Evaluation of Hyaluronan from Different Sources: Streptococcus z

Aviva Shiedlin,*,† Russell Bigelow,† William Christopher,† Saman Arbabi,‡ ... Ronald V. Maier,‡ Norman Wainwright,§ Alice Childs,§ and Rob...
0 downloads 0 Views 173KB Size
Biomacromolecules 2004, 5, 2122-2127

2122

Evaluation of Hyaluronan from Different Sources: Streptococcus zooepidemicus, Rooster Comb, Bovine Vitreous, and Human Umbilical Cord Aviva Shiedlin,*,† Russell Bigelow,† William Christopher,† Saman Arbabi,‡ Laura Yang,† Ronald V. Maier,‡ Norman Wainwright,§ Alice Childs,§ and Robert J. Miller† Genzyme Corporation, Cambridge, Massachusetts 02139, Department of Surgery, University of Washington School of Medicine, Seattle, Washington, and Marine Biological Laboratory, Woods Hole, Massachusetts Received March 17, 2004; Revised Manuscript Received August 6, 2004

Sodium hyaluronate (HA) is widely distributed in extracellular matrixes and can play a role in orchestrating cell function. Consequently, many investigators have looked at the effect of exogenous HA on cell behavior in vitro. HA can be isolated from several sources (e.g., bacterial, rooster comb, umbilical cord) and therefore can possess diverse impurities. This current study compares the measured impurities and the differences in biological activity between HA preparations from these sources. It was demonstrated that nucleic acid and protein content was highest in human umbilical cord and bovine vitreous HA and was low in bacterial and rooster comb HA. Macrophages exposed to human umbilical cord HA produced significantly higher amounts of TNF-R relative to control or bacterial-derived HA. These results indicate that the source of HA should be considered due to differences in the amounts and types of contaminants that could lead to widely different behaviors in vitro and in vivo. Introduction Sodium hyaluronate or hyaluronan (HA) is widely distributed in extracellular matrixes1,2 and has been shown to affect cell migration,3-5 cell adhesion,6,7and angiogenesis.8-10 Because of the diverse role HA can play in orchestrating cell function,12-13 many investigators have looked at the effect of exogenous HA on cell behavior in vitro.14-16 HA can be isolated from several sources (e.g., Streptococcus zooepidemicus, rooster comb, bovine vitreous humor, and human umbilical cord)17-22 and therefore can possess diverse, species-dependent impurities. This study compares the measured impurities and differences in biological activity between HA preparations from various sources.

Figure 1. Calibration plot of RNA versus integrated peak area at 257-nm absorbance. [RNA] is the concentration in the working standard (before acid treatment).

Materials and Methods Materials. HA powders isolated from Streptococcus zooepidemicus, rooster comb, bovine vitreous, and human umbilical cord were purchased from Sigma Chemical Company. Additional bacterially derived HA was purified from the fermentation of Streptococcus zooepidemicus at Genzyme Corporation. Molecular Weight Measurement. HA molecular weight was determined with a size exclusion chromatography/ multiangle laser light scattering (SEC/MALLS) GPC system.20 The GPC system consisted of a pump (Pharmacia LKB 2150), columns (Bondagel E- High & E-500), and two * To whom correspondence should be addressed. [email protected]. † Genzyme Corporation. ‡ University of Washington School of Medicine. § Marine Biological Laboratory.

Figure 2. Calibration plot of DNA versus integrated peak area at 257-nm absorbance. [DNA] is the concentration in the working standard (before acid treatment).

E-mail:

detectors in tandem: Light Scattering Detector (DAWN DSP by Wyatt Technology) and Refractive Index Detector (Hewlett-Packard 1047A). Sample injection was performed

10.1021/bm0498427 CCC: $27.50 © 2004 American Chemical Society Published on Web 10/15/2004

Biomacromolecules, Vol. 5, No. 6, 2004 2123

Evaluation of Hyaluronan from Different Sources

Figure 3. Production of TNF-R from rabbit alveolar macrophages in different HA preparations. Table 1. Measurements of Molecular Weight, Polydispersity, Endotoxin, Protein, and Nucleotides for HA (per 1 mg HA) Preparations from Different Sources

HA M.W (x 106 Da) PI (Mw/Mn) Endotoxin (EU/mg HA) Total Protein (µg/mg HA) [RNA] (µg/mg HA) [DNA] (µg/mg HA) a

human umbilical cord a

bacterially derived (Genzyme)

bacterially derived (Sigma)

rooster comb

bovine vitreous

1.3 ( 0.1 1.3 ( 0.1 >100 47.7 ( 3 6.7 ( 0.1 16.8 ( 4.5

1.6 1.1 100 36.2 1.9 1.1

Three lots were tested. b BDL ) 1(µg/mg HA).

