*3.2. Optimization of "Heat Treatment"*

Subsequently, the second step in the isolation process development was to optimize temperature and time of isolation; together, this process can be called "heat treatment". Within the first experiment focused on SDS concentration, "heat treatment" was represented by the application of a temperature of 70 ◦C for 2 h, but despite quite high yields, using different temperature can be more effective, or lowering the temperature could be more economically suitable, eventually. Based on previous results, time of exposition was set at

2 h, strain *H. halophila* was exposed to temperature 50 ◦C, 70 ◦C (as control—for comparison see Table 1), 90 ◦C and combination of cooling-down at 4 ◦C and following heating-up at 70 ◦C. The optimization process for *S. thermodepolymerans* included a little different strategy, when, in contrast to the halophilic strain, a lower temperature of 30 ◦C was also tested (we hypothesized that even low temperature could damage cell envelope of thermophilic bacterium). Other tested temperatures were 50 ◦C, 70 ◦C, 90 ◦C (as control—for comparison see Table 1) and also a combination of cooling-down at 4 ◦C and following heating-up at 70 ◦C as same as for *H. halophila*. The results of the experiments are listed in Table 2.

**Table 2.** Effect of different "heat treatments" on PHA purity and yield during isolation in 5 g/L SDS solution.


PHA content in the biomass: *H. halophila* (82.27 ± 0.09) wt.%, *S. thermodepolymerans* (69.57 ± 0.79) wt.%.

Based on the results presented in Table 2, the most successful "heat treatment" for isolation of PHA from the halophilic strain considering the purity of isolated material was a temperature of 90 ◦C for 2 h, when the purity was higher than 99 wt.%. Despite a little lower purity after isolation at 70 ◦C, the yield of the material was higher than for the highest tested temperature 90 ◦C and also higher than for the combination of cooling-down and following heating-up. For subsequent isolation experiments, a treatment representing 2 h of heating at 70 ◦C was applied for strain *Halomonas halophila*.

The optimization process for *S. thermodepolymerans* included a little different strategy; in contrast to the halophilic strain, we tested a lower temperature (30 ◦C), but the effectivity of isolation was not too high (the purity was around 67 wt.% and the yield was 0.73). Based on these data, the most successful "heat treatment" was keeping the sample at 90 ◦C for 2 h; here, the purity exceeded 99 wt.%, hence, it was very similar to using a temperature of 70 ◦C for *H. halophila*. The combination of cooling followed by heating was not too successful for the thermophilic strain, whereas exposition to low temperature presumably led to the development of enhanced resistance to thermal treatment as was already reported in the literature [31]. This effect was described for mesophilic bacterial strains *E. coli* [32] and *C. necator* [33] after treatment at low temperature (0–5 ◦C). It was proved that enhanced resistance against elevated temperature was caused by the changes in the fatty acids composition in cell membranes [31]. Despite the fact that the effect was not described for thermophilic strains in detail, it is very likely that also in this case the low-temperature treatment resulted in active modulation of cell membrane lipids, which negatively impacted the susceptibility of the bacterial cells to hypotonic lysis. For our experiment, lower efficiency of combined "heat treatment" can be observed compared with keeping the cells at 90 ◦C for 2 h. To conclude, the most suitable "heat treatment" procedure for the thermophilic strain *S. thermodepolymerans* was keeping the isolation mixture at 90 ◦C for 2 h.

#### *3.3. Testing of Different Ratios SDS Solution: Biomass*

After optimization of SDS concentration and "heat treatment", the following step was focused on the influence of initial biomass amount on the purity of isolated material and the yield of the process. This is very important since PHA production is usually performed as a high cell density cultivation [34], therefore the robustness of the isolation process with respect to the amount of biomass employed per isolation batch is a very important factor determining the viability and feasibility of the process. Therefore, we tested different initial concentrations of biomass, while keeping the volume of SDS solution constant (5 g/L). Moreover, an increasing portion of biomass per isolation batch can also reduce the cost of the isolation process (lower SDS amount) and significantly decrease the environmental burden caused by the presence of SDS in wastewater.

