**2. Results and Discussion**

The use of (*R*)-(-)-1,2-isopropylideneglycerol as an acceptor in the coupling step with the trichloroacetamidate sugar gave high stereospecificity for the synthesis of Sulfavant R (**2**) [12]. However, a similar reaction with (*S*)-(+)-1,2-isopropylideneglycerol led to a low diastereoselective yield because of opening and closure of the acetonide ring under acidic conditions due to the boron trifluoride catalyst. As described in Scheme 1, formation of a complex between boron trifluoride and the acetonide oxygen atoms can start a fast and reversible process of racemization at C-2 of glycerol. The rearrangemen<sup>t</sup> rate was comparable to formation of the glycosidic bond, thus resulting in 20% epimerization of the glycosylated product. In order to overcome this issue, we prepared an alternative acceptor for the glycosylation reaction (Scheme 2).

*S O*GLVWHDUR\OJO\FHURO

**Scheme 2.** Synthesis of the acceptor (*S*)-1,2-*O*-distearoylglycerol.

Starting from the usual *S*-1,2-isopropylideneglycerol, the first step was the protection of the primary hydroxy group by benzylation with benzyl bromide and sodium hydride. Treatment with Dowex H<sup>+</sup> led to the diol derivative (**5**) from which the distearoyl-benzyl intermediate (**6**) was obtained by acylation with stearic acid and *N,N'*-dicyclohexylcarbodiimide. Without affecting the chirality of the stereocenter at C-2 of glycerol, debenzylation by hydrogenolysis on palladium gave the (*S*)-1,2-*O*-distearoylglycerol acceptor with 35% overall yield. The new acceptor was immediately coupled with the peracetylated glucosyl-trichloroacetimidate donor without loss of stereospecificity. As shown in Scheme 3, after deacetylation of the sugar moiety by hydrazinolysis, the synthesis of the sulfolipid followed the same sequence of reactions previously reported to give pure Sulfavant S [12] (Figure 1).

**Figure 1.** 1H-NMR of pure **3** in CDCl3:CD3OD 1:1 at 600MHz.

Enantiopure Sulfavant S (**3**) was tested for the activation of hDCs derived from blood monocytes [13]. DCs are the most efficient antigen-presenting cells (APCs) [16–22], often called "nature's adjuvant" [23] for their ability to induce activation and specific expansion of CD4+ helper T (Th) and CD8+ cytotoxic T (CTL) lymphocytes. The search for substances able to activate and prepare DCs to face pathogens represents a key tool in the development of new molecular adjuvants for vaccines against tumors or infections [24–26]. The effect of compound **3** on hDC maturation was measured by upregulation of CD83, an integral membrane protein belonging to the immunoglobulin superfamily and selectively expressed on mature DCs [27]. The activity of pure Sulfavant S (**3**) was similar to that of Sulfavant R (**2**) [12], with maximum CD83 expression at nanomolar concentration (Figure 2).

**Figure 2.** Percentage of CD83+ cells stimulated by pure Sulfavant S (**3**); asterisks indicate significant differences from untreated cells; \* *p* < 0.5, \*\*\*\* *p* < 0.0001.

In addition, as previously observed for compounds **1** and **2**, Sulfavant S (**3**) gave a bell-shaped dose activity curve that is common to other lipophilic drugs and reflects the formation of aggregates in the aqueous media [28–32]. In confirmation, Dynamic Light Scattering (DLS) measurements showed that pure **3** formed very small colloidal particles with a hydrodynamic radius of about 50 nm. The size of these particles was very similar to those we have previously observed for the epimer Sulfavant R (**2**) [12], thus proving that pure isomers have very similar biological and chemo-physical properties. Notably, both compounds were 1000-fold more active than Sulfavant A (**1**), which is composed of a 1.3:1 mixture of **2** and **3**.

In order to test the effect of the aggregation on the immunological response and to test our hypothesis on the reduction of the activity due to mixing of the two active epimers, we analyzed the effect of different combinations of Sulfavant R and Sulfavant S on hDC maturation.

Sulfavant R (**2**) was slightly more active than Sulfavant S (**3**), but their mixtures were always less effective in triggering CD83 expression than the pure molecules (Figure 3A). The addition of the *S* epimer (**3**) to the *R* epimer (**2**) determined a linear decrease of the activity in the range from 100% to 40% of *R.* Further additions revert the response (20% and 0% of *R*) as expected for the formation of mixture where the *S* epimer became progressively predominant. Furthermore, in complete agreemen<sup>t</sup> with the response elicited by **1**, the immunomodulatory activity increased to micromolar concentrations by mixture of **2** and **3** with a 1.3:1 ratio, which is identical to the composition of Sulfavant A (**1**) (Figure 3B). On the whole, these experiments proved that the mixing ratio can significantly affect the activity of amphiphilic compounds. In the case of Sulfavants, there is an incredibly large difference between potency of pure products and their mixture, thus determining erroneous evaluation of the therapeutic potential.

**Figure 3.** Effect of combination of **2** and **3** on hDC maturation. (**A**) CD83+ DC cells triggered by different combinations of **2** and **3** at an overall concentration of 0.1 μM; (**B**) CD83+ DC cells triggered by 1.3:1 mixture of **2** and **3** at concentrations from 1 nM to 100 μM; asterisks indicate significant differences from untreated cells; \*\*\* *p* < 0.001, \*\*\*\* *p* < 0.0001.

In order to test the role of aggregation on changes of the biological activity of **1–3**, we characterized the suspension of pure products and their mixture in water by Dynamic Light Scattering (DLS) (Figure 4).

**Figure 4.** Effect of combination of **2** and **3** on the hydrodynamic radius evaluated by means of Dynamic Light Scattering.

DLS measures showed a constant increase of the aggregate size moving from Sulfavant R (**2**) (about 50 nm) to the 1:1 mixture of Sulfavant R (**2**) and Sulfavant S (**3**) (about 170 nm). In analogy with the effect on the biological activity, further additions of compound **3** shrank the particle up to 50 nm of the pure Sulfavant S (**3**). This surprising behavior was linked to the different way that Sulfavants self-organize in aqueous solution [12]. The evidence that almost identical bioactive molecules (like epimers) could dramatically lose the biological activity as a result of their mixing represents a new paradigm to be evaluated in the study of bioactive lipophilic substances. We have previously suggested that the stability of the final aggregates has a relevant role in influencing the effective bioavailability of free active molecules in equilibrium with self-aggregates [12]. The current results obtained with a mixture of pure Sulfavant R (**2**) and Sulfavant S (**3**) prove, for the first time, that supramolecular organization of mixture of active epimers in aqueous solutions can bias evaluation of their biological and pharmacological potential.
