Automated Solution-Phase Syntheses of Alpha 1→2, 1→3 Type Rhamnans and Rhamnan Sulfate Fragments

Abstract

Rhamnan and rhamnan sulfate are naturally occurring carbohydrates that have important biological functions and possible therapeutic applications, but studies are limited to the microheterogeneous mixtures from natural sources. This work reports the first synthesis of any sulfated rhamnan fragments and successful automation of the process with a recently developed automated solution-phase approach using N-iodosuccinimide/trimethylsilyl triflate (NIS/TMSOTf) promotor and levulinoyl ester deprotection conditions. The automated solution-phase activation/deprotection approach was initially able to create alpha 1→2, 1→3 type rhamnan di- and trisaccharide in moderate yields. Once these targets were achieved, a process to use SO₃•pyridine complex in DMF for sulfation compatible with an automated solution-phase liquid handling system was developed and successfully applied to carbohydrate sulfation to create two rhamnan sulfate fragments with differing monosulfation patterns.

Graphical Abstract

1. Introduction

The automated syntheses of peptide and nucleic acid biopolymers have become commonplace within the organic chemistry field. The ability to efficiently and rapidly access these biopolymers has been instrumental in expanding our understanding of their biological functions. It has also allowed for the exploration of these molecules in non-natural applications such as biomaterials and drug candidates. [1, 2] Unfortunately, the use of automated technology to create other types of organic molecules is still in its infancy and presents unique challenges compared to peptides or nucleic acids. Specifically, the development of automated syntheses for carbohydrates has experienced difficulties from both a chemical and practical standpoint. The extensive number of conceivable linear and branched structures significantly contributes to the complexity of carbohydrate syntheses.[3] Many of these linkages are difficult to form and require non-commercially available building blocks for precise chemical control to obtain appropriate regio- and chemoselectivity. Also, once the desired carbohydrate is formed, purification and isolation of a single structure is a time-consuming affair.

Despite these practical and chemical challenges, several automated methods have been developed for the synthesis of carbohydrates. Many automated oligosaccharide syntheses have been modeled after solid-phase peptide synthesis, which uses a polymeric or gold-based support for chain extension.[4–7] Several solution-phase automated methods have also emerged alongside the solid-phase methods. These methods include the use of electrochemical glycosylation, chemoenzymatic strategies, and the use of fluorous tags to facilitate intermediate purification.[7–16] The fluorous-tagged strategy has been successful in providing a number desired oligosaccharide targets both manually and on an automated system with efficient usage of building block equivalents.[17–20] Most recently, this fluorous-tag-based approach was used to create an efficient automated process for the synthesis of an α−1→2 rhamnan di- and trisaccharide. The developed process focused on adapting a commonly used N-iodosuccinimide/trimethylsilyltriflate (NIS/TMSOTf) promoter system for thioglycoside activation for use on a solution-phase handling platform.[21] In addition, it was previously shown that it was possible to selectively deprotect the levulinoyl ester from the formed di- and trisaccharide rhamnan structures in the presence of glycosidic byproducts. This deprotection strategy eliminates the need for a purification event after glycosylation and improves the overall efficiency of this synthetic strategy.

With the success of the developed automation approach, we next wanted to see if we could extend the scope of this automated NIS/TMSOTf-promoted glycosylation and deprotection strategy to the creation of rhamnan fragments with alpha 1→2 and 1→3 linkage type patterns as well as site-selectively sulfated versions of these types of oligomers. Naturally occurring rhamnans contain both α−1→2 and α−1→3 linkages and have been found modified with sulfates at various positions on the rhamnose ring [22–25] (Figure 1); however, such sulfated rhamnan structures have not yet made by chemical or enzymatic synthesis. Rhamnan sulfate structures have been linked to a variety of possible therapeutic applications as anti-coagulant, antiviral, anti-obesity, and anti-tumor agents.[26–29] The unmodified rhamnan structures are frequently found in virulent strains of both Gram-positive and negative bacteria.[22, 23] It has also been shown that an α−1→3 linked rhamnan disaccharide on a multivalent scaffold reduces the binding activity of a horseshoe crab lectin to P. aeruginosa PA01, a virulent Gram-negative bacteria that contains rhamnan structures on its outer cellular membrane.[30] Based on these findings, smaller rhamnan fragments displayed on a multivalent scaffold could serve as diagnostic or therapeutic tool. In an effort to provide alternative, well-defined chemical structures for screening, a variety of synthetic di- and trisaccharides with varying linkages and sulfation patterns are needed.

Figure 1.

Rhamnan and rhamnan sulfate fragment structures.

