPa, 60 C) superdormant spores. This study contributes to the development of mild spore inactivation processes by shedding light on overlooked parameters: post-HP incubation conditions. Post-HP conditions significantly influenced the physiological state of spores, likely due to varying enzymatic activity. This finding may explain inconsistencies in previous research and shows the importance of reporting post-HP conditions in future research. Furthermore, the addition of post-HP conditions as HP process parameter may open up new possibilities to optimize HP-based inactivation of spores for potential industrial applications in the food industry.


High pressure treatment of bacterial spores
Mild inactivation of bacterial spores is a major challenge in food processing due to their extreme resistance. Bacterial spores are mainly formed by the genera Bacillus and Clostridium and are highly resistant to heat, dehydration, and chemical or physical treatments (Reineke and Mathys, 2020;Setlow, 2006;Setlow and Johnson, 2019). If spores survive a food decontamination treatment, they can eventually germinate and grow out depending on food composition and storage conditions (Wells-Bennik et al., 2016). This can lead to food spoilage or food safety problems (Wells-Bennik et al., 2016). Direct inactivation of spores usually demands intensive heat treatment, which negatively impacts food quality, in particular color, texture or nutritional value (Doona and Feeherry, 2007;Gratz et al., 2021;Ordóñez-Santos and Martínez-Girón, 2020). Mild, less heat-intensive spore inactivation strategies are therefore of high interest (Gratz et al., 2021;Jermann et al., 2015;Sevenich and Mathys, 2018). A promising mild strategy would be a germinationinactivation strategy based on high pressure (HP) processing (Delbrück et al., 2021a(Delbrück et al., , 2021bZhang et al., 2020). HP-triggered germination transforms resistant metabolically inactive dormant spores back to sensitive germinated spores (Zhang et al., 2020). The idea of the germination-inactivation strategy is therefore to trigger spore germination under HP so that spores lose their resistance and become susceptible to mild inactivation processes (Løvdal et al., 2011;Zhang and Mathys, 2019). The main advantages of isostatic HP compared to other germination stimuli, such as nutrients or exogenous dipicolinic acid (DPA), are that i) food can be treated homogenously, ii) no additional ingredient needs to be added to the food product and iii) the HP treatment can serve simultaneously as inactivation treatment of germinated spores as well as of other microorganisms, depending on the process parameters (Balasubramaniam et al., 2015;Zhang et al., 2020). The current model for physiological nutrient germination of Bacillales spores starts with activation of germinant receptors (GRs) in the inner membrane, resulting in release of ions and DPA from the spore core and partial core hydration. This leads to a loss of wet heat resistance. DPA release activates cortex-lytic enzymes (CLEs), which degrade a thick peptidoglycan layer called the cortex. The spore's core is fully hydrated.
Small acid-soluble proteins (SASPs) of the α/β-type that protect spore DNA are degraded enzymatically. Eventually, later germination events and outgrowth take place (Christie and Setlow, 2020;Moir, 2006;Setlow et al., 2017). The HP germination or inactivation mechanism depends on the applied HP process parameters, such as pressure, temperature and HP dwell time, and the bacterial species or strain (Black et al., 2005(Black et al., , 2007bDelbrück, 2022;Reineke et al., 2013b;Wuytack et al., 1998). Three HP-mediated germination or inactivation mechanisms have been classified according to three pressure and temperature ranges, where these mechanisms are predominant. The HP germination-inactivation model is based on data from Bacillus species, mainly from the germination model organism B. subtilis (Lenz and Vogel, 2015;Reineke et al., 2013b). The three ranges are as follows: i) Moderate high pressure (mHP) of 50-300 MPa and 30-50 • C triggers germination via GR activation similar to nutrients (Black et al., 2005;Delbrück et al., 2021a;Wuytack et al., 1998) and leads to incomplete spore inactivation even after long pressure dwell times (>1 h). ii) Very high pressure (vHP) at 400-600 MPa and moderate temperature (<60 • C) triggers germination by a direct release of DPA independently of GRs. Germination potentially stops after cortex hydrolysis and full hydration without SASP degradation (Paidhungat et al., 2002;Reineke et al., 2013b;Wuytack et al., 1998). Inactivation of germinated spores is inefficient under these conditions (Delbrück, 2022;Reineke et al., 2011). iii) vHP at >600 MPa and high temperature (>60 • C) also triggers the GR-independent release of DPA but possibly does neither allow cortex nor SASP degradation. The elevated temperature compared to the other pressure and temperature ranges leads to rapid inactivation (Reineke et al., 2013a(Reineke et al., , 2013b. A possible lack of cortex or SASP degradation in vHP-treated spores is assumed to be due to inactivation of cortex-or SASP-degrading enzymes by these HPs or temperatures (Black et al., 2007a;Reineke et al., 2013b;Wuytack et al., 1998). It is questionable if pathway iii) should be termed "germination" if this pathway only includes non-physiological DPA release but lacks other physiological germination steps (GR activation, cortex degradation or SASP degradation). Accordingly, Margosch et al. (2004) reported that the loss of DPA at combined vHP (600-800 MPa) and high temperature (>60 • C) is predominantly due to a physicochemical rather than a physiological process. The three HP germination or inactivation mechanisms above may be predominant at the stated pressure and temperature ranges, but an overlap of these mechanisms is also possible, depending on the HP parameters (Reineke et al., 2013b). Due to this overlap and a continuous transition from one predominant mechanism to the other (e.g., between 200 and 400 MPa (Reineke et al., 2012(Reineke et al., , 2013b(Reineke et al., , 2013c), pressure ranges for "moderate" and "very" HP are not clearly defined. Others have defined vHP as 400-800 MPa, for example (Black et al., 2007a). The classification of the HP germination or inactivation mechanisms described above is widely used but overly simplified. It does not consider other important influencing factors, such as the HP dwell time. After short HP treatment times (≤120 s), for example, cortex and SASP degradation seemed possible at 600 MPa and 60 • C, in contrast to the classification above (Reineke et al., 2012). In addition, the effect of post-HP conditions, i.e., the temperature and time after HP treatments, has not been considered or reported sufficiently, even if they may strongly influence HP-triggered enzymatic germination processes. This might have contributed to the contradictory results on the activity of CLE or SASP-degrading enzymes in HP-treated spores. Furthermore, a fraction of an isogenic spore population, called high pressure superdormant (HPSD) spores, might also not or only very slowly germinate under pressure due to heterogeneous spore behavior (Delbrück et al., 2021b;Zhang and Mathys, 2019). Please note that superdormancy is a relative term describing a subpopulation with decreased germination capacity compared to a germinated subpopulation. Therefore, "superdormant" is defined by the germination and isolation conditions in the corresponding experiment (e.g., type of germination stimulus, stimulus intensity, stimulus duration). In conclusion, HP superdormancy and the exact spore germination or inactivation mechanisms under HP are still not fully understood. Knowledge of these mechanisms and the physiological state of HP-treated spores is key to designing efficient and safe HP decontamination processes.