by an autosampler (Waters) with 50 µL injection volume. The mobile phase was an aqueous solution of sodium sulfate (0.15 M, pH ) 4.5) at a flow rate of 0.3 mL/min. Data from light scattering and refractive index detectors were collected and processed with ASTRA 4.70 software (Wyatt Technology), and weight-average molecular weight (Mw), number-average molecular weight (Mn), and polydispersity (Mw/Mn) were calculated. Polydispersity measurements are subject to system limitations. The accuracy of the measurements depends on column resolution; calculated Mn values will be somewhat higher than true values because of imperfect resolution. This overestimation will lead to an underestimation of polydispersity. Macrophage Activation Assay. Rabbit alveolar macrophages were obtained by bronchoalveolar lavage and adhered to tissue culture plates or to tissue culture plates previously coated with 0.4% (aq.) HA.21 After 2 h, macrophages were stimulated with endotoxin (lipopolysaccharide [LPS]) to induce a proinflammatory phenotype. The supernatant was harvested at 24 h and assayed for TNF-R using an L929 mouse fibroblast cytotoxicity bioassay. Nucleic Acid Content in HA Powders. A reverse phase HPLC assay was used to detect the amount of DNA and RNA in the HA samples. This assay determines the amount of guanine that is liberated after acidic hydrolysis of the samples.22 Guanine is liberated from DNA more readily than RNA, allowing independent quantitation of these polynucleotides. RNA (Baker’s Yeast, Sigma R71256) and DNA (Herring Testes, Sigma D6898) were used as standards. Free guanine is quantified by HPLC with detection at 257 nm.

The HPLC procedure uses a column (Brownlee, Reverse Phase Spheris ODS 220 × 4.6 mm), a pump (SP8800, Spectra-Physics), and a UV-vis detector (Spectra 200, Spectra-Physics). Sample injection was preformed by autosampler (SL 9, Shimazu) with a 100 µL injection volume at a flow rate of 1 mL/min. gradient table time [min]

A%

B%

0.00 10.00 12.00 13.00

0 40 40 0

100 60 60 0

Mobile phase: A ) 0.01 M NaH2PO4, 80% MeOH; B ) 0.01 M NaH2PO4, pH 5.5 Sample Preparation. An equal amount of 6 N HCl (RNA condition) or 0.06 N HCl (DNA condition) was added to sample or standard, followed by heating in a boiling waterbath for 5 min. Samples were then diluted 1:12 with mobile phase B before injection onto the HPLC. Figures 1 and 2 are calibration plots of RNA and DNA. Bacterial Endotoxin Test. LPS levels were measured using the kinetic turbidimetric method of Limulus Amoebocyte Lysate (LAL) by Marine Biological Laboratory, Woods Hole, MA. Purification/Analysis of Proteins in HA Powders. Total protein amounts were determined by the Coomassie Blue Protein Assay (Pierce). The contaminating proteins were separated from the HA in the human umbilical cord preparation by ultrafiltration (500 000 MWCO, AG Technology)

2124

Biomacromolecules, Vol. 5, No. 6, 2004

Shiedlin et al.

Figure 4. HPLC tracings for RNA from human umbilical HA and bacterially derived HA.

Figure 5. HPLC tracings for DNA from human umbilical HA and bacterially derived HA.

followed by concentration of the retentate using a second ultrafiltration step (3000 MWCO, AG Technology) then analyzed by SDS-PAGE. A major band, at around 65 kDa, was removed from the gel by dry-blot transfer to a PVDF membrane. This protein band was digested with trypsin and subjected to sequence analysis by MALDI (University Mass. Med. School, Worcester, MA). Additional separation of contaminating proteins in the human umbilical cord HA was performed by digestion of

the HA solution with hyaluronidase (Sigma) and concentration of the solution was performed by ultrafiltration (5000 MWCO, Vivascience). The remaining solution was loaded to SDS-PAGE (Figure 7). Results and Discussion Sodium Hyaluronate Modulates Macrophage Activation. The amount of TNF-R produced by alveolar macrophage cultures as a function of HA sources and added endo-

Biomacromolecules, Vol. 5, No. 6, 2004 2125

Evaluation of Hyaluronan from Different Sources

Figure 6. HPLC tracings for the three lots of human umbilical cord derived HA treated with DNA conditions. Table 2. Results of the MS-Fit Search peptide m/z 509.6 649.8 665.8 689.8 712.8 928.3 1164.5 1250.7 1284.9 1306.9 1441.1 1481.2 1569.2 1569.2 1641.5 Figure 7. SDS-PAGE gel of protein isolated from the umbilical cord HA. Band at 65 kDa was used for protein sequencing analysis.