2. Results and Discussion

2.1. Building Block Syntheses

Even with automation of the oligosaccharide assembly itself, half or more of the chemical steps remain in the syntheses of properly protected building blocks required for the various automation platforms. In order to create rhamnan libraries, building blocks amenable to the formation of both α 1→2 and 1→3 linkages are needed. The initial acceptor and donor for α 1→2 linkage syntheses were reported in a prior study using established methods.[21, 31] However, we envisioned obtaining all of the necessary building blocks for these rhamnan structures in a divergent manner from a common intermediate (1) found in the previously reported synthesis for donor 2 (Scheme 1). To that end, the initial acceptor 3 was easily obtained in two steps from donor 2 in high yield (86%). This approach eliminates the need to use a separate pathway as shown in prior work. The other donor 6 could also be effectively reached from the advanced diol intermediate 1 in the α 1→2 donor pathway (Scheme 1). In lieu of a selective benzylation at the 3-OH, a selective p-methoxybenzyl (PMB) protection with dibutyltin oxide was performed to allow for subsequent selective deprotection.

Scheme 1.

Divergent building block syntheses from common intermediate 1.

The protecting group initially chosen for the 2-OH position was a levulinoyl (Lev) ester to provide both neighboring group participation and ease of removal using the proven automated method. The protection of the 2-OH position with Lev was carried out in high yield (75%). However, selecting a protecting group to occupy the 3-OH position posed a greater difficulty, considering a selective deprotection would be needed to form the desired α 1→ 3 linkage in a subsequent step. Initial protection at the 3-OH position using a PMB ester proved the glycosylation conditions were too acidic, resulting in loss of the PMB altogether in a nonproductive manner. Further investigation using a fluorenylmethyloxycarbonyl (Fmoc) protecting group also proved problematic as the reaction was low yielding and difficult to purify. Finally, it was determined that a levulinoyl (Lev) ester was sufficient to protect the 3-OH position with a benzoyl (Bz) ester present at 2-OH (Scheme 1). The Bz protecting group provided neighboring group participation to obtain the desired alpha stereoselectivity, while also providing orthogonality to the Lev ester. In addition, the Bz ester reduces the amount of orthoester side product formation compared to use with an acetyl (Ac) group and should be easier to remove than a pivaloyl (Piv) ester. Benzoylation was performed in moderate yields (60%) from intermediate 4 and was followed by an oxidative cleavage that removed the PMB group in good yield (81%). The final Lev deprotection was completed as expected to provide the final donor 6 in moderate yield (64 %).

2.2. Solution-Phase Automation

With all the building blocks in hand, we first tested to see if the previously developed automated approach [21] could be employed without modification to expand the rhamnan fragment library. First, the solution-phase process was tested for its viability to form disaccharide 7. The automation protocol was performed with acceptor 3 and donor 6, a less activated donor than the original differentially protected donor 2. Only small reaction time modifications were made to the previously developed program for use on this glycosylation system. Analytical HPLC traces and mass spectroscopy revealed the presence of the desired disaccharide and upon isolation via preparative HPLC, the rhamnan disaccharide 7 was isolated in 62% overall yield with an average step yield of 79% (Scheme 2 and SI). The success of this run demonstrated that a modified system could be used to create a different rhamnan disaccharide with a less activated donor. Upon successful synthesis of the disaccharide, the next step involved formation of the trisaccharide 8 with an α 1→2, α 1→3 linkage pattern. The same automated program was utilized with acceptor 3, donor 6 (used in the first glycosylation), and donor 2 (used for the second glycosylation). Traditional flash chromatography was performed to isolate trisaccharide 8 in 23% overall yield with an average step yield of 69%.

Scheme 2.

Automated solution-phase syntheses of rhamnan fragments for alpha 1→2 and 1→3 linkage types.

2.3. Solution-Phase Automation Protocol for Carbohydrate Sulfation

Previous studies using solid- or solution-phase automation platforms have focused solely on unmodified rhamnan backbone fragments.[32, 33] However, no rhamnan sulfate fragments syntheses have been reported to date. Considering sulfation of biologically isolated polysaccharides typically occurs at the 2-OH and 3-OH positions, these sites were the main focus of our proposed modifications to rhamnan disaccharides. In order to access these derivatives, however, sulfation conditions amenable to a solution-phase handling platform needed to be developed. In the literature, oligosaccharide sulfation is generally carried out via addition of a sulfur trioxide (SO₃) complex in a nitrogen-containing basic solvent (pyridine or DMF).[34, 35] The most commonly reported conditions include SO₃•pyridine (Py) complex in pyridine, SO₃•trimethylamine (TME) or SO₃•triethylamine (TEA) complex in DMF at 50–70 °C.[17, 34, 36–40]