Flow cytometry analysis
Flow cytometry (FCM) is a powerful method to investigate the heterogeneous physiological states within an HP-treated spore population. It enables high-throughput analysis of several thousand individual cells within a few seconds. FCM uses a fluid stream to separate cells and measures the scattered and fluorescent light of each single cell. For the quantification of germinated, presumably inactivated or superdormant spores after HP treatment, the following two fluorescent stains are mostly used (Black et al., 2005;Borch-Pedersen et al., 2017;Mathys et al., 2007;Reineke et al., 2013a;Zhang et al., 2020): i) SYTO16, a membrane-permeant stain that shows a green fluorescence enhancement upon binding to nucleic acids by intercalation (Kong et al., 2010;Quyen et al., 2019;ThermoFisher, 2014); and ii) Propidium iodide (PI), a positively charged and thus membrane-impermeable dye that emits red fluorescence upon intercalation into nucleic acids (Kumar and Ghosh, 2019;Molecular Probes, 2006;Stiefel et al., 2015). SYTO16 or PI poorly stain (super)dormant spores due to their low inner membrane permeability and the presence of the cortex or DNA-binding proteins (α/β SASP) (Mathys et al., 2007;Zhang et al., 2020). In contrast, SYTO16 stains germinated spores with an intact inner membrane, whose nucleic acids are accessible due to cortex and SASP degradation (Black et al., 2005;Kong et al., 2010;Zhang et al., 2020). PI stains cells with a compromised inner membrane, i.e., potentially inactivated spores with a (partially) degraded cortex (Reineke et al., 2013a;Zhang et al., 2020). In membrane-compromised cells, both dyes can theoretically enter the spore core. However, PI emits stronger fluorescence than SYTO16, because i) PI has a stronger binding affinity to DNA and might (partially) displace SYTO16, or as ii) SYTO16 might be quenched by fluorescence resonance energy transfer to PI, similar to SYTO9 (Mathys et al., 2007;Stocks, 2004). The staining behavior of HP-treated spores and thus interpretation of FCM data is still not fully clear. For example, two SYTO16-positive subpopulations with high and intermediate SYTO16 intensity have been observed after HP treatments (Borch-Pedersen et al., 2017;Mathys et al., 2007) and have been categorized as germinated spores with "high physiological fitness" and "partial sublethal damage", respectively, after mHP treatments (150 MPa, 37 • C, 10 min, B. subtilis (Zhang et al., 2020)). However, the structural nature of this damage is not known. Furthermore, an FCM protocol worked for superdormant spore quantification after mHP treatments (150 MPa, 37 • C, ≤40 min, B. subtilis, (Zhang et al., 2020)). However, the same FCM protocol led to irreproducible FCM data after vHP treatments (550 MPa, 60 • C, 1 min, B. subtilis) in our laboratory due to the similar staining behavior of vHPgerminated and superdormant spores. The limited ability to distinguish between superdormant and inactivated or germinated spores prevents the accurate quantification of superdormant spores using FCM. However, accurate analysis and characterization of HPSD spores is essential. HPSD spores prevent the industrial implementation of the HP-based germination-inactivation strategy since these ungerminated spores can survive the mild inactivation step. HPSD spores could consequently grow out during storage of a food product and cause food safety or spoilage issues. Out-growing superdormant spores could also present an issue in other sectors (medical, pharmaceutical, and (bio)chemical) (Delbrück et al., 2021b).

Aim
The aim of this study was to obtain a better understanding of the FCM methodology for HP-treated spores and ultimately of HP germination and inactivation of bacterial spores. We investigated the dependence of FCM spore subpopulations on i) the HP dwell time at 550 MPa and 60 • C, ii) the post-HP temperature and time and iii) on CLEs and SASPdegrading enzymes. A HP treatment temperature of 60 • C was chosen as the temperature optimum for Bacillus vHP germination (500 MPa) has been previously shown to be 60 • C, as measured by DPA release and SYTO16 staining (Black et al., 2007b). The effect of post-HP incubation (ice, 37 • C) was also investigated for moderate HP (150 MPa, 38 • C, 10 min). Based on these experiments, we were able to separate the superdormant subpopulation from other subpopulations. Hence, we provide an optimized FCM protocol for the reproducible quantification of vHP (550 MPa, 60 • C) superdormant spores. Furthermore, this study helps to elucidate the physiological state of FCM subpopulations and allows to refine the HP germination and inactivation mechanisms of B. subtilis. Our work thereby contributes to the development of a potential HPbased germination-inactivation strategy for bacterial spore elimination with minimal thermal input.

