Biomacromolecules 2005, 6, 1851-1857
1851
Articles Structural Characterization of r-Terminal Group of Natural Rubber. 1. Decomposition of Branch-Points by Lipase and Phosphatase Treatments Lucksanaporn Tarachiwin,† Jitladda Sakdapipanich,*,†,‡ Koichi Ute,§ Tatsuki Kitayama,§ Takashi Bamba,| Ei-ichiro Fukusaki,| Akio Kobayashi,| and Yasuyuki Tanaka† Department of Chemistry, Faculty of Science, Mahidol University, Bangkok 10400, Thailand, Institute of Science and Technology for Research and Development, Mahidol University, Salaya, Nakhonphatom 73170, Thailand, Department of Chemistry, Graduate School of Engineering Science, Osaka University, Toyonaka, Osaka 560-8531, Japan, and Department of Biotechnology, Graduate School of Engineering, Osaka University, Suita, 565-0871, Japan Received January 26, 2005
Deproteinized natural rubber latex (DPNR-latex) was treated with lipase and phosphatase in order to analyze the structure of the chain-end group (R-terminal). The enzymatic treatment decreased the content of longchain fatty acid ester groups in DPNR from about 6 to 2 mol per rubber molecule. The molecular weight and intrinsic viscosity were reduced to about one-third after treatment with lipase and phosphatase. The Huggins’ k′ constant of the enzyme-treated DPNR showed the formation of linear rubber molecules. The molecular weight distribution of DPNR changed apparently after treatment with lipase and phosphatase. 1H NMR spectrum of rubber obtained from DPNR-latex showed small signals due to monophosphate, diphosphate and phospholipids at the R-terminus. Treatment of DPNR-latex with lipase and phosphatase decreased the relative intensity of the 1H NMR signals corresponding to phospholipids, whereas no change was observed for the signals due to mono- and diphosphates. The residual mono- and diphosphate signals as well as some phospholipid signals after lipase and phosphatase treatments indicate that mono- and diphosphate groups are directly linked at the R-terminus with the modified structure, expected by aggregation or linking with phospholipid molecules. Introduction Natural rubber (NR) from HeVea brasiliensis shows a bimodal molecular-weight distribution (MWD), the profile of which varies with the clone.1 This characteristic bimodal distribution has been presumed to be due to branching in NR, i.e., the high molecular-weight fraction consists of trior tetra-functional branched molecules.2,3 It has been reported that these branch-points originate from small amounts of abnormal groups such as amino,4 epoxide,5 and aldehyde6 groups presented in the rubber molecule. Reactions of these groups are presumed to cause hardening of solid NR during long-period storage, so-called storage hardening. However, our previous study showed that both the low and high molecular-weight fractions of the bimodal MWD in NR were composed of branched molecules, although the degree of branching depends on the molecular weight.7 * Corresponding author. E-mail:
[email protected]. Tel: +66-28893116. Fax: +66-2889-3116. † Department of Chemistry, Faculty of Science, Mahidol University. ‡ Institute of Science and Technology for Research and Development, Mahidol University. § Department of Chemistry, Osaka University. | Department of Biotechnology, Osaka University.
In the previous study, we reported that NR contains two kinds of different functional groups at the initiating- and terminating-ends of the rubber chain, i.e., ω- and R-termini, respectively.8 A series of evidence suggested that the ω-terminus bonds with a peptide or unidentified functional group, whereas the R-terminus consists of phosphoric ester and two long-chain fatty acid ester groups.9 Both terminal ends have been presumed to originate the branch-points and soft-gel in NR.9 Deproteinization of NR with a proteolytic enzyme decomposes the branch-points and gel-containing proteins to form linear and branched molecules.10 The addition of small amounts of a polar solvent to rubber solutions results in solubilization of the gel fraction by decomposition of the hydrogen-bonding between proteins associated with the rubber molecule.9 Transesterification of deproteinized natural rubber (DPNR) in solution with sodium methoxide decomposes the residual branch-points containing ester linkages and liberates long-chain fatty acids as methyl esters.11 A skewed unimodal MWD of DPNR rich in the high molecular weight fraction changes to a bimodal MWD accompanied with the reduction in weight-average molecularweight (M h w) by 30% after transesterification.10 This finding
10.1021/bm058003x CCC: $30.25 © 2005 American Chemical Society Published on Web 06/04/2005
1852
Biomacromolecules, Vol. 6, No. 4, 2005
Tarachiwin et al.
