A Repeating Sulfated Galactan Motif Resuscitates Dormant Micrococcus luteus Bacteria

ABSTRACT

Only a small fraction of bacteria can autonomously initiate growth on agar plates. Nongrowing bacteria typically enter a metabolically inactive dormant state and require specific chemical trigger factors or signals to exit this state and to resume growth. Micrococcus luteus has become a model organism for this important yet poorly understood phenomenon. Only a few resuscitation signals have been described to date, and all of them are produced endogenously by bacterial species. We report the discovery of a novel type of resuscitation signal that allows M. luteus to grow on agar but not agarose plates. Fractionation of the agar polysaccharide complex and sulfation of agarose allowed us to identify the signal as highly sulfated saccharides found in agar or carrageenans. Purification of hydrolyzed κ-carrageenan ultimately led to the identification of the signal as a small fragment of a large linear polysaccharide, i.e., an oligosaccharide of five or more sugars with a repeating disaccharide motif containing d-galactose-4-sulfate (G4S) 1,4-linked to 3,6-anhydro-α-d-galactose (DA), G4S-(DA-G4S)_n≥2.

IMPORTANCE Most environmental bacteria cannot initiate growth on agar plates, but they can flourish on the same plates once growth is initiated. While there are a number of names for and manifestations of this phenomenon, the underlying cause appears to be the requirement for a molecular signal indicating safe growing conditions. Micrococcus luteus has become a model organism for studying this growth initiation process, often called resuscitation, because of its apparent connection with the persistent or dormant form of Mycobacterium tuberculosis, an important human pathogen. In this report, we identify a highly sulfated saccharide from agar or carrageenans that robustly resuscitates dormant M. luteus on agarose plates. We identified and characterized the signal as a small repeating disaccharide motif. Our results indicate that signals inherent in or absent from the polysaccharide composition of solid growth media can have major effects on bacterial growth.

INTRODUCTION

Only a tiny minority of environmental bacteria can grow on agar plates, and the overwhelming majority that fail to grow are usually described as unculturable (1, 2). These unculturable cells form a heterogeneous group, including injured and dead cells but also cells that are viable but in a dormant state. Therefore, for many of the unculturable cells, the issue is not that they cannot grow but that they cannot autonomously initiate growth (3).

Under adverse conditions, such as nutrient deprivation or antibiotic treatment, bacteria can enter a metabolically inactive state that allows them to survive. Exit from this dormant state typically requires an external or autocrine factor that signals the return of favorable conditions and initiates growth (4, 5). Micrococcus luteus, a widely distributed Actinobacteria species, has become a popular model organism for this phenomenon. The ability, or inability, of M. luteus to initiate growth attracted attention through its connection with a small secreted protein called Rpf (resuscitation-promoting factor) (6). Rpf was initially discovered during studies on the human pathogen Mycobacterium tuberculosis, whose dormant (persistent) form often frustrates antibiotic treatment. In M. tuberculosis, rpf knockouts had a growth defect and difficulty exiting dormancy (7, 8). While M. tuberculosis is a slow-growing pathogen with five different rpf genes, M. luteus, which can be easily cultured, has only one (9). Rpf proteins are not the only known signals for exiting dormancy; muropeptides, soluble glycosylated proteins representing fragments of cell wall peptidoglycan, have also been widely studied as resuscitation agents for Bacillus subtilis and other Gram-positive bacteria (9, 10). Since dormancy is common in environmental bacteria, it seems likely that resuscitation signals are also widespread and that many macromolecular and small-molecule growth factors remain to be discovered. In addition to the molecular signals, many other aspects of dormancy and resuscitation remain enigmatic, and even the term dormancy itself can refer to a range of different physiological and metabolic states (11). The issue of what dormancy and related terms describe and how or whether they can be distinguished from antibiotic-tolerant persister cells or the so-called viable but nonculturable state is far from resolved (4, 12, 13). For example, the transition of the cytosol from liquid to glass-like properties and its fluidization by cellular metabolism have recently been proposed to explain the shift between active growth and the passive dormant state of a cell (14). Since some forms of dormancy are considered to be linked to latency and persistence of various bacterial pathogens, investigating the signals and mechanisms by which bacteria may enter into and exit from dormancy is important for our understanding of human infectious diseases (13, 15, 16). Expanding our general understanding of the different signals and forms of dormancy may ultimately provide a clearer picture of what dormancy is and how different forms of dormancy may affect microbial ecosystems and human diseases. In this study, we sought to focus on identifying a factor responsible for initiating growth in M. luteus.

