*2.10. Statistical Analysis*

Data analysis and the calculations of results was carried out using Microsoft Excel software, version 2013 (Microsoft Corporation, Redmond, WA, USA) and Minitab 16 Statistical Software (Minitab Inc., State College, PA, USA). Data is presented as arithmetic means ± standard deviations.

#### **3. Results and Discussion**

#### *3.1. In Vitro Antioxidant Capacity*

The in vitro antioxidant capacity of the flesh and seed are given in Table 1. In all three tests, the methanol extractions gave higher results than the ethanol extractions, which were higher than the water extractions apart from the FIC on the seed. The percent radical scavenging activity of the green plum flesh and seed was measured at a concentration of 1 mg/mL of extract powders, and the flesh gave 93.6 ± 0.6% in methanol, 92.5 ± 0.6% in ethanol and 92.7 ± 1.0% in water. The seed had similarly high levels at the same concentration of extract powder with methanol 93.2 ± 0.3%, ethanol 92.4 ± 0.7% and water 90.8 ± 1.3%. These levels are similar to that of the Kakadu plum, and higher than the Australian native wild lime, finger lime and almost twice that of the Davidson's plum in non-acidified extracts [27]. The radical scavenging activity is much higher than the chelating ability in both the flesh and seed at the same concentration, which could affect its application in food products.

**Table 1.** In vitro antioxidant capacity of green plum flesh and seed freeze dried powders.


*n* = 9; \* FIC % Chelating at concentration of 1 mg/mL of extract; N/A not available, mean values of each column with a different alphabet letter are significantly different (at *p* < 0.05).

The TPC of the unbound extraction was eight times higher than the Australian desert lime and lemon aspen, nearly three times higher than the quandong and riberry and one and a half times higher than both types of Davidson's plum studied by Konczak et al. [28]. It was lower than the Kakadu plum (approx. 158 g GAE/kg DW) but over twice the content of the blueberry control [28].

## *3.2. Antimicrobial Activity*

Table 2 shows the antimicrobial results of the disc diffusion assay and Figure 1 shows the flesh results.

**Table 2.** Antimicrobial activity by disc diffusion assay of the green plum flesh and seed, results as diameter of inhibition zone in mm using a 6 mm disc.


*n* = 3 All controls had no inhibition. NI: No inhibition observed. Mean values of each column with a different alphabet letter are significantly different (at *p* < 0.05).

**Figure 1.** Disc diffusion assay of flesh against top row, *S. aureus* and bottom row *E. coli*, from left to right extracts are methanol, ethanol and water, with controls in top right hand corner of plates. *n* = 3.

The flesh results show that it has much higher antimicrobial activity than the seed. The flesh inhibited the growth of bacteria in all extracts for both types of bacteria. The seed only gave a low amount of inhibition for the gram positive *S. aureus* and none for the *E. coli*. Aboriginal Australians use parts of *B. obovata* plant as bush medicine for its antimicrobial properties and this work adds to the growth inhibition found by Barr et al. [6] for the leaves and twigs. Figure 1 shows that in the water extractions the flesh both inhibited growth and also appears to have turned them a brown colour.

Further investigation of the antimicrobial properties was carried out using SEM imaging on the same bacteria types and the water extraction of the flesh. Figure 2 shows the SEM images of the controls and green plum flesh water extract treatment of *S. aureus* and *E. coli*.

The *S. aureus* control cells showed smooth round surfaces and membrane integrity. The treated *S. aureus* showed shape changes, with misshapen cells, broken open cells, swelling and changes to the cell structure. There was significant leakage of cytoplasmic inclusions and some cells showed this occurring. There were cells with changes to their morphology with deep wrinkles in them, as well as indentations and changes to their size from swelling. One had clearly broken open and the cell wall opened out, and others appeared to be broken bits of cell wall. Although there were a few cells that appeared more intact, their numbers were much lower than in the control. The *E. coli* control cells generally displayed smooth surfaces with occasional surface pitting and possible cell death from 31 h of growing. In the image from the treatment there is evidence of complete cell

collapse, pitting and breaking in the side wall of the cells, misshapen cells and leakage of cytoplasmic components. These indicate that there was an almost complete collapse of the cell structure, which was accompanied by cell lysis. The images show that the flesh extract in water does have antimicrobial activity. Cell disintegration may be what caused the dark colour in the flesh water extract disc diffusion plate.

**Figure 2.** Scanning electron microscopy images of (**a**) *S. aureus* control; (**b**) *S. aureus* with treatment from green plum flesh water extract; (**c**) *E. coli* control and (**d**) *E. coli* with treatment from green plum flesh water extract.

These images show changes that are similar to those shown by Zhang et al. [29] in their study on the antimicrobial activity of D-limonene. The deformation and distortion caused by the green plum extract indicates strong antimicrobial activity on the cellular integrity, similar to that shown by the combined treatment of D-limonene nanoemulsions with nisin [29].
