Increase in Lens Gangliosides Due to Aging and Cataract Progression in Human Senile Cataract

Increase in Lens Gangliosides due to Aging and Cataract Progression in Human Senile Cataract

Manabu Ogiso,* Nobuyuki Sairo,t Kenji 5udo,-(- Hideo Kubo4 Shusuke Hirano,* and Michiji Komorot

Gangliosides were isolated from human senile cataractous lenses by solvent extraction, DEAE-Sepha- dex column chromatography, and thin-layer chromatography. The content and composition of gangliosides were examined in individual lens tissues. Three predominant gangliosides, GM3, GM1, and GDI a, were tentatively identified in comparison with authentic brain gangliosides, and several unidentified gangliosides were also recognized. The increase in ganglioside content per mg of protein content in cataractous lenses was found to be influenced by two physiologic parameters: aging and cataract progression. The mature cataractous lenses showed a higher ganglioside level on a protein basis than the immature lenses compared with the same age group. On the basis of statistical analysis, an age-dependent increase in ganglioside concentration was recognized in both mature and immature lens groups. The relative increase in slow-moving polysialogangliosides on thin-layer chromatography seemed to be caused by the maturation of cataract. The sugar composition of one of the polysialogangliosides was found to be glucose, galactose, and sialic acid in the molar ratio of 2:1:4; this suggests the presence of a unique ganglioside species in human cataractous lens. Invest Ophthalmol Vis Sci 31:2171-2179,1990

Ocular gangliosides in lens tissue were first re- chromatography (GLC). In a subsequent study, Tao ported by Feldman and Feldman1 and have been and Lee8 identified the neuraminic acid moiety of characterized in cow, pig, rat,23 and rabbit4 lenses. these gangliosides as N-acetylneuraminic acid. More The human lens is reported to have a total ganglio- recently, Swindell et al,6 also using TLC, tentatively side concentration 4-11 times higher than cow, pig, identified GM3, GM1, and GDI a as major ganglio- or rat lenses.2 Various clinical studies have also been sides and demonstrated the presence of several lesser- done on gangliosides in human senile cataractous intensity resorcinol-positive bands in cataractous nu- lenses.'"3i5~7 Windeler and Feldman3 and Sarkar and clei. No statistically significant differences in the Cenedella2 reported the presence of only two ganglio- three major gangliosides were observed between cata- sides, GM3 and GM1, on the basis of thin-layer chro- ractous nuclei and normal or cataractous whole matographic (TLC) analysis. No significant differ- lenses. However, some minor but possibly significant ences were observed in the content or composition of differences were found in band intensities by densito- gangliosides from normal and cataractous lenses. metric scanning. At present, no conclusive evidence In a more elabolate study using TLC analysis, Tao has been obtained to demonstrate whether a more et al7 found that the human cataractous lens showed a complex ganglioside pattern is implicated in the complex ganglioside pattern containing 11 resor- human cataractogenic process. cinol-positive bands. The presence of several fucose- Recently, we found an age-dependent increase in and N-acetylglucosamine-containing ganglioside lenticular gangliosides in normal lenses from mon- species was shown by combined TLC and gas-liquid keys between the ages of 6-15 yr.9 The ganglioside content in individual clear lenses of rhesus monkeys increased with age, and polysialogangliosides ap- From the *Deparrtments of Physiology and fOphthalmology, peared in aged lenses. The total gangliosidic sialic Toho University School of Medicine, and the ^Department of Life Science, Faculty of Science, Tokyo Institute of Technology, Tokyo, acid content of the monkey lenses was more than one Japan. half that of the human lens. The life span of the Supported in part by Grants-in-Aid for Scientific Research monkey is reported to be about 15 yr, with 1 yr in the (No.01771443) from the Ministry of Education, Science and Cul- life of the monkey equivalent to approximately 4.5 yr ture of Japan and for Project Research (No. 1-32) from Toho Uni- in a human life.10 Thus, monkey lenses provided versity School of Medicine. Reprint requests: Manabu Ogiso, Department of Physiology, more comprehensive information on changes in len- Toho University School of Medicine, Ohmori-nishi 5-21-16, ticular gangliosides due to aging over a long life span. Ohta-ku, Tokyo 143, Japan. In the current study, quantitative and qualitative