suggested that the MWD is derived from the biosynthesis mechanism of HeVea rubber. It is commonly accepted that the active chain-end of NR, consisting of diphosphate, is involved in chain elongation by the addition of isopentenyl diphosphate during the course of rubber biosynthesis. Based on these findings, it is reasonable to consider that long chain fatty acid esters are a constituent of the R-terminal group that is expected to link with the rubber molecule as a component of a glyceride and/ or phospholipid. The present work is an attempt to elucidate the molecular structure of the R-terminal group of HeVea rubber by selective decomposition of ester linkages with enzymatic reactions using lipase and phosphatase.
copolymer gel with exclusion limits of 2.0 × 107 and 4 × 105. The rubber solution was prepared by dissolving rubber into THF (LabScan, HPLC grade) at a concentration of 0.05% w/v and filtered through a Millipore prefilter and 0.45 µm membrane filter (Alltech). THF was used as an eluent with a flow rate of 0.5 mL/min at 35 ( 0.01 °C, monitoring with a refractive index detector. cis-1,4-Polyisoprene (Polymer Standard Service GmbH, Germany) was used as a standard sample. Intrinsic viscosity was measured by the dilution method with a single bulb SCHOTT (CT52) Ubbelohde viscometer at 30 ( 0.02 °C. The viscosity average-molecular-weight was calculated according to the Mark-Houwink’s equation as given in (1):12
Experimental Section
[η] ) kMav
Freshly tapped natural rubber latex (FL-latex) was diluted to 20% dry rubber content (DRC) and subjected to enzymatic deproteinization. FL-latex was incubated with a protease (KP3939, Kao) 0.04% w/v in the presence of 0.5% v/v Triton X-100 at 37 °C for 12 h, followed by centrifugation at 20 300 g for 30 min and recentrifuged in a similar condition after redispersion with 0.2% Triton X-100 to make 10% DRC. DPNR-latex (10% DRC) was treated with a crude lipase (Candida rugosa, activity ≈ 901 units/mg solid, Sigma) at 37 °C by adjusting to pH 7.2 using phosphate buffers and was incubated for 45 h. The lipase-treated DPNRlatex was centrifuged twice at 8680 g for 30 min, followed by coagulation with methanol. Free fatty acids or nonlinked fatty acids in the rubber were removed by acetone extraction in a Soxhlet apparatus for 24 h under a nitrogen atmosphere to get acetone extracted-DPNR (AE-DPNR). An alkaline phosphatase (from Bovine Intestinal Mucosa, activity ) 4000-6000 DEA unit/mg protein, Sigma) was reacted with DPNR in toluene solution by using a phase-transfer technique at room temperature (ca ∼ 25 °C). The alkaline phosphatase 10% w/v was dissolved in distilled water. The mixture was added into 1% w/v rubber in toluene solution together with 0.01% w/v (n-butyl)4N+Cl- as a phase transfer catalyst. The mixture was vigorously stirred for 45 h at room temperature. The rubber was collected and purified by precipitation of the rubber solution using an excess of methanol and then dried in vacuo. Transesterification of the rubber from DPNR was carried out in 0.5% w/v toluene by reaction with freshly prepared sodium methoxide (NaOCH3) with stirring at room temperature for 3 h. The resulting transesterified-deproteinized natural rubber (TE-DPNR) was purified according to the procedure described above. The content of long-chain fatty acid esters was determined by FTIR using a JASCO FT/IR 460. A calibration curve was obtained for a series of mixtures of methyl stearate and synthetic cis-1,4-polyisoprene (Kuraprene IR10). The content of fatty acid ester groups per weight of rubber was determined by the intensity ratio of peaks at 1739 cm-1 (Cd O) to 1664 cm-1 (CdC). The molecular weight of NR samples was determined by size exclusion chromatography (JASCO-Borwin) using two columns in series, packed using polystyrene-divinylbenzene
Here, k and a are constant values of 33.1 × 10-5 and 0.71, respectively. The Huggins’ k′ constant was calculated using eq 2: ηsp/c ) [η] + k′[η]2
(1)
(2)