Bacterial strains and spore preparation
The B. subtilis strains used in this study are listed in Table 1. Strains were recovered from cryocultures on tryptic soy agar (TSA) plates and grown overnight in tryptic soy broth (TSB) with or without antibiotics (Table 1). Spores were sporulated and harvested as described by Zhang et al. (2020). Mutant strains and the wild-type strains to be compared were sporulated at the same time. For experiments with mutants, mutant and wild-type spores were purified using a Nycodenz® (Axis-Shield, Oslo, Norway) density gradient centrifugation according to Delbrück et al. (2022). Spore suspensions showed ≥97% dormant spores as determined by FCM or phase contrast microscopy (DM6, Leica Microsystems, Wetzlar, Germany). Spores were kept in sterile Milli-Q water at 4 • C and washed every 14 days. The same spore batch was used for experimental replicates. Two different PS832 spore batches were analyzed in total (batch 1 used for Sections 3.1, 3.2.2 and 3.2.3; batch 2 used for Section 3.2.1). For Section 3.2.2, two batches of BKK02600 and BKK22930 were analyzed.

High pressure treatment
To trigger germination or inactivation, dormant spores were exposed to HP at 150 MPa and 38 • C for 10 min or at 550 MPa and 60 • C for ≤20 min. HP samples contained approximately 10 9 CFU/mL spores in 50 mM N-(2-acetamido)-2-aminoethanesulfonic acid (ACES) buffer (pH 7.0; ThermoFisher, Kandel, Germany). The spore concentration was approximately 10 8 CFU/mL for experiments with mutant strains due to a small yield of mutant spores in the sporulation process. A spore suspension volume of 1.5 mL was HP treated in 1.5 mL cryotube vials (Nunc A/S, Roskilde, Denmark). For parallel treatment of two spore samples, 0.6 mL spores were HP treated in 0.5 mL reaction tubes (Eppendorf, Hamburg, Germany). Vials and tubes were sealed using sealing tubes (Nunc CryoFlex Tubing, Nunc A/S, Roskilde, Denmark). The temperature of the HP unit (modified Model U111, Unipress, Warsaw, Poland) was controlled by immersing the two vessels in a water bath (Huber CC410, Offenburg, Germany), as described by Zhang et al. (2020). The temperature during HP treatment was measured in the control vessel using a thermocouple inside of a dummy tube containing 50 mM ACES buffer. The number and size of dummy tubes in the control vessel were equal to those of sample tubes in the sample vessel. Sample and dummy tubes were loaded in the two vessels exactly at the same time to ensure that the measured temperature profiles were representative of those in spore sample tubes. Propylene glycol (40% v/v) was used as the pressure-transmitting medium. The compression and decompression rates were approximately 15 MPa/s and 45 MPa/s, respectively, as shown by the representative pressure/temperature profiles in Supplementary material Fig. A.1. Untreated spores ("No HP" control) were kept on ice during HP treatments.

Post-HP incubation
To investigate the influence of post-HP incubation on the physiological state of spores, the total volume of an HP-treated samples was split in aliquots. The different aliquots were incubated at different temperatures (on ice, at 37 • C or 60 • C) for up to 4 h. The spores HPtreated for 50 s were analyzed immediately after HP treatment, while the spores HP-treated for 6 min were analyzed for the first time after 15-20 min post-HP incubation to allow FCM measurements of both samples within a day. During the following 4 h of post-HP incubation, 20 μL of the aliquots were taken every 20 or 30 min for immediate FCM analysis. In later experiments, aliquots were placed on ice after a defined incubation time at 37 • C or 60 • C before FCM analysis, as further incubation on ice did not influence the FCM results. The "No HP" control was aliquoted and incubated in parallel with HP-treated samples.

Flow cytometry analysis
To analyze and quantify various subpopulations of spores after HP treatment, a cytometry-based method was used as described in detail by Zhang et al. (2020). In short, spores at a concentration of 10 7 CFU/mL were stained with PI (1.5 μM; Molecular Probes, Leiden, Netherlands) and SYTO16 (0.1 μM; Molecular Probes, Leiden, Netherlands) and analyzed in an LSR Fortessa cytometer (BD Biosciences, Franklin Lakes, USA). The number of recorded events was 10,000 or 15,000. The event rate was 6000 ± 4000 events/s. The following controls were measured (Supplementary material Fig. A.2): i) stained and unstained 0.1 μmfiltered Milli-Q water to assess background signals, ii) untreated unstained spores as control for autofluorescence and sample homogeneity, iii) untreated stained spores to set the gate for (super)dormant spores and as control of dormant spore purity. FCM data were visualized as pseudocolor density plots using fcs files as raw data, FlowJo V10.6.1 software (FlowJo LLC, Oregon, US) and a gating scheme similar to Zhang et al. (2020). The absolute position of gates in SYTO16-vs.-PI plots was adapted to the absolute position of subpopulations due to variations in PI-signal height (H) and SYTO16-H between samples. However, the relative position of gates in these plots was kept constant for each experiment. HP treatment, post-HP incubation and FCM analysis were performed within one day and repeated independently for at least three days.

Plate count for superdormant spore quantification
In addition to FCM, plate count was used for superdormant spore quantification. Plate count can be used over a broad range of HP dwell times due to a lower limit of detection compared to FCM. Superdormant spores are defined by this method as the heat resistant fraction within an HP-treated sample. A "No HP" control and HP-treated samples were heat treated at 80 • C for 10 min to inactivate all cells except (super)dormant spores. HP-treated samples were heated immediately after each HP treatment. Spores were diluted using maximum recovery diluent (Sigma Aldrich, Steinheim, Germany) in log 10 -steps, plated on TSA and grown for 48 h at 37 • C. The relative number of superdormant spores was calculated by dividing the concentration of colony-forming units of an HP-and heat-treated sample by those of the heat-treated "No HP" control. HP treatments and plate counts were repeated for three days.