We became interested in M. luteus and its complex dormancy behavior when we noted that, although it could not spontaneously initiate growth on agarose plates, it grew readily on agar plates. This difference suggested that something that was lost from agar in its conversion to agarose could resuscitate dormant bacteria.

RESULTS

In an initial experiment with 1.5% agar and agarose plates, we made a surprising discovery, i.e., whereas Escherichia coli and Bacillus subtilis grew well, with no discernible differences, on both solid media, Micrococcus luteus could be grown only on agar plates and not on plates solidified with the same amount of agarose (Fig. 1; also see Fig. S1A in the supplemental material). This effect was reproducible with various types of agarose (Fig. S1B), different media, and different inoculum sizes (see the supplemental material). These findings were especially puzzling because agarose is a polysaccharide prepared from agar.

FIG 1.

(A) Comparison of M. luteus growth on agar and agarose plates with NBE after 5 days of incubation. (B) Resuscitation of M. luteus preincubated for various times on NBE agarose plates and then transferred to NBE agar plates, where the numbers of CFU were assessed. (C) Idealized chemical structures of different extremes of the agar/agarose polysaccharide complex.

Even after prolonged incubation for more than 1 month, M. luteus did not form visible colonies. To investigate how the viability of M. luteus was influenced by agarose, we sampled the plates at different time points. CFU were counted by diluting a defined sample in medium and plating it on agar plates. After a slight initial increase, the CFU recovered from the sample decreased by more than 3 to 4 log units within 250 h on agarose plates (Fig. 1B), indicating that M. luteus cells were still viable after days on agarose but viability was lost continuously with increasing incubation time. When we prepared plates from a mixture of agar and agarose, the growth of M. luteus could be detected only in the presence of mixtures with more than 10% agar. Therefore, we concluded that the effect on M. luteus growth is not due to an inhibitory component of agarose but may result from a growth-activating constituent of agar.

Agar is usually described as a linear polysaccharide with alternating (1→4)-linked 3,6-anhydro-α-l-galactose and (1→3)-linked β-d-galactose units, with additional modifications by added sulfate and pyruvate groups. Lesser amounts of d-glucuronic acid and 6-O-methy-d-galactose have also been reported. Those modifications of the basic agar structure generate continuous variations in composition, and the structures are usually grouped into three main categories, i.e., (i) a sulfated galactan fraction with a large proportion of 3,6-anhydro-α-l-galactose substituted by galactose sulfate, (ii) a pyruvated agarose fraction with d-galactose substituted by 4,6-O-(1-carboxyethylidene)-d-galactose and small amounts of sulfated galactose, and (iii) an agarose fraction that is almost neutral in charge, with very small amounts of sulfate and pyruvate groups (Fig. 1C) (17).