2171

Downloaded from iovs.arvojournals.org on 10/01/2021 2172 INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / Ocrober 1990 Vol. 31

changes in lens gangliosides of the human senile cata- ner (Shimadzu, Kyoto, Japan).16 Two-dimensional ract were explored taking account of the age-depen- TLC was done according to the method of Ohashi.17 dent increase. Ganglioside content increased accord- The solvent systems were chloroform/methanol/ ing to two physiologic parameters: aging, as pre- 0.02% CaCl2 (60:35:8, v/v) as the first dimension and viously described in the monkey lens,9 and l-propanol/28% ammonia water/water (75:5:25, v/v) cataractous progression leading to lens opacification. as the second. The ganglioside nomenclature used The appearance of slow-moving polysialoganglio- here is that of Svennerholm.18 sides on TLC was occasionally accompanied by the maturation of senile cataracts. In addition, one of Ganglioside Mapping them was estimated to have a characteristic sugar composition by GLC analysis. Preliminary reports of To purify ganglioside species, crude ganglioside these studies have been presented.'' fractions underwent ganglioside mapping by high- performance liquid chromatography (HPLC) with an DEAE-Iatrobeads column (4.6 X 250 mm; Iatron, Materials and Methods Tokyo, Japan).19-20 The gangliosides were applied to a Human Cataractous Lenses column equilibrated with methanol and separated with a programmed gradient of ammonium acetate Human senile cataractous lenses from patients (40-50 raM) in methanol at the flow rate of 1 ml/ aged 37-84 yr were obtained from the operating min. One-milliliter fractions were collected, and an room of Toho University Hospital after cataract sur- aliquot of each fraction was analyzed by TLC. The gery and kept frozen at -80°C until use. In some polysialoganglioside fraction was collected and puri- experiments, the nuclear and cortical regions were fied by preparative TLC on a HP-TLC plate. separated under a binocular. The cortical region contained capsular and epithelial cells. GLC of Gangliosides Extraction and Quantitation of Gangliosides The GLC was used to identify the products obtained after methanolysis of the recovered polysialo- Gangliosides were extracted from individual lenses ganglioside. An aliquot of the ganglioside fraction using organic solvent extraction as described else- 9 was treated with 1 M of anhydrous methanolic hy- where. The ganglioside fraction was separated from drochloric acid at 80°C for 16 hr. The glycosides of total lipid extract by DEAE-Sephadex (Pharmacia- liberated hexosamines and methylneuraminate were LKB, A-25, acetate form; Uppsala, Sweden) column 12 converted to N-acetyl derivatives according to the chromatography, and partially purified by a re- method of Kozulic et al.21 Fatty acid methyl esters verse-phase SEP-PAK C18 cartridge (Waters Asso- 13 were recovered from the acid methanolysate by hex- ciates, Millford, MA) after mild alkaline hydrolysis. ane extraction and analyzed isothermally by a Shi- Total gangliosidic sialic acid content was estimated as madzu gas chromatograph GC-7A at 200 and 250°C the content of N-acetylneuraminic acid (NANA) on a fused silica capillary column of ULBON using the thiobarbituric acid assay described by War- 14 HR-101 (0.24 mm X 25 m; Chromato Packings ren. The values calibrated by Warren's equation Center, Tokyo, Japan). Methylglycosides were ana- were corrected against the contents of ganglioside lyzed as their trimethylsilyl (TMS) derivatives on the standards used. Protein content was determined same column. The column temperature was pro- using bovine serum albumin as a standard according 15 grammed to increase from 170-230°C at 2°C/min. to the method of Lowry et al. Results TLC of Gangliosides Total Gangliosidic Sialic Acid Content Samples corresponding to one-tenth each of the in Human Senile Cataractous Lenses ganglioside fraction were developed using a precoated high-performance TLC (HP-TLC) silica gel 60 plate Ganglioside fractions from 13 individual whole (5641; Merck, Darmstadt, FRG, Rahway, NJ) in a cataractous lenses obtained by intracapsular cataract solvent system of chloroform/methanol/0.25% extraction were prepared by the successive steps of CaCl2 solution (55:45:10, v/v). The ganglioside spots solvent extraction, DEAE-Sephadex column chro- were visualized by heating at 95 °C after spraying with matography, and SEP-PAK Ci8 cartridge purifica- resorcinol HC1 reagent. The intensity of spots was tion. The cataractous lenses were obtained from pa- quantitated by densitometric scanning of the TLC tients ranging in age from 49-84 yr. An aliquot of plate at 580 nm with a Shimadzu CS-930 TLC scan- each sample was analyzed for total NANA content