Dynamics of spore subpopulations depending on the HP dwell time
To better understand the effect of vHP on spores, the dynamics of FCM spore subpopulations were analyzed after treatments at 550 MPa and 60 • C for 0-20 min followed by post-HP incubation on ice.
Five spore subpopulations (R1-R5) were observed depending on the HP dwell time, i.e., the holding time at 550 MPa, (Fig. 1) and post-HP incubation conditions. Previous literature assigned i) PI-and SYTO16negative spores (R1) as mainly culturable (super)dormant or presumably unculturable spores with inactivated CLEsdepending on HP treatment conditions, ii) spores with high SYTO16 signals (R2) as highly culturable, germinated spores with presumably hydrolyzed cortex and hydrolyzed α/β SASPs, iii) spores with medium SYTO16 signals (R3) as germinated spores with partial growth capacity or germinated sublethally injured spores and iv) PI-positive spores (R4, R5) as membranecompromised, i.e., mostly inactivated spores with a partially degraded cortex (Borch-Pedersen et al., 2017;Mathys et al., 2007;Reineke et al., 2012Reineke et al., , 2013aZhang et al., 2020). Spores with medium SYTO16 signals appeared as one subpopulation, whereas Borch-Pedersen et al. (2017) observed two distinct subpopulations with medium SYTO16 signals. A differentiation between two PI-positive subpopulations, as seen in this study for HP dwell times of 4-20 min ( Fig. 1), has not been reported before. An explanation could be that different HP treatment conditions (pressure, temperature, HP dwell time or post-HP incubation) were investigated in this study and previously. The above stated growth capacity of subpopulations refers to the ability of mHP-treated spores to germinate and grow in a nutrient-rich medium (150 MPa, 37 • C, 10 min (Zhang et al., 2020)), but has not yet been validated for vHP-treated spores. Validation of the growth capacity of FCM subpopulations for each chosen treatment condition is important as PI-positive signals do not necessarily mean cell death (Rosenberg et al., 2019;Shi et al., 2007).
Germination was increasingly triggered with increasing HP dwell time. The number of unstainable (super)dormant spores (R1) decreased, and the number of germinated spores (R2) increased until subpopulation R1 almost disappeared ( Fig. 1: A, B, C). Meanwhile, the subpopulation with medium SYTO16 intensities (R3) started to increase ( Fig. 1: B, C). Then, the number of PI-positive, presumably inactivated spores (R4, R5) started to increase ( Fig. 1: D). These temporal shifts resembled the threestep inactivation model for mHP-treated spores (150 MPa, 37 • C, 3-120 min) describing i) a germination step (shift R1 → R2), ii) an intermediate step of potentially sublethal damage (shift R2 → R3), and iii) inactivation (shift to PI-positive region) (Borch-Pedersen et al., 2017;Mathys et al., 2007;Reineke et al., 2012;Zhang et al., 2020). However, after longer vHP treatments (i.e., approximately 2.5 min), the dynamics differed between mHP-and vHP-treated spores. More PI-positive spores (R4, R5) were detected compared to those in mHP treatments. For example, 9.8 ± 0.3% were PI-positive (R4 + R5) after 10 min at 550 MPa, 60 • C and post-HP incubation on ice, but only 4.3 ± 0.9% after 10 min at 150 MPa, 38 • C and post-HP incubation on ice (n = 3, Supplementary material Fig. A.4: M, Fig. A.7). This could indicate a stronger effect of vHP (550 MPa, 60 • C) on the inner membrane than of mHP (150 MPa, 37-38 • C), which is quite plausible (Reineke et al., 2013a). Furthermore, two instead of one PI-positive subpopulation (R4, R5) appeared ( Fig. 1: D). In addition, spores reappeared in region R1 with increasing vHP dwell time, leading to a distinct, unstainable subpopulation in R1 after 4 min under vHP ( Fig. 1: D, E, F). This reappearing R1 subpopulation, however, was not superdormant spores according to plate count; after 2.5 min or 20 min HP dwell time, the number of superdormant spores was reduced by 4.8 ± 0.3 or 6 ± 1 log 10 units, respectively (Supplementary material Fig. A.3). Hence, superdormant spores were still present after these vHP dwell times but below the limit of detection of FCM of ca. <0.1% of total events corresponding to >3 log 10 units of dormant spore reduction. Superdormant spores should therefore not be detectable in gate R1 under these conditions. A reappearing subpopulation (R1) has been reported before and has been hypothesized to be spores with inactivated CLEs that have lost most of their DPA (Mathys et al., 2007;Reineke et al., 2013aReineke et al., , 2013c. To validate this, the role of cortex and SASP degradation in the appearance of FCM subpopulations was analyzed (Sections 3.2.2 and 3.2.3).

Dynamics of spore subpopulations depending on the post-HP incubation and cortex-or SASP-degrading enzymes
The appearance of FCM subpopulations was not reproducible in initial vHP germination experiments (550 MPa, 60 • C, ≤20 min), when post-HP incubation conditions were not defined. PI and SYTO16 staining are assumed to be influenced by germination enzyme activity, which is why a varying extent of enzyme activity was hypothesized to have caused irreproducible FCM plots. Enzymatic activity is generally temperature dependent. Hence, the effect of post-HP temperature and time on vHP-treated spores was analyzed to confirm that germination enzyme activity can proceed after HP treatments and influences FCM results. Spores were treated at 550 MPa and 60 • C for 50 s or 6 min. Aliquots of each sample were incubated immediately after the vHP treatment at temperatures with expected low (ice, 60 • C) or high (37 • C) enzymatic activity for up to 4 h before FCM analysis (Section 3.2.1). The effect of the post-HP temperature (ice or 37 • C for 2.5 h) on mHP-treated spores (150 MPa, 38 • C, 10 min) was also studied. To confirm that the FCMrelevant germination enzymes were CLEs and SASP-degrading enzymes, wild-type and mutant spores without these enzymes were sporulated, vHP-treated (550 MPa, 60 • C) and post-HP incubated (Sections 3. 2.2 and 3.2.3). A 45-50 s or 6 min HP dwell time was chosen to represent all different subpopulations observed between 0 s and 20 min dwell time (Fig. 1).