Using procedures reported in literature, both types of conditions were evaluated to assess the stability of the rhamnan disaccharide and physical properties of the reactions. Liquid-phase handling platforms require the reaction components to be in solution, in the case of the fluoroustag-based approached specifically at two different points in the automated process (starting reagents and final reaction mixture). Ideally, the starting materials/reagents would be stable in solution for an extended period of time at ambient temperature. This criteria is necessary to allow for the correct amount of compound to be transferred at the programmed point in the automation sequence. Once the reaction is complete, the final reaction mixture is removed from the reaction vessel by the liquid-handling system. The solvent is then evaporated from the final reaction mixture and can be dissolved in an alternative solvent to allow for facile transfer by the liquid handling system.

Both sulfation conditions (SO₃•Py complex in pyridine and SO₃•TME complex in DMF) were carried out manually on bench using the previously synthesized rhamnan disaccharide 9 (Scheme 3 and SI). It was observed that both sets of conditions started as suspensions that solubilized over time as the reaction was heated. With the considerable lack of solubility at ambient temperature observed, alternative sulfation conditions amendable to the solution-phase automated platform needed to be explored. Further investigation revealed that the automated solid-phase synthesis of keratan sulfate fragments utilized 0.5 M SO₃• Py in pyridine:DMF (1: 1) solution at 50 °C.[41] The 0.5 M SO₃•Py solution was prepared as reported, but was also observed to be a suspension at ambient temperature. While heating the SO₃•Py suspension solubilized the complex, a suspension reformed upon cooling. Once again, solubility issues hindered use with the solution-phase handling system. Additional exploration resulted in discovery of a less frequently used method involving SO₃•Py in DMF at 50 °C for oligosaccharide sulfation.[42] When implemented, the SO₃•Py complex was found to be soluble in DMF at ambient temperature. Subsequently, the compatible conditions were tested on the sulfate disaccharide intermediate 9. Fortunately, this sulfation solution did allow the formation of the desired sulfated rhamnan disaccharide, as confirmed by mass spectroscopy analysis.

Scheme 3.

Manual sulfation trial reactions.

A new automation protocol incorporating these compatible conditions was then developed. The automated sulfation protocol included a 4 h reaction at 60 °C, followed by a cooling macro to 25 °C, and finally a purification by the in-line FSPE (see Experimental Table S3 and S4). Unlike previous reaction protocols, no evaporation cycle was included due to the reversible nature of the reaction and FPSE purification method. Considering that dimethylformamide (DMF) is a commonly used loading solvent for FPSE purifications, the sulfation reaction mixture could be directly loaded on the column without the need to remove a non-FSPE compatible solvent. The sulfation protocol was appended to the end of the modified NIS/TMSOTf-promoted glycosylation and Lev deprotection process to efficiently synthesize the desired rhamnan sulfate disaccharides. This new protocol was initially utilized to create 10 from acceptor 3 and donor 2 (Scheme 4A). Sequential automation runs afforded the mono-sulfated disaccharide 10, which was confirmed via analytical HPLC and mass spectroscopy analysis (SI). Purification conditions were then developed to provide a pure compound (10) in 31% overall yield with an average step yield of 68%. The automation process was further extended to obtain disaccharide 11 from acceptor 3 and donor 6. Gratifyingly, the desired rhamnan sulfate disaccharide 11 was effectively isolated in 33% overall yield with an average step yield of 69% (Scheme 4B). No further process optimization was required to achieve useable quantities of the desired products using automated synthesis, a criteria required for automation of oligosaccharide synthesis to become as valuable as automated nucleic acid and peptide synthesies.

Scheme 4.

Automated solution-phase synthesis of monosulfated rhamnan disaccharides 10 and 11.

3. Conclusion

This work demonstrates the feasibility of using an automated solution-phase-based approach for the first synthesis of any rhamnan sulfate fragments. This work expands the target scope of the originally developed NIS/TMSOTf and Lev deprotection automated approach to provide access to several different rhamnan and rhamnan sulfate fragments. An unmodified rhamnan disaccharide as well as a trisaccharide with a α 1→2, α 1→3 linkage were synthesized in 62% and 23% overall yield, respectively, with only minor reaction time modifications from the programs originally used in the prior automation work. The prior protocol was extended through the development of the first automated carbohydrate sulfation protocol for a solution phase-based automation platform to thereby allow for the synthesis of two disaccharide rhamnan sulfate fragments. Importantly, extensive optimization was not required to This work should enable not only the production of chemically well-defined rhamnan and rhamnan sulfate structures for biological assay and tool development, but also the solution-phase-based automation of other sulfated carbohydrates.