Downloaded from iovs.arvojournals.org on 10/01/2021 No. 10 GANGLIO5IDE5 IN HUMAN SENILE CATARACTS / Ogiso er ol 2173

Table 1. Relationship between lens ganglioside of water content in the specimens kept at -80°C and content and progression of human senile cataracts determined in protein residue after solvent extraction. Sialic Age Sialic acid Protein acid/protein (years) Stage (nmoles) (mg) (pmoles/tng) TLC Analysis of Lens Gangliosides 49* incipient 40.1 51.2 783 60 incipient 48.0 52.6 913 An aliquot corresponding to one-tenth of the gan- 79 incipient 43.5 56.5 770 72 immature 33.8 50.8 665 glioside fraction from each individual lens was ap- 73 immature 40.9 51.5 795 plied to TLC analysis using an HP-TLC plate (Fig. 1). 74 immature 41.1 50.8 810 Several resorcinol-positive bands were revealed in all 78 immature 54.3 53.2 1020 81 immature 69.6 54.0 1288 the fractions from cataractous whole lenses. Three 84 immature 62.3 55.6 1120 predominant gangliosides, GM3, GM1, and GDla, 64 mature 46.8 48.2 970 were tentatively identified in comparison with the 72 mature 67.6 53.3 1268 73 mature 48.5 53.4 908 chromatographic behavior of authentic brain gangli- 74 mature 68.5 50.0 1370 osides. Lesser amounts of minor components were recognized in different samples but were not identi- * Diabetic cataract. The ganglioside content in individual lenses was determined in duplicate fied. Slow-moving gangliosides, designated as lower as the amount of NANA liberated from gangliosides in 50 mM H2SO4 at than GDla in the text, were observed, especially in 80°C for I hour. Protein content was determined using the protein residue after solvent extraction. The data arc arranged according to cataractous stage mature cataracts. Two-dimensional TLC provided of the lens and by age within stages. more comprehensive information on the characteris- tics of the ganglioside species (Fig. 2). The mixing using the thiobarbituric acid procedure of Warren.14 experiment (Fig. 2B) indicated that the predominant Average NANA content and NANA content per mg species in the slow-moving gangliosides did not co- of lens protein showed wide distributions, ranging migrate with authentic brain GDlb and GTlb, sug- from 33.8-69.6 pmol and from 665-1370 pmol/mg, gesting a different structure for polysialogangliosides. respectively (Table 1). The concentration of ganglio- Estimates of the relative amounts of each gangliosides in mature cataracts (1129 ± 112 pmol/mg) was side species were made by the densitometric scanning not significantly different from that in immature cat- method of Ando et al.16 Apparent differences in gan- aracts (950 ± 95 pmol/mg) {P > 0.05). In this study, glioside composition between immature and mature protein content was used to compensate for the loss cataracts were indicated only by the appearance of

Fig. 1. HP-TLC separation of gangliosides from lenses of 49- to 84-year-old cataraclous patients. An aliquot corresponding to one- tenth of individual lens gan- — GMa gliosides was developed in a solvent system of chloro- |tvtff-t form/methanol/O.25% CaCl2 {55:45:10, v/v). The gangliosides were visualized by resorcinol-HCI spray. The arrow indicates the slow- moving ganglioside designated in the text. IN = incipient; IM = immature; M = mature cataract. 72 74 72 73 64 79 73 [Age]

M |M IM IM M IN M [stage]

Downloaded from iovs.arvojournals.org on 10/01/2021 2174 INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / Ocrober 1990 Vol. 01

•#«

CNJ I 1st Fig. 2. Two-dimensional HP-TLC profiles of lens gangliosides. (A) Gangliosides from the mature cataractous lens of a 72-year-old patient were developed twice in solvent systems of chloroform/methanol/0.02% CaCl2 (60:35:8, v/v) (first dimension) and once l-propanol/28% ammonia water/water (75:5:25, v/v) (second dimension). (B) A mixture of lens gangliosides and authentic brain gangliosides, GM1 (a), GDla (b), GDlb (c) and GTIb (d), was developed.