Post-HP incubation temperature and time
FCM spore subpopulations were strongly influenced by incubation time and temperature after vHP treatments (550 MPa, 60 • C). There were two subpopulation shifts along the SYTO16 axis during post-HP incubation at ≤60 • C: i) a shift from R3 to R2 and ii) a shift from R1 to R3 (see shifts in Fig. 2: C, E). The shift R3 ➔ R2 was observed after vHP treatments of 6 min or shorter and the shift R1 ➔ R3 was observed after vHP treatments of 4 min or longer (Fig. 2: E, Supplementary material Fig. A.4). On ice, subpopulations were most stable, and only a slight shift R3 ➔ R2 within the first 2.5 h after a short vHP treatment (50 s) was observed (Supplementary material Fig. A.5: A-C). Regardless of this shift, the superdormant subpopulation (R1) was not separated from the germinating subpopulation in R3 after incubation on ice ( Fig. 1: C). After a longer vHP treatment (6 min The increase in SYTO16 levels during post-HP incubation on ice or at 37 • C might indicate that enzymatic reactions (cortex or SASP degradation, see Sections 3.2.2 and 3.2.3) were triggered by vHP treatment and continued after vHP treatment. A continuing germination of vHPtreated spores has also been observed in terms of DPA release (600 MPa, 40 • C, ≤60 min (Reineke et al., 2013c)). Under chosen conditions (550 MPa, 60 • C), cortex degradation might even only start during or after decompression from 550 MPa. If cortex degradation was possible under a pressure of 550 MPa and 60 • C, most spores would have been PIor SYTO16-positive after long HP dwell times immediately after decompression. This was not the case after a 20 min HP dwell time ( Fig. 1: F).
Along the PI-axis, there was no clearly visible shift of subpopulations after vHP treatment (550 MPa, 60 • C) for HP dwell times <6 min and post-HP incubation at ≤37 • C for 2.5 h (e.g., Fig. 2: C). However, after a 6 min HP dwell time, the reappearing subpopulation (R1) mainly shifted to medium PI intensities (R4) at 37 • C, while the highly PI-positive subpopulation (R5) hardly increased ( Fig. 2: E). The post-HP temperature-induced shift R1 ➔ R4 at 37 • C diminished after a 20 min HP dwell time (Supplementary material Fig. A.4: G, N). The dependency of the shift R1 ➔ R4 on the HP treatment time suggests that this shift could be due to an enzymatic reaction, as HP can influence enzyme activity by changing their protein structure (Balasubramaniam et al., 2015). The shift R1 ➔ R4 further indicates that spores in R1 after a 6 min dwell time at vHP are not superdormant spores, as the proportion of the superdormant (metabolically dormant) spore population should be independent of post-HP incubation at temperatures ≤60 • C. A stable size and shape of subpopulation R1 were indeed observed for a short HP dwell time (50 s) after separation from other subpopulations. Subpopulation R1 after a 50 s HP dwell time was hence identified as superdormant spores. An additional indication that this subpopulation is superdormant was the accordance of the percentage of superdormant spores measured by plate count (e.g., 7.8 ± 1.7%, n = 3) to the percentage of events in R1 relative to all events in SYTO16-PI plots measured by FCM (e.g., 6.8 ± 0.6% for post-HP incubation at 37 • C, n = 10).
Post-HP incubation at 60 • C shifted vHP-germinated spores (R2, R3) to PI-positive regions (R4, R5) ( Fig. 2: D, F). This indicates increasing membrane damage and heat inactivation of germinated spores (Borch-Pedersen et al., 2017;Mathys et al., 2007;Reineke et al., 2012). Spores in region R3 had almost disappeared after 20 or 40 min post-HP incubation at 60 • C in the 50 s vHP-treated sample (similar to Fig. 2: D), as they were shifted to region R2 before shifting to PI-positive regions during later post-HP incubation. In contrast, spores in R3 were still present in the 6 min vHP-treated sample even after 4 h post-HP incubation at 60 • C (Fig. 2: F). The fast disappearance of spores in region R3 in the 50 s vHP-treated sample led to a fast separation of superdormant spores (R1) from other subpopulations. Hence, vHP superdormant spores can be quantified best by FCM after post-HP incubation at 60 • C for 20-40 min.
At mHP (150 MPa, 38 • C, 10 min), the FCM results also depended on post-HP incubation. For example, the shift R3 → R2 to higher SYTO16 levels was also observed for post-HP incubation at 37 • C for 2.5 h of wild-type-like B. subtilis PS533 spores (Supplementary material Fig. A.7). A continuing germination after mHP treatments was also reported by Zhang et al. (2020) in terms of increasing SYTO16 fluorescence during the first 90 min on ice after decompression from 150 MPa, 37 • C, 3 min or by Kong et al. (2014) in terms of decreasing phasecontrast brightness (140 MPa, 37 • C, 30 s).

