Figure 27. 2,4,6-Tripyridyl-s-triazine

In the original method two determinations were made: One with the sample and the other with water and TPTZ (control). The periodate consumed was calculated by difference. In the modification proposed in this work a fixed volume of reactive is titrated in the spectrophotometric cuvette with the sample previously diluted and the periodate concentration is obtained by a sample addition method (details in Material and Methods). With this method, an estimation of error can be made, and possible deviations in volume addition, instrumental response etc are minimized.
Another factor to be taken into account is the low solubility of PIA in pure water. Structural determinations of PIA required large volumes of bacterial culture (900 mL to several liters according the authors) (46, 62) which is not suitable for analytical purposes. In this work we use as extractant a solution of N,N-dimethylacetamide with 5% lithium chloride (68,69). This solution is one of the few capable of solubilize chitin (of similar structure of PIA but with ß(1-4) links of N-acetylglucosamine). With this solvent and the extraction method proposed significant differences in periodate consumed can be measured from initial samples of 50 mL.

Chromatographic calibration.
A separation with a Sephacryl S-300 (5,8 cm x 1.5 i.d.) was made. The column was previoully calibrated with dextran standards of MW 11600 and 48600 respectively.
Dextran polymers contains a1-6 glucose linked units branched with a1-3 in about 5% having vicinal hydroxyls susceptible to oxidation by NaIO₄. The fractions collected were oxidized for 12 h, and the periodate consumed was obtained by difference with a control. Since the long oxidation periods and the relatively high temperatures one could expect an overoxidation of the hydroxyl groups passing from aldehyde to carboxylic acid. This effect should be avoided in structural determinations, but in our case we can take advantage of it, because the greater periodate consumption results in a more sensitive determination. An additional benefit is obtained from the overoxidation. Since the oxidation was made in an unbuffered solution and deionized water was used in the chromatographic elution, the acidic groups formed should lower the pH in the fractions containing the polysaccharide. This is what is observed in practice; so that a simple pH inspection in each fraction informs us about of the position of the polysaccharide in the series.
We can see in Figure 28 that the minimum in periodate concentration (maximum polysaccharide concentration, fraction 7,) coincides with the minimum of pH value

Figure 28. Calibration with a standard dextran.Variation of pH and periodate concentration after the oxidation of the chromatographic fractions.

A similar correlation was observed with dextran standard Mw 48600 (Figure 29). The minimum appeared in fraction 4, about 7mL.

Figure 29. Calibration with dextran, Mw 48600.

The difference between the periodate consumed by each fraction and the control allowed us to draw a chromatographic profile

Figure 30. Calibration with dextran Mw 48600 (480 mg/L). Injection, 2mL

Extracts analysis
An extract of S. Epidermidis RP62 (PIA producer) with added glucose at 8 h (25) was analyzed. (see Material and Methods for the extraction procedure). The process of chromatographic separation and quantification was the same as for the dextran standards. The titration of the fractions collected yielded very precise results. Linear regression values for the fractions 4 & 7 , and the final [IO₄ ^-] are shown in Table 6 and plotted in Figure 31

Table 6. Titration data of the fractions 4 and 7

fraction	n	r	Veq ± tsa, µL	Cx ± t sb mM
T4	6	0.99996	393 ± 3	7.34± 0.08
T7	8	0.99997	430 ± 2	6.71± 0.05

a Student t , p = 0.05. s from [5] in Material and Methods
b s from Sm (see Material and Methods)

Figure 31. Titration curves of the fractions 4 and 7.The equivalence volume is obtained from the intercept to the volume axis

Figure 32. Chromatographic profile of the culture extract (see Material and methods)

The excellent linearity of the curves and the small errors allows us to establish significant differences between the fractions and make comparative studies with different samples as it is derscribed in the following paragraph.

Effect of allicin
Once established the analytical method,the next step was the study of PIA production in bacterial extracts of cultures containing subinhibitory dosis of allicin and compare the values of periodate consumed with that of control cultures without allicin. In the first experiment we prepared two 50 mL cultures of S epidermidis. One with allicin at 4 µg/mL , and the other without allicin. (see Material and Methods). In all assais PIA extraction was done from the biofilm adhered to the walls of the container and the bacteria precipitated after centrifugation. The extracts were passed trhrough a Sephacryl S-300 column proceeding as described before. The results are shown in Figure 32. The maximum of control appears in fraction 6. In allicin treated samples the maximum appears in fraction 7. This could be due to an interfering species or to a PIA with different molecular weight. In all cases, the periodate consumed by allicin added extracts is less than of the control samples. This is especially significant in the fraction 6 .

Figure 33. Periodate consumed by the fractions 5-9 after chromatographic separation of the extracts with allicin (4 µg/mL). and a control without allicin: Extractant, 1mL NNDMA

A second procedure for PIA extraction was carried out. The bacteria were grown in test tubes, (five tubes with allicin and another five without allicin) . The tubes were centrifuged an washed.The first tube was extracted with 2 mL of NNDMA/LiCl using glass microbeads and sitirred with a vortex. The extracts were passed to a second tube and vortexed again. The process was repeated with the other tubes and the final extract was centrifuged and passed trhrough a 0.45µ filter. This solution was passed by a Sephadex “desalting” column and then was chromatographed with a Sephadex S-300 (18x 1cm) . In Figure 34 appear the values of periodate consumed by the fractions.

Figure 34. Quantification of the chromatographic fractions T4-T9. Bacteria were grown in test tubes. Extraction with 2mL NNDMA

º PIA is the main agglutinating factor in the biofilm (for the strain studied ,RP 62) and in the cultures with allicin the biofilm production lowers in a considerable extent


º The analysis with periodate (a PIA oxidizer) yield lower values in allicin treated samples.


º The results of the “distance effect” point out that small amounts of an active agent (allicine or some allicin derivative ) produces a considerable -and long lasting- inhibition on biofilm formation, which suggest an action on the outer side of the bacteria. And that’s for two reasons. First, the enzimes embedded in the membrane-transmembrane- would be more accesible to allicin than those in the inner part of the cell . On the other hand enzymes in the cytoplasm are more protected by reducing molecules as NADH or glutation than those on the outer membrane. This could explain the persistent effect of biofilm inhibition we observe in growing bacteria (see figure 20)


º Finally, the hypothesis of an enzimatic inhinbition by blocking sulfhydril active centres is reinforced with the assays with iodoacetic acid and the biofilm inhibition with N-acetylcysteine (54)

Ica D

MVKPRQRQYP¹⁰ TVTSYLNIVR²⁰ ESLFITISGV ³⁰FWMYCIVVMI⁴⁰ VYIGTLINSQ⁵⁰
MESVITIRIA⁶⁰ LNVENTEIYK⁷⁰ LFGWMSLFVL⁸⁰ IIFIFFTFSL⁹⁰ AFQKYKKGRD¹⁰⁰ I