Based on the results listed in Table 3, it is possible to state that the process is very robust regarding the initial biomass concentration; purities of material isolated from different biomass amounts were greatly similar. The same goes for almost all yields for both strains. Therefore, the developed isolation procedure can be advantageously applied on biomass obtained from high-cell-density cultivation (e.g., fed-batch cultivation).


**Table 3.** Effect of different biomass content on PHA purity and yield applied during 2 h lasting isolation in 5 g/L SDS solution at 70 ◦C for *H. halophila* and 90 ◦C for *S. thermodepolymerans*.

PHA content in the biomass: *H. halophila* (75.5 ± 0.1) wt.%, *S. thermodepolymerans* (68.5 ± 0.0) wt.%.

#### *3.4. Effect of Different Isolation Approaches on the Molecular Weight of the Polymer*

The following experiment was focused on the determination of the molecular weight of obtained PHA materials; the main goal was to evaluate the influence of the isolation procedure on the molecular weight of the polymer.

Isolation of polymer using SDS solution (5 g/L) did not manifest in the decrease of molecular weight of PHA as can be seen in Table 4. For comparison, PHA was isolated from dried biomass using a procedure involving 12 h lasting incubation in chloroform at 70 ◦C in a thermoblock. To eliminate differences from sample preparation, all analyzed PHA samples were subjected to the same treatment. Despite the slight differences, values of molecular weight were in principle very similar at least for polymers isolated from *S. thermodepolymerans*. Only a little decrease can be observed which could be caused by enhanced temperature 90 ◦C used for isolation with SDS. In the case of *H. halophila*, isolation with SDS led to enhancement of molecular weight up to 1.5-times. It should be pointed out that the Mw of the polymer produced by *H. halophila* is very high as compared to other PHA producers [8] including *S. thermodepolymerans*. It can be assumed that the higher molecular weight of polymer after SDS isolation treatment might be caused by enhanced solubility of longer chains in chloroform after elimination of other cells components. 12 h lasting isolation at 70 ◦C in chloroform is possibly more efficient for shorter polymer chains with lower molecular weight. PDI values were very low for all studied isolation setups, indicating the high uniformity of generated biopolyesters in terms of molecular mass distribution, which is a beneficial feature for further polymer processing.


**Table 4.** The molecular weight of isolated PHA polymer.

<sup>2</sup> PDI (polydispersity index) is defined as Mw/Mn. Mw: weight average molecular mass; Mn: number average molecular mass.

#### *3.5. Determination of Material Purity by Infrared Spectroscopy*

FTIR technique was applied on samples of dried biomass, PHA polymer isolated by optimized process for each strain (2 h lasting treatment at 70 ◦C for *H. halophila* (HH) or 90 ◦C for *S. thermodepolymerans* (ST) in solution with 5 g/L SDS), and also for commercial P(3HB) (Biomer) as a reference material. Result are shown in Figures 1 and 2.

**Figure 1.** FTIR spectra of original biomass of *H. halophila* (black), polymer isolated via optimized approach (red) and commercial P(3HB) sample (Biomer) (green); arrows show absorption bands for components of biomass (3300; 1640 and 1540 cm<sup>−</sup>1—proteins).

**Figure 2.** FTIR spectra of original biomass of *S. thermodepolymerans* (blue), polymer isolated via optimized approach (magenta) and commercial P(3HB) sample (Biomer) (green); arrows show absorption bands for components of biomass (3300; 1640 and 1540 cm<sup>−</sup>1—proteins).

To evaluate the purity of the polymer isolates from FTIR spectra, characteristic spectral signatures of proteins can be used. These can be found at about 3300 (referred to as amide A band), 1640 cm−<sup>1</sup> (referred to as amide I band) and 1540 cm−<sup>1</sup> (referred to as amide II band). All these characteristic bands are marked with arrows in the biomass spectra in Figures 1 and 2. As expected, the protein signal is completely absent in spectrum of the reference P(3HB) material. In the FTIR spectra of polymer isolates obtained from both tested strains, significant decrease of the protein signal, compared to the original biomass, indicates high effectivity of the polymer purification during the extraction process. As can be seen from the comparison of Figures 1 and 2, lower content of the residual biomass was found for the polymer isolated from *H. Halophila*. This is in good agreement with the results of GC-FID. Based on the gravimetric yield of the chromatographic analysis, we determined polymer purity (86.4 ± 1.4) wt.% for *S. thermodepolymerans* and (91.3 ± 1.5) wt.% for *H. halophila*.