CLEs
Two semiredundant CLEs are responsible for cortex depolymerization in germinating spores of Bacillus species, namely, SleB and CwlJ, and some species have multiple CwlJ homologs (Christie and Setlow, 2020). Our results support the general understanding that cortex hydrolysis by CLEs is required for the appearance of PI-or SYTO16-positive FCM subpopulations (R2-R5) (Black et al., 2005;Kong et al., 2010;Reineke et al., 2012). Without SleB and CwlJ, >99% of mutant spores remained PI-and SYTO16-negative after both vHP treatments (45 s or 6 min) (Fig. 3: I-L). The results therefore support that the PI-and SYTO16negative subpopulation (R1), which appeared after longer HP dwell times (≥4 min, Fig. 1), could be spores with inactive CLEs. Furthermore, the increase in subpopulation R1 after ≥4 min HP dwell time (Fig. 1) likely indicates an increase in CLE inactivation with an increase in HP dwell time at 550 MPa and 60 • C.
When one of the two CLEs is missing, cortex degradation relies on the other remaining enzyme. Hence, cortex degradation in ΔsleB and ΔcwlJ mutants depends on CwIJ or SleB alone, respectively. With only one CLE, cortex hydrolysis seemed somewhat possible but less effective, as fewer spores were detected in region R2 compared to the wild-type after 45 s under pressure and incubation on ice ( Fig. 3: A, M, Q). The absence of PI-or SYTO16-positive ΔsleB spores after 6 min under vHPindependent of the post-HP temperatureindicated that CwlJ might be inactivated after 6 min under vHP ( Fig. 3: S, T). Spores without CwlJ, however, shifted to higher PI and SYTO16 intensities during post-HP incubation at 37 • C, suggesting at least partial activity of SleB after 6 min under vHP (Fig. 3: O, P). Hence, CwlJ seemed more pressure sensitive than SleB in our study. Contradicting vHP sensitivities of the CLEs have been reported in previous studies using SYTO16 staining; SleB was suggested to be less active than CwlJ upon vHP treatment (550 MPa, 37 • C, 60 min, post-HP conditions unknown (Reineke et al., 2012(Reineke et al., , 2013a; 500 MPa, 50 • C, ≤5 min, post-HP frozen for unknown time (Black et al., 2007b)). Due to this contradiction, our experiments were repeated from sporulation to FCM analysis, but the results of both biological replicates were consistent. It is difficult to explain the reason for the conversely measured pressure sensitivities of SleB and CwlJ in our study and in previous studies. This could be because different strains were used previously (B. subtilis FB114, FB115; lack additionally nutrient-germinant receptors), or different treatment conditions (pressure, HP temperature, HP dwell time, post-HP incubation conditions) were applied. It has been thought that CwlJ is more temperature sensitive than SleB (90 • C, ≤60 min, Atrih and Foster, 2001). However, heat sensitivity of enzymes might not necessarily correlate to HP sensitivity, as heat and pressure sensitivity of a whole spore do not necessarily correlate (Margosch et al., 2004;Nakayama et al., 1996;Nguyen et al., 2011).  3. Dynamics of spore subpopulations in regions R1-R5 depending on the post-high-pressure-(HP) temperature and cortex-lytic enzymes (CLEs). Wild-type (WT) and mutant spores, lacking one or both CLEs (CwlJ and SleB), were HP treated at 550 MPa and 60 • C for 45 s or 6 min. Strain 168 is WT to BKK22930 and BKK02600 and PS832 to FB113, respectively. Aliquots of HP-treated samples were incubated at 37 • C or on ice after HP treatment. Arrows in flow cytometry plots highlight shifts of subpopulations during post-HP incubation at 37 • C compared to incubation on ice. The experiment was repeated independently for at least three days, and representative flow cytometry plots are depicted. See Supplementary material Fig. A.2 for FCM controls and Table A.3 for percentages of events in each subpopulation.
HP treatments of mutant spores suggest that post-HP incubation temperature is important for CLE activity, especially those of SleB. Incubation at 37 • C compared to 4 • C (ice) after 45 s under pressure led to more highly SYTO16-positive spores (R2) in the ΔsleB mutant ( Fig. 3: Q, R) and to even more in the ΔcwlJ mutant, in which cortex degradation relied on SleB (Fig. 3: M, N). CLE activity and thus the post-HP temperature seemed to also be important for the PI-positive subpopulation R4 in wild-type spores; the shifts from R1 to R3 or R4/R5 during post-HP incubation at 37 • C of 6 min vHP-treated wild-type spores (Fig. 3: D, H) indicate that CLEs might be inactive in spores in region R1 but can become somewhat active during post-HP incubation at 37 • C.
Generally, we hypothesize that post-HP incubation on ice or at 37 • C facilitates recovery from reversible, HP-induced structural changes in proteins. It possibly contributed to higher SleB, CwlJ and SASPdegrading enzyme activity (Section 3.2.3) at 37 • C compared to post-HP incubation on ice. HP can change the secondary or tertiary protein structure, which can either positively or negatively affect enzyme function (Balasubramaniam et al., 2015;Queirós et al., 2018). HPinduced structural changes, i.e., HP enzyme inactivation, can be either reversible or irreversible (Balny and Masson, 1993;Juarez-Enriquez et al., 2015). Protein conformation and activity depend on the temperature-time combination (Akasaka, 2006). If CLE activity was not affected by HP treatment and was increased by the post-HP temperature alone, ΔsleB spores would have been PI-or SYTO16-positive after post-HP incubation at 37 • C independent of the HP dwell time. This was not the case ( Fig. 3: R, T). Hence, vHP seemed to have altered the enzyme structure and activity of CLEs, or at least of CwlJ. A temperature of 37 • C, which is the optimal growth temperature of B. subtilis (Ratkowsky et al., 1983), might stabilize the native protein conformation. A temperature of 37 • C might thereby promote back-folding from the HPinduced structure to the native protein conformation in contrast to temperatures of 4 • C (ice) or 60 • C. This recovery of proteins could then result in increased enzyme activity at 37 • C compared to 4 • C (ice) or 60 • C, as indicated by an increase in PI-or SYTO16-positive spores (Figs. 2, 3, 4). It seems reasonable that the longer the vHP treatment time is, the more the protein structure of CLEs changes. Hence, recovery from such structural changes might take longer, or proteins might be irreversibly changed the longer the HP treatment. Such reduced enzyme recovery might explain the decrease in SYTO16-(R2, R3) or PI-positive (R4) spores with increasing HP dwell time (Fig. 1). The time for enzyme recovery could also explain the retarded germination in terms of DPA release at 600 MPa and 40 • C (Reineke et al., 2013c). Different proteins might not recover equally well from vHP-induced changes, as different proteins have different structures and thus change differently under vHP. Hence, the less pronounced shift R3 → R2 during post-HP incubation at 37 • C in ΔsleB spores compared to ΔcwlJ spores could imply that CwlJ was more damaged by vHP or could recover less well from vHP damage. However, lowered CLE activity in wild-type spores could also be the result of indirect CLE inactivation due to HP inactivation of CLE-associated proteins. For example, SafA, GerQ, CotE or YlxY seem to influence CwlJ, while YpeB, YlaJ and YhcN seem to influence SleB (Amon et al., 2020;Christie and Setlow, 2020;Johnson and Moir, 2017;Sayer and Popham, 2019). Further studies are required to validate these hypotheses.
The results only allow statements about SASP-or cortex-degrading enzymes, but other enzymes might theoretically also better recover at a temperature in the range of the optimal bacterial growth temperature. Hence, the recovery of a whole cell might be improved by such temperatures. Indeed, a temperature dependence of recovery after HP treatments was observed for vegetative cells of Listeria or Salmonella species. More viable cells were present at increased storage temperatures (4 • C vs. 20-30 • C) (Ritz et al., 2006;Schottroff et al., 2018). This implies that the time and temperature between germination and subsequent inactivation treatment must be well adjusted to prevent the growth of non-irreversibly inactivated spores or sublethally injured vegetative bacteria.