toxins are compared in Figure 3. In the absence of endotoxins, there is no detectable TNF-R in the culture media of the control or bacterially derived HA. Macrophages exposed to the HA from human umbilical cord produced a significantly larger amount of TNF-R relative to control or bacterially derived HA. The addition of endotoxin (10 ng/mL) results in an increase in the TNF-R production in all macrophage cultures, which is an indication that the cells are still capable of producing this cytokine. TNF-R in the culture media of the control or bacterial-derived HA are almost identical. TNF-R level of umbilical cord HA is significantly higher. At the highest endotoxin dose (100 ng/mL), the macrophages treated with bacterially derived HA produced significantly less TNF-R than the control macrophages. This suggested a possibility that TNF-R production is suppressed by bacterially derived HA (umbilical cord HA was not tested at this level).

BSA m/z

delta %

509.6 -0.0062 649.7 0. 0015 665.8 -0.0024 689.8 0.0007 712.8 0.0019 928.1 0.0238 1164.3 0.0130 1250.4 0.0239 1284.5 0.0306 1306.5 0.0299 1440.7 0.0281 1480.7 0.0329 1568.7 0.0299 1569.7 -0.0316 1640.9 0.0249

start end 558 205 156 236 29 161 66 35 361 402 360 421 347 387 437

561 209 160 241 34 167 75 44 371 412 371 433 351 399 451

serum albumin peptide sequence (K)HKPK(A) (K)IETMR(E) (K)KFWGK(Y) (K)AWSVVAR(L) (K)SEIAHR(F) (K)YLYEIAR(R) (K)LVNELTEFAK(T) (R)FKDLGEEHFK(G) (R)HPEYAVSVLLR(L) (K)HLVDEPONLIK(Q) (R)RHPEYAVSVLLR(L) (K)LGEYGFQNALIVR(Y) (K)DAFLGSFLYEYSR(R) (K)DDPHACYSTVFDK(L) (R)KVPQVSTPTLVEDK(L)

These disparate results, with two different HA sources, indicate that one can get contrary results with supposedly the same material and that biological activity may be influenced by the presence or absence of impurities. Polymer Molecular Weight and Polydispersity. Table 1 compares the differences in polymer molecular weight and polydispersity for all HA preparations. The polymer molecular weight and polydispersity were similar for most HA samples, except for the bovine preparation, which had significantly lower molecular weight (0.4 × 106 Da) and higher polydispersity (1.9). Bacterial Endotoxin Test. Table 1 compares the differences in endotoxin contaminants for all HA preparations. The human umbilical cord HA, bovine vitreous and rooster comb preparations contained a high level of endotoxin contaminant. The bacterially derived HA from both suppliers was nearly endotoxin-free. Nucleic Acid Content in HA Powders. Table 1 compares the differences in nucleotide contamination for all HA

2126

Biomacromolecules, Vol. 5, No. 6, 2004

Shiedlin et al.

Figure 8. MALDI analysis of the tryptic fragments from 500 to 1800 mass-to-charge ratio.

Figure 9. MALDI analysis of the tryptic fragments from 50 to 600 mass-to-charge ratio.

sources. There are significant differences in the amounts of nucleotides among different HA samples. Human umbilical cord HA and bovine vitreous HA preparations contained DNA and RNA. Rooster comb and bacterially derived HA showed DNA and RNA levels that were below the level of detection. Figures 4 and 5 show overlays of HPLC tracings for equal amounts of human umbilical cord and bacterially

derived HA (Genzyme) treated with RNA (6 N HCl) and DNA (0.06 N HCl) conditions, respectively. There were also lot-to-lot variations between the lots from the same source. Three lots of human umbilical cord HA were analyzed for DNA and RNA content by HPLC. Two lots showed a distinct shoulder on the guanine peak when analyzed under the DNA conditions (Figure 6).

Evaluation of Hyaluronan from Different Sources

Biomacromolecules, Vol. 5, No. 6, 2004 2127

Table 3. Result Summary of MS Fragment Search for (K)YLYEIAR(R) Sequence

natures of contamination. These results indicate that the impurities in HA preparations could lead to widely different behavior of these HA samples in vitro and in vivo experiments.

unmatched ions

MH + calculated (Da)