4. Experimental Section

4.1. General Experimental

The chemicals used were purchased from Sigma-Aldrich, Oakwood Chemical, TCIAmerica, Fluorous Technologies, Inc, or Alfa Aesar. These chemicals were used as they were packaged unless otherwise stated. Any bench reactions that were moisture or air sensitive were performed in clean, oven-dried glassware under an argon environment. Ambient temperature was considered between 21–23 °C. Reaction monitoring was achieved by analytical thin layer chromatography (TLC) with Sorbent Technologies, Inc. silica hard layer (HL) TLC plates with unmodified silica 60 (250 μm) with 254 μm fluorescent indicator. Purifications using flash chromatography utilized ZEOprep 60 ECO flash silica gel (40–63 Å) purchased from ZEOCHEM or with a Teledyne ISCO Combiflash Rf 200 column chromatography unit and columns. Fluorous solid-phase extractions (FSPE) were performed with FluoroFlash cartridges (2 grams of silica gel bonded with perfluorooctylethylsilyl chains) acquired from Fluorous Technologies, Inc. The automation solution-phase platform used was a Chemspeed ASW2000 (Chemspeed Technologies AG, Augst, Switzerland) synthesis platform with a hood, 16 reactor vials (13 mL capacity for each vial), vacuum pump, and a heating/ cooling unit (200 °C to −20 °C). NMR spectra were recorded on a Varian Inova instrument at 500 or 600 MHz for ¹H spectra while ¹³C spectra were recorded at 125 MHz. The reported chemical shifts (δ) are in parts per million (ppm) relative to deuterated chloroform. ESI-HRMS mass spectra were obtained on a Thermo Scientific LTQ-Orbitrap XL. Analytical HPLC chromatographs were obtained from an Agilent 1200 Analytical HPLC instrument and one separation was performed with an Agilent 1200 Preparative HPLC instrument.

4.2. Synthetic procedures

4.2.1. Propyl 4-O-benzyl-3-O-(4-methoxybenzyl)-1-thio-β-L-rhamnopyranoside (4)

Intermediate 1 (2.1 g, 6.72 mmol) was co-evaporated with toluene (25 mL × 2) and then was diluted in anhydrous toluene (67 mL). Dibutyltin oxide (2.02 g, 8.11 mmol) was poured into the reaction solution and the resulting suspension was heated at 110 °C overnight. After overnight heating, the suspension had become a solution that was cooled to ambient temperature and concentrated down to a crude brown oil. The crude brown oil was dissolved in anhy. N, N-dimethylformamide (67 mL) then had cesium fluoride (2.07 g, 13.63 mmol) and 4methoxybenzyl chloride (1 mL, 7.53 mmol) added. The resulting suspension was heated at 85 °C for 4 h. After 4 h, the reaction solution was cooled to ambient temperature and diluted with ethyl acetate (60 mL) and DI water (60 mL). The extracted aqueous layer was further washed with ethyl acetate (2 × 35 mL) while the combined organic layers were washed with deionized (DI) water (2 × 60 mL) and brine (60 mL). The organic layer was dried with Na₂SO₄, filtered, and the solvent removed under reduced pressure. Column chromatography was performed and gave compound 4 as a colorless, transparent oil (1.18 g, 2.73 mmol, 40%).¹H NMR (400 MHz, CDCl₃) δ 7.38 – 7.26 (m, 7H), 6.86 (d, J = 8.7 Hz, 2H), 5.24 (s, 1H), 4.87 (d, J = 11.0 Hz, 1H), 4.62 (d, J = 11.1 Hz, 1H), 4.60 (s, 2H), 4.10 – 4.03 (m, 2H), 3.80 (s, 3H), 3.79–3.77 (m, 1H), 3.45 (t, J = 9.3 Hz, 1H), 2.65 – 2.45 (m, 3H), 1.63 (m, 2H), 1.30 (d, J = 6.2 Hz, 3H), 0.97 (t, J = 7.3 Hz, 3H). ¹³C NMR (125 MHz, CDCl₃) 159.63, 138.57, 129.96, 129.78, 128.53, 128.07, 127.86, 114.13, 83.69, 80.36, 80.11, 75.48, 71.90, 70.32, 68.01, 55.43, 33.24, 23.11, 17.98, 13.53. HRMS (ESI) m/z: C₂₄H₃₂O₅SNa calculated for: 455.1863; HRMS Found: 425.1863 [M+Na]⁺