SASP-degrading enzymes
SASP-degrading enzymes might account for the difference between subpopulations with medium (R3) or high (R2) SYTO16 levels. The α/β-type SASP proteins bind to the minor groove of DNA and thereby shield the DNA from other molecules such as SYTO16 (Ki et al., 2008;Kong et al., 2010). In 1S111 spores lacking the majority of α/β SASP proteins (lack approximately 80%, Setlow, 2006), SYTO16 intensities were visibly increased, as more spores were highly SYTO16-positive (R2) compared to the wild-type after 45 s vHP treatment and post-HP incubation on ice ( Fig. 4: A, E) or after 6 min HP treatment and incubation at 37 • C ( Fig. 4: D, H). Accordingly, Kong et al. (2010) found that the maximum SYTO16 level was greater with a reduced amount of α/β SASPs during nutrient germination. In contrast, when SASP could not be degraded by the protease GPR, highly SYTO16-positive spores (R2) were not detected after 45 s of HP dwell time and post-HP incubation on ice ( Fig. 4: I), similar to Kong et al. (2010). However, high SYTO16 levels (R2) were detected after post-HP incubation at 37 • C even without GPR ( Fig. 4: J). This was unexpected according to Kong et al. (2010) but might be explained by the existence of further, slow SASP-degrading activity, e.g., by YyaC (Sanchez-Salas et al., 1992;Wetzel and Fischer, 2015). In conclusion, spores in region R2 are likely spores with active SASP-degrading enzymes and active CLEs.
Fluorescence is greater when SYTO16 binds to DNA instead of RNA or other cytoplasmic compounds (ThermoFisher, 2014). Spores with medium SYTO16 intensities (R3) might therefore be spores where SYTO16 is taken up in the spore core but not bound to DNA. The DNA of spores in R3 might be blocked by α/β SASP proteins due to inactive or hardly active SASP-degrading enzymes. The cortex must be at least partially degraded in spores of region R3 (see Section 3.2.2). Accordingly, a shift from R3 to R2 during post-HP incubation would indicate SASP degradation, potentially combined with proceeding cortex degradation or water uptake. There was little or no shift from R3 to R2 in wild-type spores after 6 min under pressure and post-HP incubation at 37 • C (Figs. 2: E, 4: D). Only spores lacking most α/β SASP proteins were detected in region R2 under these treatment conditions (Fig. 4: H). This implies that SASP-degrading enzymes might be inactivated by 6 min at 550 MPa and 60 • C and could not recover activity at 37 • C post-HP. Inactivation of SASP-degrading enzymes by vHP has been reported for 600 MPa and 40 • C for 60 min and unknown post-HP conditions (Wuytack et al., 1998). A link between subpopulation R3 and incomplete cortex or SASP degradation has been speculated for vHP-treated spores (Reineke et al., 2012). At mHP, however, most spores ended up in region R3 (150 MPa, 37 • C, 10-40 min (Zhang et al., 2020)), even if SASP seemed to be degraded in these spores (Reineke et al., 2013a;Wuytack et al., 1998). These contradictory findings could likely be explained by a post-HP dependent, reversible mHP inhibition of SASP degradation. While post-HP incubation on ice (≤6 h) after mHP treatment at 150 MPa and 38 • C for 10 min led to spores predominantly in region R3, indicative of no or only partial SASP degradation, post-HP incubation at 37 • C (2.5 h) shifted these spores to R2, indicative of complete SASP degradation (Supplementary material Fig. A.7). Hence, SASP degradation was likely observed in previous mHP experiments, depending on the exact (unreported) temperature and time between HP treatment and further analysis.
The PI-positive subpopulation R4 could be linked to cortex degradation rather than to enzymatic SASP degradation, as i) CLE activity seemed to be required for any PI uptake and ii) Δgpr spores, in which SASP-degrading enzymes seemed inactive, were detected in region R4 (Fig. 4: I). Spores in R4 might be membrane-compromised spores with a (partially) degraded cortex. The exact difference between medium (R4) and highly (R5) PI-positive spore levels is not clear. It could be a difference in the extent of cortex or SASP degradation. The latter could be indicated by a higher number of spores in R5 than in R4 in mutant spores lacking most α/β SASP proteins compared to the wild-type after 45 s HP dwell time and post-HP incubation on ice ( Fig. 4: A, E). The growing subpopulation R5 (Fig. 1) would consequently indicate increasing SASP degradation with increasing HP dwell time. This conflicts with the hypothesis that SASP-degrading enzymes are increasingly inhibited the longer the HP treatment. Other reasons for the difference between R4 and R5 might be the extent of membrane damage, the binding mode of PI (e.g., DNA or RNA, nucleotide intercalating or bound to DNA backbone), the molecular environment of PI (e.g., hydration state) or a combination of all factors. Further research is needed to unravel the exact reasons for two PI-positive subpopulations.