MH+ error (Da)

protein MW(Da)/pI

species

0/9 0/9 0/9

928.0795 928.0795 928.0795

0.4305 0.4305 0.4305

69411.2/5.92 69293.9/5.82 69366.2/6.05

porcine bovine human

Purification/Analysis of Proteins in HA Powders. All preparations, except the Streptococcus zooepidemicus HA purchased from Sigma, contain some amount of protein (Table 1), but human umbilical cord HA and bovine vitreous HA have by far more protein. Because the human umbilical cord contained the most protein, it was decided to identify the contaminating proteins in these preparations. The major protein contaminant from the human umbilical cord HA was a protein with a molecular weight of 65 kDa (Figure 7). This protein was isolated by SDS-PAGE and dry blot transferred to PVDF. This isolate was subjected to tryptic digest followed by MALDI analysis of the tryptic fragments. Table 2 and Figure 8 summarize the masses of these tryptic fragments, which are consistent with the expected tryptic fragments for bovine serum albumin. The fragmentation pattern of the 928 m/z parent mass peptide was compared to the known fragmentation patterns of human, bovine and porcine serum albumin. These fragments matched BSA and confirmed that the contaminating protein was serum albumin. Results are shown in Table 3 and Figure 9. Conclusions Two HA samples, one from bacterial source and other from human umbilical cord, gave disparate results in an in vitro test with macrophages. The two HA preparations had similar molecular weights; however, the two preparations differed significantly in amounts of contaminating protein, endotoxin, and nucleotides. Moreover, a comparison of HA from different sources or from the same source made by different companies, showed different levels and different

References and Notes (1) Patti, A. M.; Gabriele, A.; Vulcano, A.; Ramieri, M. T.; Della Rocca, C. Tissue Cell 2001, 33 (3), 294-300. (2) Iocono, J. A.; Bisignani, G. J.; Krummel, T. M.; Ehrlich, H. J. Surg. Res. 1998, 76 (2), 111-6. (3) Turley, E. A.; Torrance, J. Exp. Cell Res. 1985, 161 (1), 17-28. (4) Knudson, C. B.; Knudson, W. FASEB J. 1993, 7 (13), 1233-41. (5) Laurent, T. C.; Fraser, J.; Robert, E. FASEB J. 1992, 6 (7), 2397404. (6) Knudson, C. B. In The Chemistry, Biology and Medical Application of Hyaluronan and its DeriVatiVe; Laurent, T. C., Ed.; Portland Press: London, 1998; pp 161-168. (7) West, D. C.; Hampson, I. N.; Arnold, F.; Kuma, S. Science 1985, 228 (4705), 1324-6. (8) West, D. C.; Kumar, S. Ciba Foundation Symp. 1989, 143, 187201; discussion 201-7, 281-5. (9) Rooney, P.; Kumar, S.; Ponting, J.; Wang, M. Int. J. Cancer 1995, 60 (5), 632-6. (10) Balazs, E. A. In The Chemistry, Biology and Medical Application of Hyaluronan and its DeriVatiVe; Laurent, T. C., Ed.; Portland Press: London, 1998; pp 185-204. (11) Taylor, K. R.; Trowbridge, J. M.; Rudisill, A.; Termeer, C. C.; Simon, J. C.; Gallo, R. L. J. Biol. Chem. 2004, 279 (17), 17079-84. (12) Ehlers, E. M.; Behrens, P.; Wunsch, L.; Kuhnel, W.; Russlies, M. Ann. Anat. 2001, 183 (1), 13-17. (13) Velten, F.; Laue, C.; Schrezenmeir, J. Biomaterials 1999, 20 (22), 2161-2167. (14) Knoflach, A.; Azuma, H.; Magee, C.; Denton, M.; Murphy, B.; Iyengar, A.; Buelow, R.; Sayegh, M. H. J. Am. Soc. Nephrol. 1999, 10 (5), 1059-1066. (15) Termeer, C. C.; Hennies, J.; Voith, U.; Ahrens, T.; Weiss, J. M.; Prehm, P.; Simon, C. J. Immunology 2000, 165 (4), 1863-70. (16) Toole, B. P.; Jackson, G.; Gross, J. Proc. Natl. Acad. Sci. U.S.A. 1972, 69 (6), 1384-6. (17) Kaye, M. A. G. Nature 1950, 166 (4220), 478-9. (18) Balazs, E. A. Fed. Proc. 1958, 17, 1086-93. (19) Laurent, T. C.; Ryan, M.; Pietruszkiewicz, A. Biochim. Biophys. Acta 1960, 42, 476-85. (20) Yu, L. P.; Burns, J. W.; Shiedlin, A.; Guo, Y.; Jankowski, T.; Pradipasena, P.; Rha, C. Harnessing Biotechnol. 21st Century, Proc. Int. Biotechnol. Symp. Expo. 9th 1992, 80-4. (21) Arbabi, S.; Rosengart, M. R.; Garcia, I.; Maier, R. V. Surg. Forum 1999, 50, 487-489. (22) Nakahara, T.; Shiraishi, A.; Hirano M.; Matsumoto, T.; Kuroki, T.; Tatebayashi, Y.; Tsutsumi, T.; Nishiyama, K.; Ooboshi H.; Nakamura, K. Anal. Biochem. 1989, 180 (1), 38-42.

BM0498427