To isolate the suspected growth-inducing component from agar, we used an established protocol for the fractionation of agar into its three extreme polysaccharide fractions (17, 18). Washing agar thoroughly with dilute aqueous saline at 20°C and precipitation with ethanol yielded the sulfated galactan fraction (fraction I). Another washing step at 50°C yielded the fraction described as pyruvated agarose (fraction II). The agarose fraction (fraction IV) was obtained from the residual polysaccharide by dissolution and precipitation steps with polyethylene glycol (PEG) 6000 and ethanol. The polysaccharide that did not precipitate with PEG was collected as a separate fraction (fraction III). Fourier transform infrared (FTIR) spectroscopic analysis revealed a strong band at 1,240 cm⁻¹ for fraction I, which is characteristic of sulfated polysaccharides (Fig. 2A) (19). Elemental analyses of the sulfur content of all fractions confirmed fraction I as highly sulfated polysaccharide, with an average of 70% sulfated disaccharide units; the sulfate content decreased gradually from fraction I to fraction IV (Fig. 2B). Strong infrared (IR) absorption bands at 1,615 cm⁻¹ indicated the presence of pyruvate groups (20) in fractions I and II, and a characteristic band for 3,6-anhydrogalactose was detected at 930 cm⁻¹ (Fig. 2A) (19). Fractions I and II did not form stable gels, and all fractions were tested in growth assays as mixtures (13% [wt/wt]) with agarose. Only fraction I induced growth under those conditions (Fig. 2C), and it did so in a concentration-dependent manner (Fig. S2A). No growth occurred with fractions II, III, and IV at 13%. Although fraction IV still contained more sulfate groups than commercial agarose, growth was strongly diminished even when the fraction was used as the sole solidifying agent (Fig. S2B).

FIG 2.

Fractionation of agar into different extremes of the agar polysaccharide complex. (A) FTIR analysis of the fractions, compared to commercial agarose. (B) Elemental analysis of the sulfur contents of agar, its fractions, and agarose. The sulfate content was estimated on the assumption of a homogeneous backbone with the disaccharide agarobiose as the unit. (C) Photographs of representative experiments, showing that only the sulfated galactan fraction I induces growth of M. luteus.

To investigate whether a sulfated galactan may be responsible for growth induction, we prepared sulfated agarose using SO₃·Pyr and purified the product (Fig. 3A). IR and elemental analyses confirmed the sulfate incorporation at a statistical ratio of 0.9 ± 0.3 sulfate units per agarobiose unit. The sulfated agarose induced growth equally as well as agar (Fig. 3B), while the addition of sodium sulfate at even higher doses, compared to agarose, had no effect (Fig. S2C). These results indicate that a sulfated polysaccharide but not sulfate alone is required for M. luteus growth induction.

FIG 3.

Sulfated agarose induction of M. luteus growth. (A) Synthesis of sulfated agarose. (B) Activity of the statistically sulfated agarose, compared to unmodified agarose and agar (representative experiment at 1.5% [wt/vol] in NBE).

Fractionation of agar and sulfation of agarose coupled with plate growth assays indicated some of the features required for resuscitation activity, but it seemed likely that a smaller active fragment could be defined in much greater molecular detail. Our search was focused by noticing the close similarity of fraction I to other sulfated galactans, and we empirically found κ-carrageenan to be a useful model system. The FTIR spectrum of fraction I closely resembled that of κ-carrageenan (Fig. S3), which has a well-defined structure of alternating d-galactose-4-sulfate (G4S) and 3,6-anhydro-α-d-galactose (DA) units (Fig. 4A). In plate resuscitation assays, agarose spiked with κ-carrageenan recapitulated the behavior of agar and the active fractions described above.

FIG 4.

Evidence that the sulfated oligosaccharide motif G4S-[DA-G4S]_n with n of ≥2, is responsible for M. luteus growth. (A) Generalized structure of the oligosaccharides. (B) Growth induction activity in comparison with agar and agarose (percentages are weight/volume values).

To investigate whether a high-molecular-weight sulfated polysaccharide such as κ-carrageenan or agar was required or whether lower-molecular-weight fragments would suffice, we performed mild acid hydrolysis of κ-carrageenan with 0.1 M H₂SO₄ and isolated the <3-kDa fraction using size exclusion filters. The resulting oligosaccharide fraction was fully active in M. luteus growth induction at 0.3% (wt/vol) added to 1.5% agarose plates, indicating that lower-molecular-weight fractions would be sufficient (Fig. 4B). ¹H and ¹³C nuclear magnetic resonance (NMR) spectroscopy of a <3-kDa mixture confirmed a structure with alternating G4S and DA units (Fig. 4A; also see Fig. S5A and B).