4.2.2. Propyl 2-O-benzoyl-4-O-benzyl-3-O-(4-methoxybenzyl)-1-thio-β- L-rhamnopyranoside (S4)

Intermediate 4 (1.18g, 2.73 mmol) was dissolved in anhydrous pyridine (5.5 mL) and cooled to 0 °C via an ice-bath. To the cooled solution, benzoyl chloride (0.48 mL, 4.09 mmol) was added and the ice-bath was removed. The solution was allowed to stir at ambient temperature for 2 hours and then was quenched with methanol (0.3 mL). The quenched solution had the solvent removed under reduced pressure and was coevaporated with toluene. The resulting residue was column purified to provide S4 (1.22 g, 2.27 mmol, 84%) as a transparent, colorless oil. ¹H NMR (600 MHz, CDCl₃) δ 8.10 – 8.06 (m, 2H), 7.59 (t, J = 7.4 Hz, 1H), 7.47 (m, 2H), 7.36–7.27 (m, 5H), 7.21 (d, J = 8.6 Hz, 2H), 6.78 (d, J = 8.6 Hz, 2H), 5.67 (dd, J =3.0, 1.7, 1H), 5.29 – 5.27 (m, 1H), 4.90 (d, J = 10.9 Hz, 1H), 4.67 (d, J = 11.0 Hz, 1H), 4.62 (d, J = 10.9 Hz, 1H), 4.47 (d, J = 11.0 Hz, 1H), 4.13 (m, 1H), 3.96 (dd, J = 9.3, 3.2 Hz, 1H), 3.76 (s, 3H), 3.54 (t, J = 9.4 Hz, 1H), 2.67 – 2.53 (m, 2H), 1.65 (ddt, J = 10.5, 7.2, 3.2 Hz, 1H), 1.36 (d, J = 6.2 Hz, 2H), 0.98 (t, J = 7.3 Hz, 3H). ¹³C NMR (125 MHz, CDCl₃) δ 165.89, 159.36, 138.60, 133.30, 130.20, 130.09, 130.04, 129.88, 128.54, 128.46, 128.19, 127.79, 113.88, 82.98, 80.40, 78.30, 75.46, 71.52, 71.32, 68.54, 55.35, 33.90, 23.22, 18.21, 13.50. HRMS (ESI) m/z: C₃₁H₃₆O₆SNa calculated for: 559.2125; HRMS Found: 559.2125 [M+Na]⁺

4.2.3. Propyl 2-O-benzoyl-4-O-benzyl-1-thio-β- L-rhamnopyranoside (5)

Intermediate S4 (0.60 g, 1.12 mmol) was dissolved in a biphasic mixture of anhydrous dichloromethane (37.5 mL) and DI water (2.1 mL). To the biphasic solution, 2, 3-dichloro-5, 6dicyano-1, 4-benzoquinone (0.6 g, 2.64 mmol) was poured in and the reaction was stirred at ambient temperature for 4 hours. After 4 hours, the reaction solution was washed with saturated aqueous NaHCO₃ (3 × 35 mL) and the organic layer was dried with Na₂SO₄, filtered, and the solvent was removed under reduced pressure. The resulting crude product was purified via column chromatography to provide 5 (0.38 g, 0.91 mmol, 81 %) as a colorless, transparent oil. ¹H NMR (500 MHz, CDCl₃) δ 8.07 – 8.04 (m, 2H), 7.63–7.57 (m, 1H), 7.47 (m, 2H), 7.40 – 7.28 (m, 5H), 5.45 (dd, J = 3.4, 1.5 Hz, 1H), 5.30 (d, J = 1.1 Hz, 1H), 4.85 (d, J = 11.2 Hz, 1H), 4.77 (d, J = 11.2 Hz, 1H), 4.20 – 4.13 (m, 2H), 3.51 (t, J = 9.4 Hz, 1H), 2.68 – 2.53 (m, 2H), 2.17 (d, J = 5.0 Hz, 1H), 1.65 (m, 2H), 1.40 (d, J = 6.2 Hz, 3H), 0.99 (t, J = 7.3 Hz, 3H). ¹³C NMR (125 MHz, CDCl₃) δ 166.33, 138.27, 133.50, 130.00, 129.91, 128.70, 128.60, 128.18, 128.12, 82.76, 82.14, 75.33, 75.28, 71.22, 68.29, 33.92, 23.22, 18.26, 13.49. HRMS (ESI) m/z: C₂₃H₂₈O₅SNa calculated for: 439.1550; HRMS Found: 439.1551 [M+Na]⁺