slow-moving gangliosides, and no differences were normal monkey lenses has been expressed in terms of found in the peak area ratio of mono- to disialogan- human age and inserted in Figure 3A (dotted line). gliosides (data not shown). There was a significant difference between mature and immature groups (P < 0.01) when the ganglio- Relationship Between Ganglioside Content side concentration were adjusted to those of 60-year- and Lens Opacification old humans, according to the corresponding regression lines (Fig. 3B). Human cataractous lenses were classified into three phases, incipient, immature, and mature, according to the location and degree of opacity, as de- Distribution of Gangliosides in Cataractous Lenses termined by slit-lamp observation. The incipient, To examine the increase in gangliosides, whole immature, and mature groups roughly corresponded cataractous lenses were carefully separated into nu- to Groups I, I-II, and II-IV described by Pirie.22 When clear and cortical regions. Mature lenses appeared to the data on lens gangliosides were arranged according accumulate gangliosides in the cortical region rather to this classification, higher values of ganglioside than in the nuclear region (Table 2). In senile cata- content expressed per mg of protein were seen in ma- racts, the ratio of ganglioside content in the cortical to ture cataractous lenses from younger patients than in that in the nuclear region was greater than 1 with the immature lenses (Table 1). The close relationship be- single exception of an 82-year-old immature cataract tween aging and lens opacification was studied by lens. When the total ganglioside content of each re- plotting the ganglioside values as a function of age gion was expressed per mg of protein, the nuclear (Fig. 3A). For statistical comparison, an additional 16 region showed lower values than the cortical region. specimens were examined. In both groups, the corre- In diabetic patients, enrichment of gangliosides was lation coefficients (r) between ganglioside content observed in the nuclear region when nuclear opacity and aging were calculated to be 0.73 (P < 0.01) for was produced. A significant difference (P < 0.05) be- mature lenses (n = 13) and 0.87 (P < 0.01) for im- tween senile and diabetic cataracts was only found in mature lenses (n = 12). Diabetic cataractous lenses the ratio of ganglioside content in the cortical region from 37-, 49-, and 55-year-old patients had ganglio- to that in the nuclear region. The age-dependent in- side contents within the range of the correlation. A crease of gangliosides in each region was not exam- similar correlation (r = 0.88, P < 0.01) observed in ined because of the small numbers of specimens.

Downloaded from iovs.arvojournals.org on 10/01/2021 No. 10 GANGLIOSIDES IN HUMAN SENILE CATARACTS / Ogiso er ol 2175

Fig. 3. Relationship between ganglioside content and aging in whole catarac- J tous lenses. (A) Ganglioside (x10~ ) content expressed as NANA 2.0 per mg protein was plotted against age. Regression lines 1.5 " B were calculated for immature (open circles) and ma- r=0.74 ture (closed circles) catarac- Y=23.5X-559.6 tous lenses. The regression line calculated for normal monkey lenses9 and ad- 1.0 justed for age difference has O been inserted as a dotted r=0.67 V=30.* X-13 85.1 line. Asterisks indicate diabetic cataractous lenses, x = incipient. (B) The values were calculated to be equiv- 8 0.5 alent to those at age 60 years A)—---- 0 on the basis of the ratio to «*> ft the regression lines (method 1) and the proportional change in content obtained from the regression lines (method 2). Dotted lines in- 37 50 60 70 80 dicate the equivalent values METHOD 1 METHOD 2 AGE (years) for age 60. Immature (open circles) and mature (closed circles) cataractous lenses.

Qualitative changes in ganglioside composition were Properties of Slow-Moving Gangliosides ascertained by HP-TLC, and band intensities corresponding to ganglioside species were compared be- Slow-moving gangliosides observed in mature cata- tween nuclear and cortical regions (Fig. 4). No re- racts in high proportions were separated by ganglio- gional differences were seen. Slow-moving ganglioside mapping using HPLC with a DEAE-Iatrobeads sides were recognized in both regions and in both column before compositional analysis by GLC. Fig- mature and immature cataract lenses, when the gan- ure 5 presents a map of lens gangliosides from catar- glioside samples were applied at 3-4 nmol NANA actous lenses. More than 11 resorcinol-positive bands equivalent. were recognized on HP-TLC separation, and slow-

Table 2. Distribution of gangliosides in cataractous lenses Opaque sites Age (years) Stage as ps cor n Cortex (A) Nucleus (B) A/B