In order to identify the minimal active unit, we performed further hydrolysis and fractionation experiments. Mild hydrolysis of κ-carrageenan has been reported to selectively yield odd-numbered oligosaccharides flanked by G4S units, due to the acid instability of terminal DA (21). Initial purification of the <3-kDa oligosaccharide fraction by size exclusion chromatography resulted in one active fraction, which we identified by mass spectrometry as a mixture containing mainly the pentasaccharide, followed by the trisaccharide and the heptasaccharide, as well as traces of the nonasaccharide and the undecasaccharide (Fig. 5). The presence of mainly the pentasaccharide was supported by NMR analysis using corresponding reference spectra (Fig. S6) (22). Further hydrolysis of the <3-kDa mixture followed by size exclusion chromatography resulted in highly pure trisaccharide (Fig. S7A to C), which, however, exhibited only low levels of growth induction activity (Fig. S7D). Pure commercial neocarrabiose-4-O-sulfate and neocarratetrose-4^1,3-disulfate did not result in growth induction (Fig. S8A) and neither did a <1-kDa fraction collected by dialysis of the initial <3-kDa mixture, indicating the requirement for at least a pentasaccharide (1.03 kDa) or heptasaccharide (1.41 kDa). Therefore, we conclude that an oligosaccharide fragment with the structure given in Fig. 4A, with n of ≥2, is the minimal active motif.

FIG 5.

Oligosaccharide fractions obtained by size exclusion chromatography of hydrolyzed κ-carrageenan. (A) Activity of the fractions in comparison to crude hydrolyzed κ-carrageenan after 3-kDa filtration. Each well contained solid medium composed of 4 ml NBE with 12 mg/ml agarose and 3 mg/ml levels of material from fractionation and hydrolysis. Only fraction B induced growth of M. luteus. (B) ESI-mass spectrometric analysis, in negative ionization mode, of active fraction B, indicating the presence of sulfated trisaccharide, pentasaccharide, and heptasaccharide forms. (C) Expanded view of the region around m/z 360 to 370, showing traces of the nonasaccharide and undecasaccharide forms.

To investigate whether sulfated polysaccharides in general were capable of inducing the growth of M. luteus, we investigated a broad diversity of polysaccharides in 25% and 50% (wt/wt) mixtures with agarose, adding up to 1.5% (wt/vol) in plates. In a control experiment without additives with 0.75% (wt/vol) agarose, no growth was observed after 48 h, indicating that gel strength was not an important factor. In addition to κ-carrageenan, ι-carrageenan, with alternating G4S and 3,6-anhydrogalactose-2-sulfate units, and λ-carrageenan, with galactose-2-sulfate and galactose-2,6-disulfate units, induced growth within 48 h at both mixing ratios. Chondroitin sulfate A containing alternating β-glucuronic acid and N-acetyl-β-galactosamine-4-sulfate units only induced growth at a 50% mixing ratio with agarose. In contrast, the sulfated glycosaminoglycan heparin barely induced growth at 50%, and other sulfated polysaccharides, such as dextran sulfate and β-cyclodextrin sulfate, were entirely inactive (Fig. S8B).

DISCUSSION

We discovered that M. luteus was unable to initiate growth on plates with agarose as the solidifying agent but stayed viable on those plates, although viability slowly declined over days. When the viable but nongrowing cells were transplanted to a plate containing specific sulfated polysaccharides, growth resumed. This observation fits some of the typical definitions of dormancy and resuscitation (23), suggesting that M. luteus requires a signal to initiate growth. We identified this class of previously undescribed signals as sulfated polysaccharides. Our results demonstrate that not every sulfated polysaccharide induces growth and a specific secondary structural motif appears to be required for activity. A major finding in our analysis is that helical structures with a pitch of 25 to 28 Å are a shared feature of the active polysaccharides κ-carrageenan, ι-carrageenan, and chondroitin sulfate A (24, 25). In contrast, the secondary structures of the inactive saccharides are very different. For example, β-cyclodextrin is a small cyclic heptasaccharide, heparin has clusters of three adjacent sulfate groups every 16.7 Å (26), and dextran sulfate exhibits substantial branching by 1,3-linked side chains (27). In the animal kingdom, receptors for sulfated polysaccharides are known to discriminate between different sugar backbones and even different sulfation patterns (28, 29). It is thus likely that resuscitation of M. luteus involves the recognition of specific sulfated galactans. While the secondary structure and the charge of the sulfated saccharides may be initiating factors, the molecular mechanism of recognition of these sulfated saccharide motifs by M. luteus and the signal transduction leading to growth remain to be discovered.