4.2.4. Propyl 2-O-benzoyl-4-O-benzyl-3-O-levolinoyl-1-thio-β- L-rhamnopyranoside (6)

Compound 5 (0.47 g, 1.13 mmol) was placed under argon and dissolved in anhydrous dichloromethane (11.3 mL). The solution then had the following reagents added in the listed order: levulinic acid (0.175 mL, 1.69 mmol), N-(3-dimethylaminopropyl)-N’-ethylcarbodiimide hydrochloride (0.446 g, 2.32 mmol), and 4-dimethylaminopyridine (0.082 g, 0.68 mmol). The reaction solution was stirred at ambient temperature overnight. The reaction solution was washed with DI water (3 × 25 mL), 1 N HCl (3 × 25 mL), satd. aq. NaHCO₃ solution (3 × 25 mL), and brine (25 mL). The organic layer was extracted, dried with Na₂SO₄, filtered, and the solvent was removed under reduced pressure. The resulting crude material was purified by column chromatography (slow gradient of 20% diethyl ether in pentane to 50% diethyl ether in pentane). Compound 6 was isolated as a clear, colorless oil (0.37 g, 0.72 mmol, 64 %). ¹H NMR (600 MHz, CDCl₃) δ 8.07 – 8.04 (m, 2H), 7.62 (m, 1H), 7.49 (m, 2H), 7.36 – 7.27 (m, 5H), 5.58 (dd, J = 3.2, 1.6 Hz, 1H), 5.37 (dd, J = 9.7, 3.3 Hz, 1H), 5.26 (d, J = 1.2 Hz, 1H), 4.75 (d, J = 11.2 Hz, 1H), 4.65 (d, J = 11.2 Hz, 1H), 4.28 – 4.22 (m, 1H), 3.65 (t, J = 9.5 Hz, 1H), 2.76–2.69 (m, 1H), 2.68 – 2.53 (m, 3H), 2.53 – 2.39 (m, 2H), 2.09 (s, 3H), 1.65 (m, 2H), 1.38 (d, J = 6.2 Hz, 3H), 0.99 (t, J = 7.3 Hz, 3H). ¹³C NMR (126 MHz, CDCl₃) δ 206.32, 171.83, 165.66, 138.11, 133.54, 129.97, 129.88, 128.68, 128.56, 128.06, 127.95, 82.54, 79.08, 75.12, 72.82, 72.77, 68.50, 37.99, 33.69, 29.87, 28.10, 23.13, 18.19, 13.49. HRMS (ESI) m/z: C₂₈H₃₄O₇SNa calculated for: 537.1917; HRMS Found: 537.1919 [M+Na]⁺.

4.3. Automation Platform Procedures and Compounds

4.3.1. General Procedures and Automation Set-Up

These procedures and the automation set-up information can be found in the supporting information on page S5–8.

4.3.2. 2-(4-((4,4,5,5,6,6,7,7,8,8,9,9,10,10,11,11,11-heptadecafluoroundecyl)oxy)phenoxy)ethyl 2-O-benzoyl-4-O-benzyl-α-L-rhamnopyranosyl (1→2)-3,4-di-O-benzyl-α-L-rhamnopyranoside (7)

This compound was formed using the automation platform protocol outlined in Table S1. Purification was achieved utilizing a Phenomenex Luna 5 μm C18(2), 100 Å, LC Column 250 × 21.2 mm on the Agilent 1200 Preparatory HPLC. Elute gradient: 0–30 min 95:5 acetonitrile (ACN):H₂O (with 0.1% TFA), 30–50 min 95–100:5–0 ACN:H₂O, 50–62 min 100% ACN. The collected material had the solvent removed under reduced pressure and compound 7 was isolated as a clear, colorless oil (0.017 g, 0.013 mmol, 62% overall yield). ¹H NMR (600 MHz, CDCl₃) 8.08 – 8.03 (m, 2H), 7.61 (m, 1H), 7.48 (m, 2H), 7.39 – 7.26 (m, 12H), 7.19 (d, J = 7.3 Hz, 2H), 7.13 (m, 1H), 6.86 – 6.78 (m, 4H), 5.52 (dd, J = 3.2, 1.7 Hz, 1H), 5.12 (d, J = 1.3 Hz, 1H), 4.88 (d, J = 10.9 Hz, 1H), 4.86 – 4.83 (m, 2H), 4.73 (d, J = 11.1 Hz, 1H), 4.65 (s, 2H), 4.62 (d, J = 10.8 Hz, 1H), 4.31 (dd, J = 9.4, 3.4 Hz, 1H), 4.09 – 4.02 (m, 3H), 3.96 (t, J = 5.9 Hz, 2H), 3.94 – 3.84 (m, 3H), 3.79 – 3.70 (m, 2H), 3.47 (dt, J = 13.8, 9.4 Hz, 2H), 2.29 (m, 2H), 2.06 (m, 2H), 1.37 (d, J = 6.2 Hz, 3H), 1.30 (d, J = 6.2 Hz, 3H). ¹³C NMR (126 MHz, CDCl₃) δ 166.34, 153.34, 153.11, 138.65, 138.46, 138.35, 133.44, 130.06, 129.96, 128.67, 128.57, 128.49, 128.47, 128.22, 128.20, 128.08, 127.78, 127.68, 127.61, 115.92, 115.60, 99.22, 99.19, 81.81, 80.32, 79.88, 75.55, 75.37, 74.98, 73.39, 72.35, 70.69, 68.23, 67.93, 67.12, 66.01, 28.13, 20.82, 18.37, 18.17. HRMS (ESI) m/z: C₅₉H₅₇O₁₂F₁₇Na calculated for: 1303.3471; HRMS Found: 1303.3452 [M+Na]⁺