Here, ηsp/c, [η], c, M h v and k′ represent the reduced viscosity, intrinsic viscosity, concentration expressed as g/dL, viscosityaverage molecular-weight, and the Huggins’ constant, respectively. The 1H NMR measurement was performed at 750 MHz. All experiments were carried out on a Varian Inova-750 spectrometer. A 5 mm 1H-{13C-15N} indirect probe was used for the 1H NMR measurement. All spectra were recorded at 50 °C of samples in C6D6 solution at a concentration of 1% w/v with tetramethylsilane as an internal standard. All measurements were based on 512 scans obtained at a pulse repetition time of 10 s and 50° pulse width. The spectrum accumulation was carried out using a 128K data matrix. Results and Discussion The presence of acyglyceride and/or phospholipid at the rubber chain end, assumed by a series of supporting evidence, can be confirmed by a structural change of the rubber chain after treatment with lipase and phosphatase. These enzymes have high efficiency to hydrolyze selectively acylglycerides, including phospholipids and monophosphate esters, respectively. The rubber particles in latex are known to be covered with lipid and protein layers.13 Direct reaction of lipase and phosphatase on the rubber particles was difficult due to the presence of a protein layer on the surface of the rubber particles. Consequently, it was necessary to remove the proteins in latex before using these enzymes for enzymatic deproteinization. The effect of lipase and phosphatase was analyzed for DPNR-latex. It is known that lipase (triacylglycerol acylhydrolase or triacylglycerol lipase) is able to hydrolyze selectively the ester linkages of tri-, di-, and monoglycerides at the C1 and C3 positions in triacylglycerides, whereas phosphatase decomposes monophosphate ester linkages, as shown in Figure 1. However, it is very important to confirm the efficiency of these two enzymes using phospholipid mol-
Decomposition of Branch-Points
Figure 1. Decomposition position of (a) lipase on the acylglyceride and (b) phosphatase on monophosphate ester group.
Biomacromolecules, Vol. 6, No. 4, 2005 1853
Figure 3. FTIR spectra of (a) L-R-phosphatidic acid treated with phosphatase, (b) L-R-phosphatidylcholine treated with phosphatase and (c) L-R-phosphatidic acid monosodium salt.
Figure 2. FTIR spectra of L-R-phosphatidylcholine before and after lipase treatment.
ecules, such as L-R-phosphatidylcholine (PC) and L-Rphosphatidic acid sodium salt (PA), as representative models of phospholipids in NR, assuming the linkage between a phospholipid and rubber chain is at the R-terminus. The result obtained from enzymatic treatments would give useful information on the structure of phospholipids presumed to be at the rubber chain-end. To establish that the changes observed in the phospholipids were truly the effect of the enzyme added, the experiment without-enzyme and boiledenzyme control samples have been carried out. However, no difference in the structural characteristics of both samples was observed, so the without-enzyme DPNR-latex was used as a control sample for all experiments without the growth of bacterial activity. The FTIR spectra of PC before and after treatment with lipase are shown in Figure 2. The CdO stretching band of fatty acid ester in PC shifted from 1743 to 1710 cm-1 with a residual small band at 1743 cm-1 after the reaction with lipase. This indicates that lipase treatment removed most fatty acid ester groups, but not completely. As mentioned above, lipase decomposes fatty acid esters at the C1 and C3 of triglycerides. The shift of the CdO band after lipase treatment indicates that lipase has a capability to decompose fatty acid esters at the C1 position of phospholipid molecules and then liberates free fatty acids. Figure 3 shows the FTIR spectra of PC and PA treated with phosphatase. These two kinds of phospholipids are expected to play an important role on the effect of polar
Figure 4. Content of long-chainfatty acid ester in DPNR treated with lipase at different lipase concentrations.