The requirement for growth-inducing factors is most likely a consequence of the danger faced by a bacterial cell's commitment to growth. A bacterial cell in a dormant state is relatively safe, a reflection of the ironic difficulty of killing something that is barely alive. Antibiotics, for example, are largely ineffective against dormant bacteria, as evidenced by the current approaches to treating tuberculosis. There are a number of studies that illustrate the sensitivity of bacteria to their solid substrates. For Vibrio parahaemolyticus, for instance, mechanical blocking of the polar flagellum on solid medium triggers the production of lateral flagella that enable swarming motility (30). Solid substrates also have been reported to induce morphological changes (31), antibiotic production (32), and activation of two-component signal transduction pathways (33). M. luteus is ubiquitously distributed and is well known to thrive in marine environments, among many other habitats (34). Sulfated galactans are produced by a wide variety of organisms, including a large diversity of seaweeds but also marine plants and animals (35). Other sulfated polysaccharides with lower-growth inducing activity, such as chondroitin sulfate, are found in human skin (36), another habitat of M. luteus. It is thus plausible that M. luteus encounters various types of sulfated polysaccharides with growth-inducing, resuscitating activity in its diverse natural environments.

In conclusion, we demonstrated that M. luteus does not grow on solid media without a resuscitation factor. This factor can be provided by sulfated galactans of the agar polysaccharide complex that are structurally similar to sulfated polysaccharides of the carrageenan type. There is some structural specificity, however, because not all sulfated polysaccharides resuscitate M. luteus. The small active fragment isolated from a carrageenan hydrolysate is G4S-(DA-G4S)_n≥2, and the activity of a small part of a much larger molecule is likely to be a common theme.

MATERIALS AND METHODS

Materials.

Chemicals and solvents were generally purchased from Sigma-Aldrich, EMD Chemicals, or Carl Roth GmbH. Solvents were anhydrous or high-performance liquid chromatography (HPLC) grade, and chemicals were of reagent grade or better and were used without further purification. Temperatures were measured externally.

Agarose I (biotechnology grade) was purchased from AMRESCO (Solon, OH). ι-Carrageenan (product no. C4014), λ-carrageenan (product no. 22049), and κ-carrageenan (product no. 22048) were obtained from Sigma-Aldrich. Heparin sodium salt from porcine mucosa (≥100 IU/mg; CAS no. 9041-08-1) was purchased from Alfa Aesar. β-Cyclodextrin sulfated sodium salt (product no. 389153), chondroitin sulfate A sodium salt from bovine trachea (product no. C9819), and dextran sulfate sodium salt from Leuconostoc spp. (product no. 31404) were obtained from Sigma-Aldrich.

Instrumentation.

FTIR spectra were measured with an Alpha FTIR spectrometer (Bruker) with a resolution of 4 cm⁻¹ and 24 scans per spectrum. NMR spectra were obtained with Bruker Avance-III 400 and Bruker Avance-III 600 NMR spectrometers at ambient temperature. Chemical shifts (δ) are given in parts per million (ppm), relative to the solvent residual signal (D₂O, δ_H 4.79 ppm; dimethyl sulfoxide [DMSO]-d₆, δ_H 2.50 ppm and δ_C 39.52 ppm; MeOD-d₄, δ_H 3.31 ppm and δ_C 49.00 ppm) (37). The data obtained were processed and analyzed with Bruker Topspin 3.5 software. High-resolution mass spectrometry was carried out by the Small Molecule Mass Spectrometry facility of Harvard Faculty of Arts and Sciences with an Agilent 6210 liquid chromatography-time of flight mass spectrometry system, using NH₄OH as a mobile phase additive and using electrospray ionization (ESI) in negative ionization mode. Elemental analysis was performed by Micro-Analysis Inc. (Wilmington, DE), using O₂ flask combustion ion chromatography to quantify the sulfur content.