Updated high pressure germination or inactivation pathways
The HP germination or inactivation model for B. subtilis spores previously described the following sequence of germination steps: germinant receptor (GR) activation (triggered by mHP or nutrient germinants), release of ions and DPA (directly triggered by vHP), partial core hydration, cortex hydrolysis and full core hydration, SASP degradation, and inactivation. It was differentiated between vHP pathways at i) 400-600 MPa combined with moderate temperature (<60 • C), which potentially stops after cortex hydrolysis and full core hydration and does not lead to SASP degradation or inactivation, and ii) >600 MPa combined with high temperatures (>60 • C), which leads to inactivation without cortex or SASP degradation (Reineke et al., 2013b;Reineke and Mathys, 2020). Accordingly, spores should follow the vHP pathway i) at 550 MPa and 60 • C. However, our results show that the vHP pathway ii) was also possible at 550 MPa and 60 • C, for example, and that this model is too simplified. Our refined model of the HP germination or inactivation pathways (Fig. 5) therefore highlights that i) spores can follow different pathways with different probabilities at one pressure-temperature combination; the most likely pathway depends on the pressure-temperature combination, the HP dwell time and post-HP temperature and time, ii) cortex and SASP degradation are not a prerequisite for inactivation (in contrast to DPA release (Reineke et al., 2013c)), iii) cortex hydrolysis proceeds SASP degradation, but SASP degradation can already begin when cortex hydrolysis is not completed (Kong et al., 2010), iv) cortex and SASP degradation could not or only partially occur depending on post-HP conditions and on whether vHP has changed involved enzymes reversibly or irreversiblythe later seems more likely the longer the vHP treatment, v) lack of cortex or SASP-degradation could be not only due to HP-inactivation of CLEs or GPR proteins but also HP-inactivation of CLE-associated proteins or additional SASP-degrading enzymes, and vi) the triggered germination steps or inactivation also depend on heterogeneous spore properties within a spore population. Superdormant spores do not germinate, while other spores can already be inactivated by the same HP treatment conditions. It is important to highlight that the extent of HP germination and inactivation are also influenced by other factors, such as the suspending medium and its pH or the HP compression and decompression rate (Black et al., 2007a;Delbrück et al., 2021a;Fekraoui et al., 2021). Few spores might follow the mHP germination pathway during vHP treatment, as the mHP pressure range is passed during compression to vHPs. An important difference between mHP and vHP is that CLEs seem to be active during mHP treatments (150 MPa, 37 • C, 5 min (Reineke et al., 2013a)) in contrast to vHP treatments (550 MPa, 60 • C, ≤20 min). This might explain why mHP treatments do not necessarily require post-HP incubation at ≥37 • C for the separation of superdormant (R1) and germinated (R3) spores in FCM plots (Zhang et al., 2020). Another difference in mHP and vHP germination is that inactivation of germinated spores at mHP (150 MPa, 37 • C) is possible but slower compared to vHP (550 MPa, 60 • C) (Black et al., 2007b;Borch-Pedersen et al., 2017). A new finding of our study regarding mHP germination (150 MPa, 38 • C, 10 min) is that SASP degradation in mHP-treated spores also seems post-HP-incubation dependent. Hence, mHP can seemingly reversibly change SASP-degrading enzymes.

Conclusion
This study investigated the influence of the HP dwell time, post-HP incubation conditions and CLEs or SASP-degrading enzymes on the physiological state of B. subtilis spores after mHP (150 MPa, 38 • C) and vHP (550 MPa, 60 • C) treatments using FCM. The main conclusions are as follows: i) Post-HP temperature and time matter. Depending on the HP dwell time and post-HP conditions, different shifts of FCM subpopulations were observed. Shifts to higher SYTO16 intensities during post-HP incubation on ice or at 37 • C might rely on the activity and recovery of CLEs or SASP-degrading enzymes or their associated proteins from reversible HP-induced structural changes. Future research on HP germination or HP enzyme inactivation should refer to all HP treatment parameters (pressure, temperature, time profile) and post-HP conditions (temperature, time). This will allow to differentiate between reversible and irreversible enzyme inactivation by HP and to avoid ambiguous statements about the effect of HP on enzymes. ii) Optimization of post-HP conditions (60 • C, 40 min) enables fast, reproducible FCM quantification of vHP (550 MPa, 60 • C) superdormant spores, which limit the development of an HPbased germination-inactivation strategy. iii) The difference between the spores with medium (R3) or high (R2) SYTO16 signals might be mainly caused by different extents of SASP degradation. iv) Two PI-positive subpopulations have been observed for the first time, but their physiological differences are not fully clear. v) CwlJ seemed more pressure sensitive than SleB in our study, contradicting previous studies. vi) Based on our results, the current HP germination or inactivation model was updated and refined.
Post-HP incubation conditions as additional parameters in HP processing open up new possibilities to optimize the HP-based germinationinactivation strategy or HP-based protein modification. The conditions between germination and inactivation must be well adjusted to maximize inactivation and prevent outgrowth or potential toxin production of germinated spores.
Our work contributes to the potential development of a HP-based germination-inactivation strategy for bacterial spore elimination with minimal thermal input. Regarding the HP germination or inactivation mechanisms, the exact molecular pathways still need to be fully elucidated. This is essential for efficient further development of the HP germination-inactivation strategy. In addition, differences in HP germination and inactivation between different spore formers (other Bacilli or Chlostridia species) need to be better understood to apply HP as a universal spore inactivation technology.

CRediT authorship contribution statement
RH conceived the original idea. RH, AD, and AM planned the experiments and wrote the manuscript. RH and partially AD conducted the experiments. AM took care of funding acquisition, supervision, and project administration.

Funding
This work has been supported by the Swiss National Science Foundation SNF [Grant number: 31003A_182273].

Declaration of competing interest
None.