54* immature + 28.6 (998) 54.9(1,651) 0.52 56* incipient + 20.0 (579) 24.5 (548) 0.82 62 immature + 26.7 (833) 10.3 (469) 2.59 64* immature + 8.4 (665) 58.3(1,475) 0.14 66* incipient 77.9 (2,273) 11.2(179) 6.96 66 immature 101.8(2,436) 13.8(549) 7.38 66 mature 93.5 (3,638) 47.5(1,339) 1.97 66 mature + 72.1 (2,503) 52.0(1,505) 1.39 70 mature + 46.5(1,495) 35.3(1,193) 1.32 73 immature + 44.5(1,426) 19.2(617) 2.32 75 immature + 47.4(1,342) 27.1(938) 1.75 80* immature + 55.3(1,945) 49.2(1,411) 1.12 82 immature + 22.5 (686) 23.9(1,052) 0.94

Whole cataractous lenses were separated into cortical and nuclear regions. as = anterior subcapsular; ps = posterior subcapsular; cor = cortical; n Ganglioside content was determined as described in Table 1. The values are nuclear. expressed as total sialic acid content (nmoles), and as that per mg protein * Diabetic cataract. (pmoles/mg) in parentheses.

Downloaded from iovs.arvojournals.org on 10/01/2021 2176 INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / Ocrober 1990 Vol. 31

— -GM3

-GM1 -GD3 o 20 30 40 RETENTION TIME (mtn)

Fig. 5. Mapping of lens gangliosides from cataractous lenses. The lower panel indicates the gradient profile of ammonium acetate used for elution from the DEAE-Iatrobeads column. One milliliter fractions were collected, and an aliquot of each fraction was applied 12 12 to TLC. The plate was developed as described in Figure I. The arrowheads indicate GM3 and a slow-moving ganglioside, which Mature Immature were partially purified by preparative TLC, respectively. Fig. 4. Distribution of gangliosides in mature and immature cataractous lenses. An aliquot of each ganglioside fraction from 70-year-old mature and 73-year-old immature lenses were devel- gangliosides (Fig. 1). Ganglioside mapping revealed a oped as described in Figure I. Ganglioside patterns from cortical more complex pattern of gangliosides, consisting of at (I) and nuclear (2) regions were compared. least 15 components (Fig. 7), than that described by Tao et al.7 In earlier studies by Windeler and Feld- 3 2 moving gangliosides showed retention times incon- man and Sarkar and Cenedella, human cataractous sistent with brain GTlb. The corresponding fraction lens showed a simple ganglioside pattern consisting of and GM3, which showed an Rf-value similar to that GM3 and GM1. In contrast, recent reports by Tao et of brain GM3, were further purified by preparative TLC (arrow heads in Fig. 5). A representative GLC of the N-acetyl-O-TMS methylglycosides of the slow- moving ganglioside is shown in Figure 6. The molar ratio of the sugars was calculated from the peak areas on the GLC. The slow-moving ganglioside was found to consist of glucose, galactose, and sialic acid in a molar ratio of 2:1:4, based on the molar ratio of 1:1:1 of GM3. N-Acetylgalactosamine could not be detected. Methylation of the ganglioside was not done because of the small amount. The fatty acid composition of the ganglioside is summarized in Table 3 and compared with that of GM3, a precursor of other lens gangliosides, GM1 and GD1 a. The slow-moving ganglioside was similar in composition to GM3 except for the long-chain fatty acid moiety. 10 15 20 25 RETENTION TIME (min) Discussion Fig. 6. Gas-liquid chromatogram ofO-TMS derivatives of methylglycosides from the slow-moving ganglioside of human cataracts. The TLC separation in whole cataractous lenses Conditions of separation are given in "Materials and Methods." found GM3, GM1, GDI a, and minor unidentified Gal = galactose; Glc = glucose; SA = N-acetylneuraminic acid.

Downloaded from iovs.arvojournals.org on 10/01/2021 No. 10 GANGUO5IDE5 IN HUMAN SENILE CATARACTS / Ogiso er ol 2177