Bacterial cultures.

Micrococcus luteus ATCC 4698 was obtained from the American Type Culture Collection (ATCC) and maintained in nutrient broth E (NBE) (1 g/liter meat extract, 2 g/liter yeast extract, 5 g/liter peptone, 5 g/liter NaCl [pH adjusted to 7.4 with NaOH]) at 30°C, with continuous shaking at 250 rpm. All experiments were started from cultures that were inoculated from glycerol stocks and grown at 30°C in NBE liquid medium for 2 days, to stationary phase, before plating on supplemented or pure NBE plates with agar, agarose, or other polysaccharides as solidifying agents. Solid media were prepared by adding 1.5% solidifying agent (agar or agarose, 15 g/liter) to NBE. Escherichia coli ATCC 10798 and Bacillus subtilis ATCC 6633 were maintained in tryptic soy broth (TSB) (30 g/liter; product no. BD 211825; BD Biosciences) at 37°C. For agarose plates, 1.5% agarose I from AMRESCO was found to be most suitable; other agaroses gave similar results but might vary in their purity and sulfate contents. For agar plates and further fractionation studies, we used 1.5% Bacto agar (BD Biosciences). Plates were poured with autoclaved medium at 30 ml agar/agarose medium per 100-mm-diameter petri dish.

Different nutrient media.

Agar and agarose (1.5%) plates (diameter, 100 mm) were prepared with different nutrient media, 50 μl of a 2-day culture of M. luteus was plated, and growth was recorded over at least 5 days. The following media were used: Miller Luria-Bertani (LB) broth (Difco, BD Biosciences), NBE, and TSB (BD Biosciences). No growth occurred on any of the agarose plates, but dense growth was visible on the corresponding agar plates after 1 or 2 days of incubation at 30°C.

Dependence of inoculum size.

On NBE plates with agarose, volumes of 10 μl, 100 μl, and 1,000 μl of a densely grown culture of M. luteus were plated on 100-mm-diameter plates and incubated at 30°C. The plates were checked daily for visible growth of colonies, and no growth was recorded for the entire 25-day duration of the experiment.

Agar-agarose mixing experiments.

Freshly autoclaved NBE medium with 1.5% agar and agarose was premixed in conical tubes and poured into 100-mm-diameter plates, to give increasing percentages of agar from 0% to 100%. Fifty microliters of a densely grown culture of M. luteus was plated and incubated at 30°C. Growth was monitored, and digital photographs were taken for evaluation.

Fractionation of agar.

Agar (60 g of Bacto agar) was packed in a chromatography column and eluted with 50 mM aqueous NaCl solution. The eluate was tested with the colorimetric phenol-sulfuric acid (PSA) method until no sugar could be detected at a given temperature. For the PSA test, 500 μl of 4% aqueous phenol solution and 2 ml of 95% sulfuric acid were added to 100 μl of sample. A light yellow to dark red color indicated the presence of sugars. The first 900 ml eluting at 20°C gave the strongest signal in the PSA test. The polysaccharide was precipitated by the addition of 4 volumes of absolute ethanol, collected by centrifugation, washed with ethanol, and lyophilized (1.74 g; fraction I). The agar column was eluted exhaustively at 20°C with a total of 4,100 ml of 50 mM NaCl. The remaining material in the column was eluted at 50°C with a total of 7,150 ml of 50 mM NaCl. The first 3,050 ml was again precipitated with 4 volumes of absolute ethanol, collected by centrifugation, washed with ethanol, and lyophilized (2.67 g; fraction II). The residual nonextractable agar was lyophilized, and 10 g of dry material was liquified at 85°C in 330 ml of 50 mM NaCl. The clear solution was cooled to 65°C, and 100 g of solid PEG 6000 was added slowly. The precipitate was isolated by centrifugation while still hot and was washed twice with 200 ml of a 60°C solution of 25% PEG 6000 in 50 mM NaCl. The supernatants from the precipitation and washing steps were pooled and precipitated with 4 volumes of absolute ethanol. The precipitate was collected by centrifugation, washed four times with 300 ml (each time) of ethanol, and lyophilized (1.00 g; fraction III). The PEG precipitate was washed eight times at room temperature with 200 ml (each time) of 50 mM NaCl. PEG could be detected by adding KI₃ solution, which gives a dark complex in the presence of PEG. After six washes, there were no PEG traces in the supernatant. The insoluble material was washed twice with absolute ethanol and lyophilized (7.92 g; fraction IV).