headgroup in phospholipid. By considering the fact that phosphatase decomposes monophosphate, it is expected that the phosphate group in PC was intact, whereas that of PA was almost hydrolyzed after phosphatase treatment. The absorption band at 1240 cm-1 was assigned to PO-2 asymmetric stretching of the phospholipid bilayer.14,15 In the spectrum of PC, the PO-2 band remained even after phosphatase treatment, whereas it disappeared in PA treated with phosphatase. This indicates that phosphatase has an efficiency to decompose only monophosphate ester linkages of PA and liberates phosphoric acid, whereas phosphate ester groups in PC could not be removed by phosphatase treatment. In addition to the decomposition of monophosphate linkages in PA, phosphatase has the possibility to hydrolyze the phosphate group of polyprenyl phosphate (-CH2OP) and polyprenyl diphosphate (-CH2OPP) in rubber chains due to the presence of a monophosphate linkage. Figure 4 illustrates the relationship between the content of long-chain fatty acid ester groups in the rubber molecule from DPNR-latex against the lipase concentration by weight against the rubber weight, after incubation at 37 °C for 45 h. Here, the measurement of fatty acid ester content was carried out after extraction of the lipase treated rubber with acetone to remove free fatty acids and glycerides present as a mixture. AE-DPNR contained long-chain fatty acid ester groups in a concentration of 29 mmol/kg rubber correspond-
1854
Biomacromolecules, Vol. 6, No. 4, 2005
Tarachiwin et al.
Figure 5. Intrinsic viscosity and content of long-chainfatty acid ester in DPNR after lipase treatment at 37 °C with pH 7.2 for 45 h.
Figure 6. 1H NMR spectrum of DPNR measured in C6D6 at 750 MHz. Table 1. Structural Characteristics of DPNR, Transesterified DPNR (TE-DPNR), Lipase- and Phosphatase-Treated DPNR after Acetone Extraction
characteristics
AE-DPNR
TE-DPNR
lipase treated DPNR
phosphatase treated DPNR
ester content (mol/chain) [η] k′ M h v (×105) M h n (× 105) M h w (× 105) M h w/M hn
5.09
0.00
1.59
2.14
5.12 0.47 7.92 1.93 4.34 3.84
2.62 0.30 3.08 1.18 2.71 3.94
3.75 0.23 5.10 0.97 3.27 3.04
3.47 0.35 4.12 1.16 3.64 3.58
ing to ∼5.0 mol/ rubber chain. This decreased gradually to 17 mmol/kg rubber (∼1.5 mol/ rubber chain) with increasing lipase concentration (cf. Figure 6 and Table 1). Here, the ester content per rubber chain was calculated based on the number-average molecular-weight (M h n) values of rubber. The M h n values used for this calculation were obtained by size exclusion chromatography using polyisoprene standard, which represents the whole molecule including the branchpoints. In this experiment, it was necessary to use a large amount of crude lipase (up to 10 times the rubber weight) to decompose all fatty acid ester groups at C1 and C3 of triglycerides. This may be due to the low activity of crude lipase. As mentioned above, lipase can decompose fatty acid esters at the C1 and C3 of triacylglyceride compounds. The
decrease of ester content after lipase treatment clearly indicates the presence of acylglycerol compounds in rubber molecules. However, the ester groups could not be removed completely, retaining of about one-third of the original amount even after treatment of DPNR-latex with extremely high lipase concentrations, whereas all fatty acid ester groups are completely removed after chemical treatment such as transesterification and saponification.10 This can be explained by considering the fact that the fatty acid ester at the C2 position of triglycerides could not be decomposed by lipase treatment. This is direct evidence showing the presence of acylglycerol compounds in the rubber molecule. In view of the fact that all simple fatty acid esters and fatty acids can be extracted from NR by acetone after enzymatic treatment, the residual fatty acid esters are presumed to be linked to the rubber chain or polar lipids such as phospholipids and glycolipids, which are insoluble in acetone. In HeVea latex clonal RRIM 501, glycolipid is reported to comprise about 33% w/v of the total lipids.16 Thin-layer chromatography of the glycolipid in NR-latex showed the presence of four main spots corresponding to esterified steryl glycoside (ESG), monogalactosyl diglyceride (MGDG), steryl glucoside (SG), and digalactosyl diglyceride (DGDG).16 The 1H and 13C NMR spectra of 5-mycoloyl-R-arabinofuranosyl (1f1′)glycerol acetate (Gl-ai-II acetate) showed the carbohydrate signals ranging from 