### *3.6. Electron Microscopy Imaging*

Selected samples were analyzed using electron microscopy techniques, specifically, samples of microbial cells of *H. halophila* and *S. thermodepolymerans* and isolates of PHA from both strains.

Cryo-SEM image of *H. halophila* (Figure 3A) shows rod-shaped cells containing several granules of PHA. Most of the cells were fractured during the freeze-fracturing procedure and revealed the intracellular content of PHA. As previously described [35], PHA remain elastic even at temperatures of liquid nitrogen and can be observed being pulled out of the cells, showing "needle type" deformation (arrowhead in Figure 3) or as holes in cells (concave deformation) when the granule was pulled out of the cell completely. Cells of *S. thermodepolymerans* (Figure 3C), containing 1 or 2 granules of PHA also deformed by freezefracturing. In the image, it is possible to observe also fracture of the cell wall revealing the surface of the plasma membrane, marked with O. The image of isolated granules from *H. halophila* (Figure 3B) shows that the granules merged, but still remained elastic. PHA isolated from *S. thermodepolymerans* (Figure 3D), however, stayed separate after the extraction as individual oval granules, which were also affected by freeze-fracturing and show needle deformation.

**Figure 3.** Cryo-SEM image of: (**A**) *H. halophila* biomass before extraction of PHA granules, (**B**) PHA granules isolated from *H. halophila*, (**C**) *S. thermodepolymerans* biomass before extraction of PHA granules, (**D**) PHA granules isolated from *S. thermodepolymerans*, needle deformation of PHA granules marked with arrowhead, cells with fractured cell wall revealing the surface of plasma membrane marked with O, scale bar 2 μm.

TEM observation confirmed previous findings of cryo-SEM. *H. halophila* contained several smaller granules in their cells (Figure 4A), while *S. thermodepolymerans* contained 1–3 larger granules. Generally, it is described that larger PHA granules are beneficial for the recovery process [36], which makes *S. thermodepolymerans* of even higher interest for an eventual industrial-scale application. Images of isolated granules also confirm the shape observed in cryo-SEM: isolates from *H. halophile* merged into irregularly shaped granules, while granules from *S. thermodepolymerans* of regular oval shape stayed separate. In Figure 4B,D, it is possible to see shades surrounding the granules. Based on the results of purification, the shades could be caused by contamination originating from disrupted cellular fragments, or by PHA granule membrane remnants. It is also possible to observe a few intact cells (Figure 4B marked with arrow), thus confirming the results of the determination of purity of the isolates—for instance (91.3 ± 1.5) wt.% of PHA for *H. halophila* using standard optimized process.

**Figure 4.** TEM image of (**A**) *H. halophila* biomass before extraction of PHA granules, (**B**) PHA granules isolated from *H. halophila*, (**C**) *S. thermodepolymerans* biomass before extraction of PHA granules, (**D**) PHA granules isolated from *S. thermodepolymerans*, PHA granules marked with arrowhead, intact cells after extraction marked with arrow, scale bar 1 μm.

### *3.7. Wastewater Management—Precipitation of KDS from SDS after Isolation*

As same as other detergents, also SDS serves as a contaminant of wastewater with possible harmful effects on wastewater management. Therefore, we focused on the removal

of the irritant compound SDS from the supernatant after the isolation process. We have tested several approaches (data not shown), and the most successful procedure was simple and feasible precipitation of SDS in the form of poorly water-soluble potassium dodecylsulfate (KDS) in the presence of 2 M KCl. The efficiency of the SDS removal process was determined via measurement of SDS concentration using the protocol of Rupprech et al. (2015) [29]. Data are demonstrated in Table 5.

**Table 5.** Results of SDS determination before and after KDS precipitation.


<sup>3</sup> n.d.—no detectable.