4.3.3. 2-(4-((4,4,5,5,6,6,7,7,8,8,9,9,10,10,11,11,11-heptadecafluoroundecyl)oxy)phenoxy)ethyl 3,4-di-O-benzyl-α-L-rhamnopyranosyl-(1→3)- 2-O-benzoyl-3-O-benzyl-α-Lrhamnopyranosyl (1→2)-3,4-di-O-benzyl-α-L-rhamnopyranoside (8)

This compound was formed using the automation platform protocol outlined in Table S2. Purification was achieved utilizing flash chromatography and 8 was isolated as a clear, colorless oil (0.0119 g, 6.97 μmol, 23 % overall yield). ¹H NMR (500 MHz, CDCl₃) δ 8.07 (d, J = 7.3 Hz, 2H), 7.59 (m, 1H), 7.47 (m, 2H), 7.37 – 7.23 (m, 16H), 7.23–7.18 (m, 8H), 7.13 (m, 1H), 6.84 – 6.762(m, 4H), 5.56 – 5.54 (m, 1H), 5.17 – 5.14 (m, 1H), 5.13 (s, 1H), 4.88 – 4.44 (m, 12H), 4.30 (dd, J = 9.3, 3.1 Hz, 1H), 4.05 – 4.00 (m, 3H), 3.94 (t, J = 5.9 Hz, 2H), 3.92 – 3.80 (m, 5H), 3.77 – 3.65 (m, 3H), 3.55 (t, J = 9.4 Hz, 1H), 3.48 (t, J = 9.4 Hz, 1H), 3.39 (t, J = 9.3 Hz, 1H), 2.28 (m, 2H), 2.04 (m, 2H), 1.28 (m, J = 9.4, 6.2 Hz, 6H), 1.14 (d, J = 6.2 Hz, 3H). ¹³C NMR (126 MHz, CDCl₃) δ 165.48, 153.34, 153.09, 138.72, 138.66, 138.59, 138.09, 133.24, 129.97, 128.67, 128.58, 128.47, 128.43, 128.38, 128.32, 128.06, 127.99, 127.92, 127.84, 127.74, 127.71, 127.64, 127.47, 115.92, 115.58, 99.20, 98.95, 80.77, 80.29, 79.91, 79.76, 75.55, 75.17, 74.88, 72.81, 72.27, 72.19, 69.28, 68.50, 68.38, 68.26, 67.94, 67.12, 65.98, 29.82, 20.81, 18.29, 18.14, 17.82. HRMS (ESI) m/z: C₇₉H₇₉O₁₆F₁₇Na calculated for: 1629.4989; HRMS Found: 1629.4999 [M+Na]⁺

4.3.4. 2-(4-((4,4,5,5,6,6,7,7,8,8,9,9,10,10,11,11,11-heptadecafluoroundecyl)oxy)phenoxy)ethyl 3,4-di-O-benzyl-2-O-sulfonate-α-L-rhamnopyranosyl (1→2)-3,4-di-O-benzyl-α-L-rhamnopyranoside (10)