4.2 to 5.1 and 63.3 to 105.5 ppm, respectively, measured in CDCl3.17 However, these signals have not been detected in DPNR either the 1H or the 13C NMR spectra. This suggests the absence of glycolipids in DPNR. In higher plants, it is common that polyprenols are present as fatty acid esters.18 It was reported that the R-termini of rubbers from leaves of sunflower (Helianthus annuus)19 and Lactarius mushroom (Lactarius Volemus)20 are esterified with fatty acids. However, in the case of NR, the presence of fatty acid esterified at the R-terminus can be neglected by the fact that AE-DPNR and lipase treated DPNR showed no 13C NMR signal due to C-4 methylene carbon of the R-terminal isoprene unit linked to an ester group, which resonates at 60.9 ppm in polyprenol esters.19,20 These findings suggest that the residual fatty acid esters are included in a functional group presumably in phospholipids at the R-terminus and/or in free phospholipids, which cannot be extracted completely with acetone even after treatment with lipase. It is noteworthy that the treatment of lipase decreased the molecular weight of DPNR. Figure 5 shows the relationship between the intrinsic viscosity, [η], of the lipase treatedDPNR and the concentration of lipase, together with the longchain fatty acid ester content of DPNR. The [η] value of the lipase treated-DPNR decreased from ∼5.1 to ∼3.3 dL/g with increasing concentration of lipase. In proportion to the decrease in the [η] value, the ester content decreased gradually from ∼5.1 to 1.6 mol of ester groups per rubber molecule. Accordingly, this means that about 5 mol of ester groups per rubber molecule in AE-DPNR are reduced to about two moles per rubber molecule after lipase treatment. It is noteworthy that the presence of about two ester groups per single chain, which means a chain constituting branching, was confirmed by 13C NMR measurement of fractionated
Decomposition of Branch-Points
DPNR from FL-latex.11 In the present study, AE-DPNR and lipase treated DPNR showed the 13C NMR signal due to the methylene carbon next to the carboxyl group of a fatty acid ester, OCdOCH2, at 34.6 ppm. The relative intensity of this signal was about twice that of the C-1 methylene carbon signal from trans-1,4-isoprene units, which was confirmed to be two units per single chain.21 These findings are strong supporting evidence that NR molecules are partially linked altogether by a functional group containing glyceride backbone such as phospholipid to form branch-points. The decrease of ester content after lipase treatment was accompanied by a decrease in molecular weight. In the case of phospholipid, hydrolysis of the C1 ester group will contribute to reduce the ester content, by considering the active site of lipase. Meanwhile, the decrease in the molecular weight of DPNR should be due to the decomposition of branch-points. Consequently, it can be deduced that the removal of long-chain fatty acid ester group at C1 in the phospholipids by lipase treatment partly destroys a micelle structure and results in the decomposition of branch-points. Table 1 shows the content of long-chain fatty acid ester, [η], and molecular weight of lipase treated-DPNR compared with AE-DPNR and TE-DPNR as well as phosphatase treated-DPNR at the concentration ratio 10:1 against the rubber weight. The [η] value of the rubber from lipase treated-DPNR latex was lower than that of AE-DPNR, whereas it was higher than that of TE-DPNR. If the decrease in the molecular weight and the ester content is due to the decomposition of branch-points, it should be accompanied with the formation of linear rubber molecules. It is known that the Huggins’ k′ constant is a qualitative indicator of the presence of long-chain branching. For a given polymer, the k′ value is nearly independent of molecular weight and MWD, and it increases in proportion to the quantity of branching in the polymer chain.22,23 The k′ values of 0.450.65 are attributed to the branched molecules, whereas that of about 0.3 corresponds to the linear polymer.22 It is remarkable that the k′ value of DPNR decreased from 0.47 to 0.23 after treatment of DPNR-latex with lipase. This indicates that branch-points in the lipase treated-DPNR were entirely decomposed to form linear molecules after lipase treatment. In addition, the lipase treated-DPNR showed lower M h n and M h w values than those of the untreated AE-DPNR and lower polydispersity indices, M h w/M h n. The decrease in the molecular weight