The milk samples were collected from Errer valley, Babile District of eastern Ethiopia, about 30 km from Harar city. Errer valley is located at 9˚14'N latitude and 42˚14'E longitude at an altitude of 1300 - 1600 m.a.s.l. The milk was collected in clean and sterilized plastic containers from four households of five lactating camels at different parities and stage of lactations. The samples were pooled and packed under icebox and transported to Haramaya University Dairy Technology Laboratory. Milk samples were collected three times for each experiment and analysis was done in duplicates. The milk samples to be preserved with LPs were activated within 2 hrs of collection, as the indigenous antimicrobial activity in the freshly drawn milk is usually used up within 2 - 3 hrs due to suboptimal levels of the thiocyanate ion and hydrogen peroxide in the milk according to Codex Alimentarius Commission.

The thiocyanate concentration naturally present in camel milk samples was determined by spectrophotometer (3605, Jenway) at an absorbance of 460 nm, after deproteinisation of the milk with trichloroacetic acid (TCA) (Sigma-Aldrich T6399) and addition of ferric nitrate to form a ferric complex (orange to orange-red). Four millilitres (4.0 ml) of milk was mixed with 2.0 ml of 20% TCA solution. The mixture was mixed well and allowed to stand for 30 minutes. The solution was filtered through a filter paper (Whatman No. 40). One and half (1.5 ml) of the clear filtrate was then mixed with 1.5 ml of the ferric nitrate reagent and the absorbance was measured at 460 nm using spectrophotometer. The thiocyanate concentration was calculated from a standard curve prepared using known concentrations of sodium thiocyanate (Alfa Aesar 33388) [13].

An optimization experiment for the level of exogenous H₂O₂ to activate LPs was conducted using TBC and titratable acidity, considering 30 ppm to be used as a positive control in this experiment. Weissella confusa 22282 was selected for activation of the LPs due to its better H₂O₂ production in MRS broth and reduction in acidification rate in pasteurized camel milk. Mother culture of the strain was prepared by inoculating the strain into MRS broth for 72 hours at 30˚C and activating culture was prepared by growing the culture from the MRS broth in skim milk at a rate of 100 mg/L at 22˚C for 16 hours [21]. The culture was inoculated into raw camel milk at rate of 1% and shelf life of the milk samples was evaluated through the pH (acidification curves), TA, TBC and CC at storage times of 0, 6, 12, 18, 24 and 48 hours. Raw milk with no addition and 30 ppm H₂O₂ were used as negative and positive controls respectively (Table 1).

The TBC was done by pour plate method using standard plate count (SPC) agar (Sigma-Aldrich 70152). Plates with colonies ranging from 30 - 300 were counted and expressed as colony forming units per millilitre (cfu/ml) according to IDF [22]. The titratable acidity was determined by titrating the milk sample with 0.1 N NaOH (Himedia MB09) using a phenolphthalein indicator to an end-point of faint pink colour [23]. Coliform count was done by pour plate method using violet red bile agar (VRBA) (Sigma-Aldrich 70188). Plates with 15 to 150 cfu/mL were used for determining total coliform counts [22].

The activity of lactoperoxidase in the milk and the effect of activated LPs on Escherichia coli were evaluated by growing the bacteria (E. coli ATCC 25922) in pasteurized and boiled camel milk samples. Weissela confusa 22282 (1%, V/V) and exogenous H₂O₂ (30 ppm) were used as a source of H₂O₂ to activate the LPs in both pasteurized and boiled milk samples (Table 2). Changes in E. coli population

were evaluated at storage times of 0, 6, 12, 18, 24 and 48 hours at room temperature.

Temperatures of 63˚C/30 minutes and 80˚C/15 seconds were used for pasteurizing and boiling the milk samples, respectively. The E. coli (ATCC 25922) culture were obtained from Haramaya University Pathology Laboratory and maintained on MacConkey agar plates at 32˚C for 24 hours. Working cultures was prepared by transferring a single colony of E. coli from MacConkey agar into sterile Tryptone Soy Broth (TSB) (Sisco 24392 (TM 018)) to incubate for 24 hours at 37˚C. Contaminating inoculum of the E. coli was prepared by transferring 0.5 ml of the working culture into 100 ml sterile TSB for 24 hours at 37˚C [21]. The contaminating culture was then aseptically added to all the test samples at a rate 0.25% [24]. The respective test samples were then inoculated with 1% (V/V) W. confusa 22282 cultures and 30 ppm H₂O₂ in both pasteurized and boiled milk samples and left at room temperature for the period of 48 hours (Table 2). Cultures from the milk samples were plated on MacConkey agar and incubated at 32˚C for 24 hours [25].