Table 3. Fatty acid compositions of gangliosides function of gangliosides in cataractous lenses and the isolated from human cataractous lenses relation of ganglioside content and pattern to cataract formation remain unclear. % of Total fatty acids This study indicated that two factors, one related to Slow-moving aging and the other to lens opacification, influence Fatty acid GM3 ganglioside the increase in lens gangliosides on a protein basis. The protein content of cataractous lenses fluctuated 12:0 1.7(3.6)* 3.7 (3.8) 12:1 2.3(4.8) 4.7 (4.8) slightly between 49-84 yr and appeared to increase 14:0 9.8 (20.2) 17.5(18.0) with aging (Table 1). The age-dependent increase in 14:1 1.5(3.1) 4.7 (4.8) ganglioside content was shown in individual mature 16:0 25.7 (53.0) 36.9 (37.9) 16:1 2.1 (4.4) 7.5 (7.7) and immature cataractous lenses (Fig. 3A). A high 18:0 3.0(6.1) 12.6(13.0) correlation was observed in the immature lens group 18:1 2.4 (4.9) 9.8(10.0) (r = 0.87), and the mature groups were distributed 22:0 4.8 2.6 24:0 9.4 over the regression line calculated from immature 24:1 37.3 lenses. The greater accumulation of gangliosides in mature compared with immature lenses appears to Fatty acids are denoted as chain length: number of double bonds. * Values in parentheses are percentages of the sum of values from 12:0 to lead to the maturation of senile cataract. When the 18:1. age-dependent increase in ganglioside content was not considered, there was no significant difference in al7 and Swindell et al6 describe a more complicated terms of cataract progression. However, correction pattern of gangliosides, including polysialoganglio- for the age-dependent increase revealed a significant sides. The discrepancy with earlier studies may be difference between mature and immature cataractous explained by differences in procedural techniques, for lenses (Fig. 3B). The TLC analysis (Fig. 1) showed example, the application of small amounts of samples that the increase in gangliosides with maturation of in the dialysis step and the Folch-Suzuki partition the cataract was occasionally accompanied with the procedure.23'24 In the current study, the dialysis step appearance of polysialogangliosides. However, im- 13 was replaced by the SEP-PAK d8 cartridge method mature lenses also showed the presence of the gangli- since there appeared to be less than a critical amount osides when ganglioside samples were quantitatively needed for the gangliosides to form micelles. applied to TLC (Fig. 4), suggesting that they are more Sarkar and Cenedella2 reported a high ganglioside or less capable of synthesizing the gangliosides. Al- content in human lens (272.5 yug of NANA/g of wet though the detailed mechanism has not been clari- weight), and similar values for total NANA content fied, it can be hypothesized that gangliosides accu- were estimated in cataractous nuclei (294 jig of mulate in the lens with aging and the progression of NANA/g of wet weight), whole cataractous lenses senile cataracts, causing the relative increase in poly- (243 /ig/g) and whole normal lenses (357 Mg/g)-6 We sialogangliosides in mature cataractous lenses. estimated that the ganglioside content in whole cata- In their study of the normal bovine lens, Sarkar ractous lenses ranged from 248-1486 pmol of and Cenedella2 reported that the epithelial cell frac- NANA/mg of protein. Assuming that the wet weight tion contained an approximately tenfold greater con- of lens is 180 mg, ganglioside content as NANA centration of gangliosides than either the lens cortex would be 19.2-114.8 jzg/g. The lower values are con- or nucleus. Swanson and Albers'29 work with bovine sidered to be due to differences in ganglioside deter- lenses indicated differences in ganglioside patterns mination. The thiobarbituric acid method14 for total between the cortical and nuclear regions of the lens sialic acid estimation is dependent on the release of based on TLC analysis. Our study with monkey free sialic acids from gangliosides and provides lower lenses9 provided similar evidence: the total ganglio- values than those estimated by the resorcinol method sidic sialic acid concentration of the cortical region of Svennerholm25 as modified by Miettinen and was about four times greater than that of the nuclear Takki-Luukkainen,26 but it consumes less of the sam- region. A complex ganglioside pattern similar to that ple than the latter. In our previous study,27 much from whole lens was obtained in the ganglioside frac- lower estimates of ganglioside content in human cat- tion from the cortical region. In contrast to lenses in aractous lenses were made, presumably because of several other animal species, no significant differ- the phosphate-buffered tetrahydrofuran method28 ences in resorcinol-positive bands on TLC were ob- and the dialysis step. In addition, whole normal served between human cataractous nuclei and whole lenses from 31-37-yr-olds had no less than two thirds lens.6 In addition, ganglioside content expressed per g of the total ganglioside content of whole cataractous of wet weight was higher in cataractous nuclei than in lenses from 55-89-yr-olds. At present, physiologic whole cataractous lenses,6 suggesting a higher con-