is strong supporting evidence showing that the fatty acid ester group of triglycerides and/or phospholipids participates in the branching formation of NR, presumably by micelle formation. The presence of phospholipids will be further reported in the subsequent paper. The alkaline phosphatase was reacted with rubber in a toluene solution to increase the phosphatase efficiency on the decomposition of the R-terminal phosphate groups. It is expected a large effect of the aggregation and entanglement of rubber molecules on the difficulty of phosphate decomposition in latex state. On the other hand, rubber molecules are expected to dissolve in toluene solution and form extended chains, in which phosphatase can easily attack to the R-terminal phosphates. As discussed above, phosphatase hydrolyzes the monophosphate ester linkage, which is
Biomacromolecules, Vol. 6, No. 4, 2005 1855
included in a phospholipid containing a hydroxyl group as a polar headgroup, i.e., PA. As can be seen in Table 1, the phosphatase treatment decreased the molecular weight and [η] value as well as the k′ value of DPNR, although these values were higher than those of TE-DPNR. This indicates the decomposition of branch-points by phosphatase was not complete and also the presence of monophosphate linkage in the branch-points of rubber molecules. In other words, the phosphatase treatment of DPNR may result in the partial scission of rubber chains at the branch-points. This suggests the presence of phosphate groups at the R-termini in rubber molecules. The chain elongation of rubber molecule is believed to proceed by the addition of isopentenyl diphosphate to polyisoprenyl diphosphate.24 If polyisoprenyl phosphate is formed by hydrolysis of polyisoprenyl diphosphate stabilized in latex, it can be the origin of this phosphate linkage in the NR molecule. Based on the assumption mentioned above, at present, two possibilities are considered for the termination of chain elongation reaction in NR synthesis, which can generate functional groups to form branch-points. The first type is the formation of a linkage between phosphate groups including mono- and diphosphate groups of the rubber chain with a phospholipid. The second type is a direct linkage of a phospholipid after dephosphorylation of polyprenyl phosphates. The latter, however, can be neglected due to the ineffectiveness of phosphatase treatment for the decomposition of branch-points. The presence of mono- and diphosphate terminations without phospholipid is also negligible due to the presence of a methylene proton signal of isoprene units linked to mono- and diphosphate groups in the 1H NMR spectrum of phosphatase treated DPNR, which will be described later. The presence of mono- and diphosphate groups as well as phospholipids at the R-termini of rubber molecules was further confirmed by the 1H NMR spectrum of DPNR rubber as illustrated in Figure 6. A sharp peak appears at 3.49 ppm corresponding to the methyl proton next to the nitrogen atom of the choline headgroup of a phospholipid. In addition, two small multiplet signals resonating at 3.92 and 4.04 ppm are expected to be derived from the nonequivalent methylene protons linked to a phosphate group, CH2OP, in the glyceride structure of phospholipids containing nitrogenous and hydroxyl groups. The signals due to phosphate and diphosphate groups were detected in this spectrum in addition to those from phospholipid molecules. A sharp triplet signal at 4.09 ppm having a coupling constant of 6.5 Hz was assignable to C-4 methylene protons of a cis-isoprene unit next to a phosphate group, CH2OP. Besides, a small triplet signal resonating at 4.22 ppm is assignable to the methylene proton linked to a diphosphate group, CH2OPP, although the chemical shifts of these signals are different from those of synthetic polyprenyl diphosphates. It was found that the methylene protons linked to phosphate and diphosphate groups of phosphorylated betulaprenol-18 showed broad triplet signals at 4.41 and 4.46 ppm with coupling constant of 6.6 Hz, measured in CDCl3 and at 27 °C.25,26 The difference of this chemical shift is expected to be due to the difference of measuring conditions and the structure of the
1856
Biomacromolecules, Vol. 6, No. 4, 2005
Tarachiwin et al.
Figure 8. Proposed structure of R-terminal group of NR.
Figure 7. 1H NMR spectra of DPNR (a), lipase treated DPNR (b) and phosphatase treated DPNR (c) after acetone extraction measured in C6D6 at 750 MHz.
