**3. Results**

#### *3.1. Morphological and Physiological Characteristics of the Bacterial Isolates*

Thirty bacterial isolates (10 isolates for each medium) were randomly isolated on three different selective media, medium (A), medium (B), and medium (C). Some of morphological and physiological characteristics of bacterial isolates were obtained. In addition, 10 bacterial isolates which were isolated on medium (A) were bacilli, Gram positive, spore positive, capsule positive and negative, and motile, and the indole test, catalase test, starch hydrolysis and casein hydrolysis were positive (Table S1, Supplementary Materials). From these characteristics, all isolates seemed to be *Bacillus* spp. [21]. Finally, 10 bacterial isolates which were isolated on medium (B) were diplococcic, Gram negative, non-spore, capsule positive, and motile and non-motile, and the indole test, catalase test, starch hydrolysis and casein hydrolysis were positive (Table S2, Supplementary Materials). From these characteristics, all isolates seemed to be *Azotobacter* spp. [21].

Bacterial isolates which were isolated on medium (C) were bacilli, cocci and short rods. Gram stain, spore stain, and capsule were positive and negative. Motile and none motile, indole test, catalase test, starch hydrolysis and casein hydrolysis were positive (Table S3, Supplementary Materials).

### *3.2. Screening the Best PHB-Producing Bacterial Isolates*

The results presented in Table 1 show that all tested isolates cultivated on medium (A) produced PHB. The CDW varied from 0.98 to 2.99 g/L. PHB production ranged from 0.10 to 0.62 g/L. PHB% ranged between 6.84 and 27.55%. Sugar utilization ranged from 14.32 g/L to 18.51 g/L. The values of the conversion coefficient ranged between 0.62% and 3.90%. Significant differences were observed between bacterial isolates in most cases. Generally, the best isolate which yielded the highest PHB production with reasonable sugar utilization was No. A6. Therefore, this isolate was chosen and used in the subsequent experiments. Isolates No. A1, A2 and A6 also produced more than 20% PHB.

In addition, all tested isolates grown on medium (B) also produced PHB (Table 1). The highest values of cell dry weight, PHB production, PHB%, sugar utilization and conversion coefficient were 4.00 g/L, 0.93 g/L, 28.51%, 20.93 g/L, and 6.17%, respectively. However, the lowest values of the above parameters were 2.56 g/L, 0.73 g/L, 19.50%, 15.07 g/L, and 2.86%, respectively. It was observed that there were significant differences between the bacterial isolates; thus, the best isolate with the highest value of PHB (0.93 g/L) was No. P3. This isolate was therefore chosen for identification based on molecular biology and used in the experiments on alternative carbon sources. PHB was produced from all isolates grown on medium (C) (Table 1). The CDW values varied from 0.79 to 3.1 g/L. PHB production ranged from 0.08 to 0.73 g/L, and PHB% ranged between 6.04 and 23.67%. Sugar utilization ranged from 6.09 g/L to 8.17. The values of the conversion coefficient

ranged between 1.00% and 8.93%. Generally, the best isolate which yielded the highest PHB production (0.73 g/L) with reasonable sugar utilization was No. W8. It was noted that there were significant differences between the bacterial isolates. Therefore, this isolate was chosen and used in the subsequent experiments.


**Table 1.** Screening of different bacterial isolates to produce PHB on different media.

C.D.W: cell dry weight; PHB: Poly-ß-hydroxybutyrate; A1–10: symbol for unknown bacterial isolate grown on medium (A) [15]; P1–10: symbol for unknown bacterial isolate grown on medium (B) [16]; W1–10: symbol for unknown bacterial isolate grown on medium (C) [17]. Means followed by different letters indicate significant differences between treatments according to Duncan's test (*p* < 0.05).