Downloaded from iovs.arvojournals.org on 10/01/2021 2178 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Ocrober 1990 Vol. 31

centration ofgangliosides in the nuclear region. Thus, that palmitic (16:0) and nervonic (24:1) acids pre- human cataracts appear to possess ganglioside distri- dominated in GM3 and GM1. Tao et al7 also re- butions that differ from those of the normal lenses of ported that nervonic acid was the predominant fatty several animals, as shown in Figure 4. Table 2 indi- acid in GM3 and GM1. More recently, Swindell et al6 cates that ganglioside accumulation occurs in the determined fatty acid percentages from total lenticu- cortical region of senile cataract, probably due to an lar ganglioside extract and indicated that nervonic age-dependent increase in plasma membranes. If the acid is increased in both cataractous nuclei and whole age-dependent increase in ganglioside content in each cataractous lenses compared with whole normal region were clarified, more comprehensive informa- lenses. Our data on GM3 were compatible with those tion on lens gangliosides could be obtained. The dis- of earlier reports, but data on the slow-moving gangli- crepancy between the sum of ganglioside content in oside relating to the maturation of cataracts were not. both regions and the values from whole lenses re- Our findings indicate that the slow-moving ganglio- mains unknown, but presumably it is attributable to side which lacks long-chain fatty acids such as ner- contamination by nonsialic acids together with small vonic acid is synthesized by a different pathway. The amounts of sialic acids. Diabetic cataractous lenses fatty acid composition of other ganglioside species were characterized by a high accumulation ofgangli- was not undertaken in this study, but it was expected osides in the nuclear region, suggesting that a differ- to be the same as that of GM3 because GM1 and ent mechanism from that in senile cataracts is oper- GDI a are synthesized from GM3 by the a-pathway of ating in the accumulation of lens gangliosides. An gangliosides. attempt to locate lens gangliosides using the immu- nohistochemical approaches is currently underway. Key words: human cataractous lens, gangliosides, senile cataract For structural analysis of predominant ganglioside species, combined TLC-immunostaining using antiasialo-GM I antiserum (there was no apparent Acknowledgments cross-reactivity to gangliosides by enzyme-linked im- munosorbent assay; Seikagaku Kogyo, Tokyo, The authors thank Professor M. Hoshi, Tokyo Institute of Technology, for useful advise and T. Takagi, Toho Uni- Japan) and acid hydrolysis to remove the sialic acid versity, for his technical assistance. moiety from the ganglioside suggested that, except for GM3 and GM2, they consisted of the gangliotetra- References osyl backbone." However, slow-moving gangliosides, noted in this and a previous study," were regarded as 1. Feldman GL and Feldman LS: New concepts of human len- a mixture of gangliosides with Rf-values lower than ticular lipids and their possible role in cataracts. Invest Oph- that of GDI a. As shown in Figures 1 and 2, trace thalmol 4:162, 1965. 2. Sarkar CP and Cenedella RJ: Gangliosides in normal and cat- levels of GDIb and GTlb were detected. Judging aractous lenses of several species. Biochim Biophys Acta from these facts, antiasialo-GM 1 antiserum is consid- 711:503, 1982. ered to react immunologically exclusively with the 3. Windeler AS and Feldman GL: The isolation and partial backbones originating from contaminated GDlb and structure characterization of some ocular gangliosides. Bio- GTlb. In the current study, GLC analysis revealed chim Biophys Acta 202:361, 1970. 4. Swindell RT, Bell VC, Slaughter S, and Albers-Jackson B: that one of the slow-moving gangliosides was com- Incorporation of 14C-galactose into gangliosides of rabbit lens. posed of glucose, galactose, and sialic acid in the Curr Eye Res 6:451, 1987. molar ratio of 2:1:4 (Fig. 6). Surprisingly, the ganglio- 5. Feldman GL, Feldman LS, and Rouser G: The isolation and side was eluted in the disialoganglioside fraction by partial characterization ofgangliosides and ceramide polyhex- DEAE-Iatrobeads ganglioside mapping despite the osides from the lens of the human eye. Lipids 1:21, 1966. 6. Swindell RT, Harris H, Buchanan L, Bell C, and Albers-Jack- high proportion of sialic acid in the molecule. In ad- son B: Ganglioside composition in human cataractous nuclei. dition, other slow-moving gangliosides also had a Ophthalmic Res 20:232, 1988. similar chromatographic behavior on two-dimen- 7. Tao RVP, Shen Y-W, Kovathana N, and Cotlier E: A new sional TLC and ganglioside mapping (Figs. 2, 5). family of fucose-containing gangliosides isolated from human Thus the cataractous human lens seems to contain senile cataracts. Biochim Biophys Acta 753:89, 1983. 8. Tao RVP and Lee B-C: Isolation and identification of sialic certain ganglioside species, which have unreported acids from the gangliosides of human cataracts. Curr Eye Res sugar-chain structures as minor components. The 5:167, 1986. long-chain base, complete structure, and precursor(s) 9. Ogiso M, Saito N, Sudo K, Hirano S, and Komoto M: Age-de- of the slow-moving ganglioside examined in the cur- pendent changes in monkey lenticular gangliosides. Exp Eye rent study are under investigation. Res 50:51, 1990. 10. Benitz K-F: Measurement of Chronic Toxicity. In Methods in In their study of the fatty acid composition of cata- Toxicology, Paget GE, editor. Oxford, Blackwell Scientific, 5 ractous lenticular gangliosides, Feldman et al found 1970, pp. 82-131.