protecting group of the phosphate. These findings suggest the presence at least three phosphate functional groups at the R-termini in NR, i.e., monophosphate, diphosphate, and phospholipid groups. Figure 7 shows 1H NMR spectra of DPNR, lipase treated DPNR and phosphatase treated DPNR. In the 1H NMR spectrum of lipase treated DPNR, the phospholipid signal at 3.49 ppm disappeared after lipase treatment, whereas no change was observed for mono- and diphosphate signals. This clearly indicates that mono- and diphosphate groups are directly linked to the rubber chain at the R-terminal group, whereas the phospholipids are present in the rubber as a mixture or aggregated with the R-terminal group. Similarly, phosphatase treatment also gave no significant change in the 1 H NMR spectrum, except for the disappearance of a phospholipid signal at 3.49 ppm. Based on the fact that phosphatase decomposes monophosphate ester linkages as mentioned above, the presence of a monophosphate signal after phosphatase treatment demonstrates that the functional group at the rubber chain end or R-terminal group is not a simple monophosphate group but is modified at the hydroxyl group in the phosphate molecule. Therefore, it may be reasonable to consider that the NR molecule is terminated with mono- or diphosphate groups linked to a phospholipid, which acts to stabilize the phosphate terminal as a protective group. The decrease of the ester content and molecular weight of DPNR after phosphatase treatment suggests the presence of PA as one of the phospholipids linked to the mono- and diphosphate terminating groups. These terminal groups can constitute the branchpoints by formation of a micelle structure or the association of polar headgroups of phospholipid molecule via hydrogen bonding. This assumption can be confirmed by decomposition of the branch-points after the addition of small amounts of polar solvent into rubber solution, which will be reported in a subsequent paper. It is interesting to observe the decrease in the content of long-chain fatty acid esters in DPNR after phosphatase treatment from ∼5.1 to ∼2.1 mol per rubber chain. Ordinarily, however, it is difficult to expect that phosphatase
decomposes phospholipids, because phosphatase cannot hydrolyze phosphate di-esters as mentioned above. Although most free phospholipids are partly soluble either in acetone, methanol or toluene due to their amphiphilic character, phospholipids cannot be separated completely from rubber by ordinary precipitation. In the usual treatment, therefore, phospholipids separated from rubber molecules by phosphatase treatment can remain in DPNR. This fatty acid ester group mixed into DPNR can be counted as the ester content. However, if the branch-points contain PA, the simplest phosphoglyceride, it can be hydrolyzed by phosphatase to give acetone soluble product and contribute to reduction of ester content. Based on the above findings, it can be postulated that the R-terminus is composed of a mono- or diphosphate group linked to a phospholipid as illustrated in Figure 8. The detail analysis of the R-terminal group was further confirmed by phospholipase treatment and chemical analysis, which will be presented in the subsequent paper. Conclusion The presence of acylglyceride and/or phospholipid at the rubber chain-end as an R-terminal group was confirmed by the structural change of the rubber chain after treatment with lipase and phosphatase. The content of fatty acid ester groups and molecular weight of DPNR drastically decreased after lipase and phosphatase treatments. The removal of long-chain fatty acid ester groups, which are linked at the C1 in the phospholipid, by lipase treatment and of phosphoric ester groups by phosphatase treatment resulted in a decrease in molecular weight and the decomposition of branch-points. The decrease in the molecular weight and Huggins’ k′ constant indicated the presence of an acylglycerol compound and a phosphoric monoester linkage the R-terminal group, which form branch-points, probably through micelle formation or by the association of polar headgroups of phospholipid molecules. The 1H NMR spectra of lipase and phosphatase treated DPNR indicated the presence of mono- and diphosphate groups linked to phospholipids at the R-terminus of the rubber molecule. Acknowledgment. This work was supported by a grant from the Thailand Research Fund PHD/0142/2543 and RSA4680009. References and Notes (1) Bristow, G. M.; George, M.; Westall, B. The molecular weight distribution of natural rubber latex. Polymer 1967, 8, 609. (2) Angulo-Sanchez, J. L.; Caballero-Mata, P. Long chain branching in natural HeVea rubber-determination by gel permeation chromatography. Rubber Chem. Technol. 1981, 54, 34. (3) Fuller, K. N. G.; Fulton, W. S. The influence of molecular weight distribution and branching on the relaxation behavior of un-crosslinked natural rubber. Polymer 1990, 31, 609.