General linear model (GLM) procedure of SAS version 9.0 was employed to determine the significance between treatment means at a particular storage period. Mean separations were done using least significant difference (LSD) for variables whose F values were significantly different. Significant differences were calculated at 5% significance level. Descriptive statistics was used to calculate means and standard deviations of chemical compositions and amount of H₂O₂ produced.

The natural thiocyanate content in the present study was 22.34 ± 5.11 mg/L that falls within the ranges reported earlier (9.74 to 32.9 mg/L) [26] and (9.7 to 36.4 mg/L) [27] and higher than the level found (6.04 mg/L) [28] from Erer valley of eastern Ethiopia. Natural thiocyanate content of 7.38 was found from cow milk in Kombolchadistrict, eastern Hararghe of Ethiopia [29].

The higher thiocyanate content in the present study compared to the finding from the same area [28] might be due to differences in analytical measurements, feeding system and environmental conditions. Several factors could affect the thiocyanate concentration of milk such as age of the animal, health of the animal, species of animal, breed, lactation stage and nutritional condition among which the kind of feed supplied plays a major role [30]. Thiocyanate content might also vary among season of the year where the level in summer was higher than the thiocyanate concentration in winter [31]. Camels in the Erer valley of Babile area spend during the day outdoors browsing different types of plants including herbaceous plants, shrubs, shoots, cacti and different types of acacia trees. Acacia trees and shrubs are expected to have high contents of cyanogenic glycosides which are a precursor of thiocyanate in plants [32].

The SCN value in the current study is higher than the concentration (15 ppm) required for the activation of the LPs according to Codex Alimentarius Commission [33]. The present finding is therefore showed that LPs in camel milk can be activated by natural SCN content in the milk with the addition of only desired amount of H₂O₂ [27].

The result of the acidification for the H₂O₂ producing strains of LAB indicated that W. confusa 22282 and S. infantarius 22279 are slow acidifying strains while L. lactis 22333 showed fast acidification in pasteurized camel milk at room temperature. L. lactis 22333 attained a pH of 4.3 at around 24 hours of storage while the pH in the W. confusa 22282 and S. infantarius 22279 treated samples shown no change (pH > 6.4) until the end of the storage period. The control treatment (pasteurized milk without LAB) started to drop at about 32 hours of storage and become below pH 6.4 at the end of the storage (Figure 1). This indicated that camel milk treated with W. confusa 22282 and S. infantarius 22279 had better stability compared to the control samples.

Similarly, the result of the H₂O₂ quantification by titration showed that W. confusa 22282 produced significantly (P < 0.05) the highest amount of H₂O₂ (302.10 ± 20.55) in the MRS broth. Lactococcus lactis 22333 produced significantly (P < 0.05) the least amount of H₂O₂ (84.37 ± 24.53) in MRS broth (Table 3). All the strains except L. lactis 22333 were observed to grow yielding a deep blue colour around the colonies on PB agar due to reaction of H₂O₂ to hexacyanoferrate (III) and iron (III) [34]. The overall results of H₂O₂ indicated that W. confusa 22282 was the promising strain to use for activation of the antimicrobial system in raw camel milk.

The production of H₂O₂ by LAB in MRS broth in the current study is higher than the results reported by earlier researchers [35] [36]. Hydrogen per oxide yield of 0.4279 mg/L in MRS broth from L. Lactis [35], and Leuconostoc mesenteroides produced higher quantity of H₂O₂ (24 mg/L) in normal MRS broth at 30˚C for 48 hours of incubation period [36]. However, higher concentration of 350 mg/L of H₂O₂ from L. lactis subsp. lactis suspended in 0.5% (w/v) glucose plus 0.5% (w/v) lactate (pH 7.0) and incubated for 5 hours at 37˚C under aeration was also reported [37].

The activation of LPs by W. confusa 22282 and exogenous H₂O₂ in raw camel milk significantly (P < 0.05) reduced the rate of acidification, TBC and CC. The values of pH (Figure 2) and lactic acid percentage (Table 4) in the activated samples for 18 hours of storage were within the acceptable level of pH and lactic acid of fresh camel milk, respectively. This could be due to retarded microbial growth as a result of antimicrobial properties of the LPs in activated samples. Decrease in lactic acid production was observed in LPs activated camel milk samples [27] [28] and other finding reported that there is less production of lactic acid in the LPs activated milk samples compared to the control sample because of the inhibitory nature of LPs [38].