Based on presented data, it seems that removal of SDS from a wastewater stream by simple addition of non-toxic and cheap KCl is a very effective process since after this treatment the residual concentration of SDS in the supernatant after isolation was not detectable for almost all the samples. The formation of KDS precipitate is a very quick and straightforward method to remove SDS used for isolation of PHA granules, since the formation of the precipitate can be observed immediately after the addition of KCl into supernatants after isolation as shown in Figure 5.

**Figure 5.** KDS precipitate using 2 M KCl after isolation with different SDS concentrations—from the left: blank, 1 g/L, 2.5 g/L, 5 g/L, 10 g/L: (**A**) before centrifugation; (**B**) after centrifugation.

A different strategy to get rid of the detergent after isolation was used by the team of Samorì et al. (2015), who isolated PHA polymer from bacterial biomass. For this purpose, they used ammonium laurate (most efficient at a concentration of 200 wt.%), which acts similarly to SDS in terms of microbial cell disruption and solubilization of hydrophobic contaminants. After separation of the polymer by centrifugation, the addition of CO2 into supernatant leads to a decrease of the pH-value, and switches the detergent into poorly water soluble neutral (protonated) lauric acid with the possibility of separation by centrifugation; resulting water phase containing ammonium hydrogen carbonate can serve as a nitrogen source for microorganisms [37]. Our approach relies on the application of the lower amount of inexpensive and abundant detergent SDS, which can be simply removed from wastewater by precipitation. It should be pointed out that unlike for ammonium laurate, the removal of the SDS from the supernatant in form of KCl occurs in a non-destructive way, and it is very likely that SDS could be regenerated for instance by ultradialysis in excess of Na<sup>+</sup> ions. Hence, our concept enables not only removal of detergent from wastewater, but also holds a promise of SDS recovery, which would be very beneficial with respect to economic aspects of the PHA isolation process. Nevertheless, the

possibility of SDS recovery from KDS precipitate deserves further investigation, which is out of the scope of this preliminary work.

### **4. Conclusions**

In this work, we developed a procedure for the isolation of PHA materials from extremophilic microbial cells. The method is based on the exposition of the bacterial cells naturally containing high intracellular concentrations of compatible solutes as part of their adaptation strategy to extreme hypotonic conditions induced by the diluted solution of SDS (optimal concentration is about 5 g/L of SDS) at elevated temperature. Our results indicate that such conditions lead to disruption of the cells and release of PHA granules. Moreover, SDS apart from its cell-disruptive function also solubilizes hydrophobic cell components which would otherwise get attached as contaminants to PHA materials. The procedure seems to be simple, robust, and feasible in industrail condtions. The purity of obtained materials (usually above 95%), as well as yields of the procedure (usually about 0.9), reach high values. If even higher purity of the material is needed, additional steps such as the washing of the materials with proper organic solvents or other agents can be involved. Furthermore, since leftovers of the detergent SDS in process wastewater might dramatically decrease the positive ecological features of the process, we also focused on the removal of SDS. We have developed a simple, cheap, and safe technique that is based on the precipitation of SDS in the presence of KCl. The resulting precipitate can be simply removed by decantation or centrifugation, hence, SDS does not contaminate the wastewater of the process. Moreover, there is also the possibility to regenerate SDS which would substantially improve the economic feasibility and overal sustainability of the process. To sum up, the developed strategy for PHA isolation is compatible with Next-Generation Industrial Biotechnology concepts, it is simple, cheap, robust, provides PHA materials in high purity and at high yields and is also ecologically friendly.

**Author Contributions:** Conceptualization, S.O.; investigation, I.N., X.K., K.M., P.S., M.K. (Michal Kalina) and V.K.; writing—original draft preparation, S.O., I.N. and M.K. (Martin Koller); writing—review and editing, X.K., P.S. and M.K. (Martin Koller); visualization, S.O., I.N. and K.M.; supervision, S.O.; project administration, I.N.; funding acquisition, I.N. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by Grant FCH-K-21-6952 implemented within the project Quality Internal Grants of BUT (KInG BUT), Reg. No. CZ.02.2.69/0.0/0.0/19\_073/0016948, which is financed from the OP RDE. Also CIISB research infrastructure project LM2018127 funded by MEYS CR is gratefully acknowledged for the financial support of the measurements at the CF CryoEM.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
