*3.3. Inhibition of mRNA Expression and Secretion of Pro-Inflammatory Cytokines by the Extracts*

Because the *Spirulina* extracts proved to be capable of suppressing the production of mediators that directly affect inflammation, Figures 4–7 show the effects of the two extracts on the mRNA expression of TNF-α, IL-6 and IL-1β, which are pro-inflammatory cytokines that directly affect the inflammation of brain nerve cells. These experiments also assess the degree to which the extracts suppress the actual production of these cytokines in mouse nerve cells.

Figure 4 shows RT-PCR products that reflect the expression of mRNA in the cells treated with only LPS, cells treated with both LPS and extracts, and the control with no not treatment. As shown in Figure 4a, the amount of TNF-α increased sharply in the mouse microglial cells treated with LPS alone compared to the control without any treatment. In contrast, when LPS and *Spirulina* extracts were administered together, TNF-α expression decreased in a concentration-dependent manner. Similar to the effects on NO and PGE2 production, shown in Figures 2 and 3, the ultrasonic extract had greater inhibitory effects on expression than the extract obtained through the general extraction process. To quantitatively compare the electrophoretic bands, shown in Figure 4a, the sizes of individual bands were normalized to the beta-actin band, which is a house-keeping gene, using a program that quantitatively compares the band sizes, as shown in Figure 4b. Similar to the pattern in Figure 4a, mRNA expression relative to beta-actin decreased as the concentration of the extracts administered increased. Additionally, the ultrasonic extract had much greater inhibitory activity than the general extract, as shown in Figure 4a. In particular, the difference in TNF-α expression inhibition between the two extracts was larger higher extract concentrations of 0.1 mg/mL than at lower concentrations,

possibly, due to higher amounts of chlorophylls in the extracts. Therefore, this result indicates once again that extraction of bioactive substances sensitive to heat requires a low-temperature extraction process and that the UE is the most efficient extraction process in such cases.

**Figure 4.** Down-regulation of mRNA expression of TNF-a from LPS-induced BV-2 cells (**a**) and the relative ratio of the gene expression by normalizing with beta-actin as a house-keeping gene (**b**) by the treatment of various concentrations of the *Spirulina* extracts along with the untreated control. EE, 70% ethanol extraction at 80 ◦C for 12 h, UE, ultrasonic pretreatment with 70% ethanol at 40 kHz and room temperature for 8 h, and further extraction at 65 ◦C for 4 h. Values are presented as means ±SD; \* *p* < 0.05 and \*\* *p* < 0.01 compared with the LPS group.

**Figure 5.** The secretion of TNF-a from LPS-induced BV-2 cells by the treatment of various concentrations of the *Spirulina* extracts. EE, 70% ethanol extraction at 80 ◦C for 12 h, UE, ultrasonic pretreatment with 70% ethanol at 40 kHz and room temperature for 8 h, and further extraction at 65 ◦C for 4 h. Values are presented as means ±SD; \* *p* < 0.05 and \*\* *p* < 0.01 compared with the LPS group.

As shown in Figure 5, the secretion of TNF-α from BV-2 cells was observed to confirm whether down-regulation of mRNA expression actually reflects the suppression of TNF-α secretion in mouse central nervous system (CNS) nerve cells. Similar to the pattern of the transcription of mRNA in Figure 4, the secretion of TNF-α was also decreased as a concentration-dependent manner with the greatest reductions at high concentrations. Unusually, although the number of general extracts produced concentration-dependent reductions, the reductions were relatively smaller compared to the ultrasonic extract. Although mRNA expression suppression, as shown in Figure 4, appeared to be concentration-dependent for both extractions, the degree of TNF-α suppression was lower in the general extract treatment. This means that mRNA expression is more sensitive to the useful components in the extract, but dramatic effects occur only when the concentration of useful substances is high enough in the following transcription stage. Therefore, we hypothesize that to anticipate intracellular inhibition or enhance functions using natural products along with gene expression, the natural products should exist at least at the critical concentration necessary to affect the production of the target substance. Figures 6 and 7 show the quantification of IL-6 and IL-1β gene expression, which was determined using the Image J program with the IL-6 and IL-1β RT-PCR product bands, as shown in Figure 4a, to more easily compare the suppression levels of these genes than that allowed by the pictures of electrophoresis bands themselves. Figure 6a shows the quantitative comparison of the RT-PCR-amplified bands reflecting mRNA levels of the target and housekeeping genes after treatment with LPS and the two extracts at different concentrations. Figure 7a shows the comparison of IL-1β mRNA gene expression levels.

**Figure 6.** Relative ratios of mRNA expression of IL-6 by normalizing with beta-actin as a house-keeping gene (**a**) and secretion of IL-6 (**b**) from LPS-induced BV-2 cells by the treatment of various concentrations of the *Spirulina* extracts. EE, 70% ethanol extraction at 80 ◦C for 12 h, UE, ultrasonic pretreatment with 70% ethanol at 40 kHz and room temperature for 8 h, and further extraction at 65 ◦C for 4 h. Values are presented as means ±SD; \* *p* < 0.05 and \*\* *p* < 0.01 compared with the LPS group.

**Figure 7.** Relative ratios of mRNA expression of IL-1beta by normalizing with beta-actin as a house-keeping gene (**a**) and secretion of IL-6 (**b**) from LPS-induced BV-2 cells by the treatment of various concentrations of the *Spirulina* extracts. EE, 70% ethanol extraction at 80 ◦C for 12 h, UE, ultrasonic pretreatment with 70% ethanol at 40 kHz and room temperature for 8 h, and further extraction at 65 ◦C for 4 h. Values are presented as means ±SD; \* *p* < 0.05 and \*\* *p* < 0.01 compared with the LPS group.

These two graphs do not show the electrophoretic bands, as shown in Figure 4a, to avoid redundancy, as the graphs in Figures 6a and 7a show the quantitative comparison of the bands obtained by electrophoresis. Examination of IL-6 and IL-1β gene expression after extract administration in Figures 6a and 7a revealed that the degree of IL-1β inhibition is generally higher than that of IL-6 for both extracts, although they are similar. This suggests that *Spirulina* extract may act more selectively on IL-1β than on IL-6 transcript levels. According to this difference of IL-6 and IL-1β protein secretion, shown in Figures 6b and 7b, the amount of IL-6 secreted, which was 680.5 pg/mL when 0.01 (mg/mL) of the ultrasonic extract was administered, decreased by 29.4% to 480.3 pg/mL when 0.1 mg/mL was administered. The amount of IL-1β secretion decreased by approximately 47% (from 94.5 pg/mL to 48.2 pg/mL), indicating that the extract from UE is more effective at inhibiting IL-1β and TNF-α production among the pro-inflammatory cytokines. Moreover, similar to the trend for the suppression of TNF-α production, the UE showed higher inhibitory effects than the EE from conventional hot ethanol extraction processes, demonstrating for the first time that ultrasonic extracts are generally more effective at inhibiting inflammation of mouse nerve cells than the extract from the hot extraction process. Although the inhibitory activity was generally lower than that of the ultrasonic extracts, the *Spirulina* extract from conventional 70% ethanol extraction also inhibited the secretion of pro-inflammatory cytokines from mouse nerve cells to levels similar to or higher than those from other natural products [5,9,15,16].