The TBC and CC were significantly (P < 0.05) reduced in the treated samples compared to the non-activated control sample at 12, 18, 24 and 48 hours of storage. The TBC count in the LAB activated and H₂O₂ activated LPs samples were decreased by 1.61 and 1.66 log units, respectively compared to the non-activated sample at 18 hours of storage. On the other hand, TBC in the exogenous H₂O₂ activated LPs decreased below the initial value for up to 12 hours and then slightly increased toward the end of storage period. The LAB activated LPs sample shown nearly constant rate in bacterial reduction throughout the storage period compared to the H₂O₂ activated milk indicating that LAB can constantly produce H₂O₂ and other inhibitory components which have antagonistic effect on the microbial population. This finding can be supported by other findings [39] that the activation of the LPs may induce a longer-lasting bacteriostatic

T₁ = Raw camel milk (without preservative), T₂ = Raw camel milk + W. confusa 22282 culture, T₃ = Raw camel milk + H₂O₂. Means bearing different superscript letters within the same column are significantly different (P < 0.05).

effect due to the presence of higher levels of other indigenous antimicrobial components including the H₂O₂. It was reported that activation of LPs in camel milk decreased the multiplication of total bacteria for more than 12 hours of storage [40]. Other reports indicated that the activated LPs in milk had bacteriostatic effect against a mixed raw milk flora dominated by mesophilic bacteria [14].

On the other hand, W. confusa 22282 activated LPs retarded level of CC by 1.09 log units as compared to the non-activated at 18 hours of storage showing that the activated LPs exhibited antimicrobial effect on the coliform microorganisms in raw camel milk. This finding is in agreement with the result of other experiments [24] who reported reduction of CC by 1.14 log units in Saanen goats’ milk activated with LPs compared to the non-activated milk. No significant difference (P > 0.05) was observed in CC between the W. confusa 22282 and exogenous H₂O₂ activated LPs samples throughout the storage period which indicated that the treatments have comparable effect on the growth of coliforms in camel milk.

The activation of LPs by W. confusa 22282 and exogenous H₂O₂ in pasteurized camel milk remarkably reduced the growth rate of E. coli population compared their respective population in boiled milk throughout the storage period (Figure 3). This result indicated that the role of LP enzyme in boiled milk might be impaired in catalyzing the oxidation of thiocyanate by H₂O₂. The LPs improves keeping quality of milk pasteurized at 72˚C/15 s compared to milk heated to 80˚C/15 seconds [41]. Similar findings [42] was reported that activation of LPs extended the keeping quality of pasteurized milk (72˚C/15s) inoculated with Pseudomonas aeruginosa, S. aureus and S. thermophilus while the milk heated at 80˚C/5s and activated with the LPs had no effect on growth of these organisms. Moreover, this finding is supported by recommendation by FAO/WHO [8] that heating milk for 15 seconds at 80˚C, completely inactivates enzymatic activity of the milk might be because of the destruction of indigenous inhibitory components by the higher temperature. Indigenous enzymes are heat-labile and can easily be destroyed under the boiling conditions [43].

Similarly, the rate of E. coli growth in the LAB and exogenous H₂O₂ activated LPs in pasteurized camel milk was considerably lower than the rate in non-activated pasteurized milk (Figure 3). Reduction in microbial spoilage of LPs activated

pasteurized cow milk inoculated with E. Coli compared to the non-activated one was also reported [44]. An inhibition of the growth of E. coli for 24 hours [45] and a nearly total inhibition of E. coli O157:H7 [21] was also reported in cow milk and in commercial fermented milk and traditional Madila activated with LPs, respectively.

No remarkable difference was observed in retarding E. coli population between the W. confusa 22282 activated and H₂O₂ activated LPs in the pasteurized camel milk except a better retarding effect in the LAB activated samples toward the end of the storage period, which might be due to continuous production of H₂O₂ in the milk (Figure 3). The W. confusa 22282 treated boiled milk had shown slight reduction compared to the untreated boiled milk samples might be because of the effects of residual LP in the milk [42] or due to production of metabolites by the LAB [46]. Metabolites produced by LAB were reported to affect the growth of E. coli, S. aureus, Salmonella spp. in food items [47]. On the other hand, the H₂O₂ treated boiled milk shown remarkable reduction in E. coli counts compared to the untreated boiled samples. This clearly indicated that H₂O₂ had antimicrobial effect by itself as supported by other finding [48].