Small-scale test of polysaccharides.

To 12 ml of NBE medium were added 144 mg of agarose and 36 mg (20% [wt/wt]) of an isolated fraction; the mixture was boiled by microwaving and distributed in a 12-well plate (Corning) in 4-ml/well portions. An inoculum of 2 μl of an undiluted, densely grown, 2-day culture of M. luteus in NBE medium was plated in each well, allowed to dry for 10 min (so that there was no liquid film covering the surface), and incubated at 30°C.

Concentration dependence of fraction I.

Polysaccharide fraction I was mixed with agarose (13%, 27%, and 53% [wt/wt] of fraction I) and used at a fixed percentage of 1.5% (wt/vol) in NBE medium to prepare 60-mm-diameter NBE plates (10 ml per plate).

Sulfation of agarose.

Agarose I (AMRESCO) was dried for 30 min at 70°C and lyophilized to remove residual water. To a solution of 1.26 ml of chlorosulfonic acid (18.9 mmol) in 80 ml of dry pyridine was added 919 mg of dry agarose, and the mixture was stirred for 1 h at 100°C. After cooling in an ice bath, 50 ml of 10% NaOH (2.5 M) was added under continuous stirring, followed by acetone until phase separation. The upper phase was discarded, and the lower phase was washed six times with 70 ml (each time) of acetone. The residue was taken up in 60 ml of deionized water and heated until the solid material completely dissolved. The polysaccharide was precipitated by the addition of 4 volumes of absolute ethanol and was collected by centrifugation. The solid material was washed twice with 50 ml of absolute ethanol and lyophilized. Seven hundred milligrams of the material was suspended in 10 ml of deionized water and dialyzed twice against 2 liters of deionized water, using a Spectra/Por Float-A-Lyzer dialysis device (molecular weight cutoff, 8,000 to 10,000 Da), at 4°C for 6 h. After dialysis, the sample was lyophilized to yield 220 mg of white solid product. The product was confirmed as sulfated polysaccharide by FTIR analysis, with a strong absorption band at 1,240 cm⁻¹. Elemental analysis was performed to determine the degree of sulfation of agarose (C, 31.0 ± 0.8; H, 4.6 ± 0.2; S, 4.6 ± 1.6). The number of sulfate groups was calculated and normalized based on a sample of ι-carrageenan (C, 25.0 ± 0.3; H, 3.7 ± 0.3; S, 10.1 ± 0.9), which is known to contain two sulfate units per disaccharide unit.

Hydrolysis of carrageenan.

A suspension of κ-carrageenan (1 g) in 100 ml of 0.1 M H₂SO₄ was stirred for 1.5 h at 60°C. The reaction was allowed to cool to room temperature, and the pH was neutralized by dropwise addition of 2 M NaOH. The material was lyophilized for storage. The hydrolysate was dissolved in 50 ml of water and passed through a filter with a molecular weight cutoff of 3,000 Da (Ultracel). The flowthrough fraction was lyophilized, and 25 mg of the mixture was fractionated on a Sephadex G15 size exclusion column, using 10% ethanol as the solvent.