This compound was formed using the automation platform protocol outlined in Table S3. The compound was purified via column chromatography (elute 92:5:3 EtOAc: MeOH: DI H₂O) and the like fractions were combined and the solvent removed under reduced pressure to give 10 as a transparent, colorless oil (0.009 g, 6.6 μmol, 31% overall yield). ¹H NMR (500 MHz, CDCl₃) δ 7.38–7.26 (m, 12H), 7.25 – 7.11 (m, 8H), 6.84 – 6.77 (m, 4H), 5.40 (s, 1H), 4.98 (s, 1H), 4.83 (m, 2H), 4.77 (d, J = 11.7 Hz, 2H), 4.68 – 4.52 (m, 3H), 4.45 (m, 2H), 4.02–3.89 (m, 6H), 3.88–3.82 (m, 2H), 3.79 (d, J = 9.2 Hz, 1H), 3.71–3.64 (m, 2H), 3.61 (t, J = 9.5 Hz, 1H), 3.23 (t, J = 9.2 Hz, 1H), 2.38–2.23 (m, 2H), 2.10–2.02 (m, 2H), 1.30 (d, J = 6.4 Hz, 4H), 1.21 (d, J = 6.0 Hz, 4H).¹³C NMR (126 MHz, CDCl₃) δ 153.26, 153.06, 138.61, 138.48, 138.42, 137.24, 128.88, 128.64, 128.51, 128.49, 128.45, 128.25, 128.17, 128.15, 128.11, 127.76, 127.75, 127.72, 115.80, 115.55, 99.63, 98.82, 80.23, 79.86, 78.57, 76.66, 76.39, 75.41, 75.26, 74.53, 71.85, 71.69, 68.59, 67.88, 67.77, 67.06, 66.05, 28.29, 28.11, 27.93, 20.79, 18.10, 17.78. HRMS (ESI) m/z: C59H58O14F17 S calculated for: 1345.3281; HRMS Found: 1345.3244 [M⁻]

4.3.5. 2-(4-((4,4,5,5,6,6,7,7,8,8,9,9,10,10,11,11,11-heptadecafluoroundecyl)oxy)phenoxy)ethyl 2-O-benzoyl-4-O-benzyl-3-O-sulfonate-α-L-rhamnopyranosyl (1→2)-3,4-di-O-benzyl-α-L-rhamnopyranoside (11)

This compound was formed using the automation platform protocol outlined in Table S4. The compound was purified via column chromatography (silica, elute 92:5:3 EtOAc: MeOH: DI H₂O) and the like fractions were combined and the solvent removed under reduced pressure to give 11 as a transparent, colorless oil (0.01 g, 7.1 μmol, 33% overall yield). ¹H NMR (600 MHz, CDCl₃) δ 7.98 (d, J = 7.5 Hz, 2H), 7.41 (t, J = 6.9 Hz, 1H), 7.30–7.27 (m, 3H), 7.25–7.21 (m, 3H), 7.20 – 7.12 (m, 9H), 7.11–7.04 (m, 3H), 6.80 – 6.72 (m, 4H), 5.96 (s, 1H), 5.26 (s, 1H), 5.02 (d, J = 9.3 Hz, 1H), 4.95 (d, J = 10.8 Hz, 1H), 4.84 (s, 1H), 4.72 (d, J = 10.7 Hz, 1H), 4.62 (d, J = 11.6 Hz, 1H), 4.57 – 4.52 (m, 2H), 4.46 (d, J = 10.8 Hz, 1H), 4.06 (s, 1H), 3.92 (t, J = 5.7 Hz, 4H), 3.84–3.77 (m, 3H), 3.73 – 3.60 (m, 4H), 3.51 – 3.45 (m, 1H), 2.32–2.22 (m, 4H), 2.06–2.00 (m, 2H), 1.32 (d, J = 6.0 Hz, 3H), 1.23 (d, J = 5.9 Hz, 3H). ¹³C NMR (126 MHz, CDCl₃) δ 166.42, 153.34, 153.00, 138.83, 138.22, 138.01, 133.35, 130.29, 129.67, 129.12, 128.38, 128.29, 128.04, 128.01, 127.76, 127.64, 127.43, 115.80, 115.52, 99.05, 80.28, 79.91, 75.23, 74.96, 72.42, 71.91, 68.57, 67.75, 67.08, 66.03, 28.11, 20.80, 18.28, 17.72. HRMS (ESI) m/z: C₅₉H₅₆O₁₅F₁₇S calculated for: 1359.3074; HRMS Found: 1359.3037 [M⁻]

Highlights.

• First synthesis of any sulfated rhamnan natural product

• Development of an automated solution-phase process for carbohydrate sulfation

• Automated solution-phase syntheses of monosulfated rhamnan dimers

• Automated solution-phase syntheses of alternating 1,2 and 1,3-linked rhamnans