Decomposition of Branch-Points (4) Burfield, D. R.; Gan, S. N. Nonoxidative cross-linking reactions in natural rubber. I. Determination of cross-linking groups. J. Polym. Sci. 1975, 13, 2725. (5) Burfield, D. R. Epoxy groups responsible for cross-linking in natural rubber. Nature 1974, 249, 29. (6) Sekhar, B. C. Abnormal groups in rubber and microgel (formation). Proc. Rubber Technol. Conf., 4th 1962, 10. (7) Tangpakdee, J.; Tanaka, Y. Branching in natural rubber. J. Rubber Res. 1998, 1, 14. (8) Tanaka, Y.; Kawahara, S.; Tangpakdee, J. Structural characterization of natural rubber. Kautsch. Gummi Kunst. 1997, 50, 6. (9) Tangpakdee, J.; Tanaka, Y. Characterization of Sol and Gel in HeVea Natural Rubber. Rubber Chem. Technol. 1997, 70, 707. (10) Tangpakdee, J.; Tanaka, Y. Purification of natural rubber. J. Nat. Rubber Res. 1997, 12, 112. (11) Eng, A. H.; Ejiri, S.; Kawahara, S.; Tanaka, Y. Structural characteristics of natural rubber-role of ester groups. J. Appl. Polym. Sci. Appl. Polym. Symp. 1994, 53, 5. (12) Subramaniam, A. Gel Permeation Chromatography of Natural Rubber. Rubber Chem. Technol. 1972, 45, 346. (13) Wren, W. G. Application of the Langmuir through the study of rubber latex. Rubber Chem. Technol. 1942, 15, 107. (14) Dluhy, R. A.; Cameron, D. G.; Mantsch, H. H.; Mendelsohn, R. Fourier transform infrared spectroscopic studies of the effect of calcium ions on phosphatidylserine. Biochemistry 1983, 22, 6318. (15) Wong, P. T. T.; Mantsch, H. H. High-pressure infrared spectroscopic evidence of water binding sites in 1, 2-diacylphospholipids. Chem. Phys. Lipids 1988, 46, 213. (16) Hasma, H.; Subramaniam, A. Composition of Lipids in Latex of HeVea brasiliensis Clone RRIM 501. J. Nat. Rubber Res. 1986, 1, 30. (17) Watanabe, M.; Ohta, A.; Sasaki, S.; Minnikin, D. E. Structure of a New Glycolipid from the Mycobacterium aVium-Mycobacterium intracellulare complex. J. Bacteriol. 1999, 181, 2293.
Biomacromolecules, Vol. 6, No. 4, 2005 1857 (18) Stone, K. J.; Wellburn, A. R.; Henning, F. W.; Pennock, J. F. The characterization of ficaprenol-10, -11 and -12 from the leaves of Ficus elastica (decorative plant). Biochem. J. 1967, 102, 325. (19) Tanaka, Y.; Kawahara, S.; Eng, A. H.; Shiba, K.; Ohya, N. Initiation of biosynthesis in cis-polyisoprene. Phytochemistry 1995, 39, 779. (20) Tanaka, Y.; Mori, M.; Ute, K.; Hatada, K. Structure and Biosynthesis Mechanism of Rubber from Fungi. Rubber Chem. Technol. 1990, 63, 1. (21) Tanaka, Y.; Eng, A. H.; Ohya, N.; Nishiyama, N.; Tangpakdee, J.; Kawahara, S.; Wititsuwannakul, R. Initiation of Rubber Biosynthesis in HeVea brasiliensis: Characterization of Initiating Species by Structural Analysis. Phytochemistry 1996, 41, 1501. (22) Bristow, G. M. Huggins k′ parameter for polyisoprenes. J. Polym. Sci. 1962, 62, S168. (23) Amber, M. R.; Mate, R. D.; Purdon, J. R. A basis for the determination of branching in polymers by use of gel-permeation chromatography. J. Polym. Sci. Polym. Chem. Ed. 1974, 12, 1759. (24) Lynen, F.; Henning, U. Biological path to natural rubber. Angew. Chem. 1960, 72, 820. (25) Sakdapipanich, J. T.; Mekkriengkrai, D.; Gou, L. B.; Tanaka, Y. Structural Characterization of the Terminating Groups of some Polyprenyl Phosphates and of Rubber from the Lactarius Mushroom. J. Rubber Res. 2001, 4, 118. (26) Mekkriengkrai, D.; Ute, K.; Swiezewska, E.; Chojnacki, T.; Tanaka, Y.; Sakdapipanich, J. Structural Characterization of Rubber from Jackfruit and Euphorbia as a Model of Natural Rubber. Biomacromolecules 2004, 5, 2013.
BM058003X