#### *3.3. Identification of the Best Bacterial Isolates Based on Molecular Method*

Isolates No. A6, P3 and W8 (which were the best isolates for PHB production in the previous experiments) were identified according to the Polymerase Chain Reaction (PCR) method by Sigma Company (Cairo, Egypt) [34]. Isolate No. A6 was identified by the

16S ribosomal RNA gene and the partial sequence Query1 GAGCTTGCTCTTATGAAGT-TAGCGGCGGACGGGTGAGTAACACGTGGGTAACCTGCCCAT 60.

The sequences of this isolate were accessed through a database (www.ncbi.nlm.nih. gov, accessed on 1 September 2021) using the accession number, and it was found to belong to the *Bacillus paramycoides* strain MCCC 1A04098 (Figure 1A). Isolate No. P3 was identified using the same method, by the partial sequence Query 1 AATACCCTGCAGTCTTGACGT-TACCGGCAGAATAAGCACCGGCTAACTTCGTGCCAGCAG 60. It was identified as the *Azotobacter salinestris* strain NBRC 102611 (Figure 1B). Isolate No. W8 was also identified using the same method, by the partial sequence Query 1 GAATAACTCAGGGAAACTTGT-GCTAATACCGAATGTGCCCTTTGGGGGAAAGATTTATCG 60. It was identified as *Brevundimonas naejangsanensis* strain BIO-TAS2-2 (Figure 1C).

#### *3.4. PHB Production from Cheap Alternative Carbon Sources*

The values of carbon and nitrogen were considered during preparation of the cultivation media, which already contained these sources. In dried whey, sugar beet molasses and date molasses, the carbon contents were 70%, 51%, and 37%, respectively, and the nitrogen contents were 15%, 13%, and 14.1%, respectively. Table 2 shows the effect on PHB production from *B. paramycoides* strain MCCC 1A04098 of replacing glucose in medium (A) with the three cheap alternative carbon sources (dried whey, treated sugar beet molasses, and treated date molasses). The addition of treated sugar beet molasses was the most economically suitable, as it increased PHB production (0.98 g/L) and the conversion coefficient (6.5%). Furthermore, the utilized sugar (15.11 g/L) was moderate and slightly lower than the control value (15.88 g/L). On the other hand, dried whey yielded the lowest PHB production (0.51 g/L), even lower than the control. The values of treated date molasses showed the same pattern as treated sugar beet molasses, except for sugar utilization which was the highest among all carbon sources (16.40 g/L). It was observed that there were significant differences between bacterial strains. In the present study, the maximum production of PHB and CDW by *B. paramycoides* was 0.98 g/L and 2.84 g/L from sugar beet molasses, respectively.

**Figure 1.** *Cont*.

**Figure 1.** Genotype tree of *Bacillus paramycoides* MCCC 1A04098 (**A**), *Azotobacter salinestris* NBRC 102611 (**B**) and *Brevundimonas naejangsanensis* BIO-TAS2-2 (**C**).


**Table 2.** Effect of cheap alternatives carbon sources on PHB production by different strains of *Bacillus paramycoides*, *Azotobacter salinestris* and *Brevundimonas naejangsanensis*.

Means followed by different letters indicate significant differences between treatments according to Duncan's test (*p* < 0.05).

Fructose was replaced with alternative carbon sources in medium (B) for PHB production using the *A. salinestris* strain NBRC 102611. Table 2 showed that the use of treated beet molasses yielded the highest PHB production (1.56 g/L) and highest conversion coefficient (23.92%), with moderate sugar utilization (6.52 g/L) in comparison with the control values. Dried whey yielded the lowest PHB production (0.66 g/L). It was observed that there were significant differences between the bacterial isolates.

Sucrose was replaced with alternative carbon sources in medium (C) for PHB production from the *Brevundimonas naejangsanensis* strain BIO-TAS2-2. Table 2 showed that the use of treated date molasses yielded the highest PHB production (1.5 g/L) and the highest conversion coefficient (9.35%), with moderate sugar utilization (16.03 g/L) in comparison with the control values. Dried whey yielded the lowest PHB production (0.64 g/L). It was observed that there were significant differences between the bacterial isolates.