Downloaded from iovs.arvojournals.org on 10/01/2021 No. 10 GANGUOSIDES IN HUMAN SENILE CATARACTS / Ogiso er ol 2179

11. Saito N, Sudo K, Masago M, Yamashita A, Kobayakawa N, kawa T: Characterization of GMlb in mouse spleen. J Bio- Komoto M, Ogiso M, and Hirano S: Gangliosides in the chem (Tokyo) 96:949, 1984. human cataractous lens. J Eye (Atarashii Ganka) 6:773, 1989 21. Kozulic B, Ries B, and Mildner P: N-Acetylation of amino (in Japanese). sugar methyl glycosides for gas-liquid chromatographic analy- 12. Ledeen RW, Yu RK, and Eng LF: Gangliosides of human sis. Anal Biochem 94:36, 1979. myelin: Sialosylgalactosylceramide (G7) as a major compo- 22. Pirie A: Color and solubility of the proteins of human cata- nent. J Neurochem 21:829, 1973. racts. Invest Ophthalmol 7:634, 1968. 13. Williams MA and McCluer RH: The use of Sep-Pak C,8 car- 23. Folch J, Lees M, and Sloane-Stanley GH: A simple method for tridges during the isolation of gangliosides. J Neurochem the isolation and purification of total lipids from animal tis- 35:266, 1980. sues. J Biol Chem 226:497, 1957. 14. Warren L: The thiobarbituric acid assay of sialic acids. J Biol 24. Suzuki K: The pattern of mammalian brain gangliosides: II. Chem 234:1971, 1959. Evaluation of the extraction procedures, postmortem changes 15. Lowry OH, Rosebrough NJ, Farr A, and Randall RJ: Protein and the effect of formalin preservation. J Neurochem 12:629, measurement with the Folin phenol reagent. J Biol Chem 1965. 193:265, 1951. 25. Svennerholm L: Quantitative estimation of sialic acids: II. A 16. Ando S, Chang N-C, and Yu RK: High-performance thin- colorimetric resorcinol-hydrochloric acid method. Biochim layer chromatography and densitometric determination of Biophys Acta 24:604, 1957. brain ganglioside compositions of several species. Anal Bio- 26. Miettinen T and Takki-luukkainen I: Use of butyl acetate in chem 89:437, 1978. determination of sialic acid. Acta Chem Scand 13:856, 1959. 17. Ohashi M: A comparison of the ganglioside distributions of fat 27. Sudo K, Sugai M, Komoto M, Asou H, and Hirano S: Ganglio- tissues in various animals by two-dimensional thin layer chro- sides in human and monkey lenses. Jpn J Ophthalmol 30:197, matography. Lipids 14:52, 1979. 1986. 18. Svennerholm L: Chromatographic separation of human brain 28. Tettamanti G, Bonali F, Marchesini S, and Zambotti V: A new gangliosides. J Neurochem 10:613, 1963. procedure for the extraction, purification and fractionation of 19. Iwamori M and Nagai Y: A new chromatographic approach to brain gangliosides. Biochim Biophys Acta 296:160, 1973. the resolution of individual gangliosides: Ganglioside map- 29. Swanson AA and Albers B: Proteolytic activity on lens proteins ping. Biochim Biophys Acta 528:257, 1978. and gangliosides in mammalian lens. Interdiscipl Topics Ger- 20. Nakamura K, Hashimoto Y, Suzuki M, Suzuki A, and Yama- ont 12:241, 1978.

Downloaded from iovs.arvojournals.org on 10/01/2021