- Email: [email protected]
Protein microanalysis of animal tears
Research in Veterinary Science 2000, 68, 207–209 doi:10.1053/rvsc.1999.0358, available online at http://www.idealibrary.com on
Protein microanalysis of animal tears S. HEMSLEY*, N. COLE*†, P. CANFIELD*, M. D. P. WILLCOX† *Department of Veterinary Anatomy and Pathology, University of Sydney, Sydney, NSW 2006, Australia, †Cooperative Research Centre for Eye Research and Technology and Corneal and Contact Lens Research Unit, School of Optometry, University of New South Wales, Sydney, NSW 2052, Australia SUMMARY Sub-microlitre volumes of normal koala, mouse, dog, rat and cat tears were fractionated using size exclusion-high performance liquid chromatography (SE-HPLC), giving reproducible profiles which were different for each species. Microlitre volumes of tears were also fractionated using sodium dodecylsulphate-polyacrylamide gel electrophoresis (SDS-PAGE), resulting in good separation of individual tear proteins with a species specific distribution. Tears from koalas with conjunctivitis and mice with keratitis were similarly examined and showed mostly quantitative changes. These simple, rapid techniques gave reproducible results and, in contrast to conventional separation techniques, used easily obtainable volumes (as little as 0·75 µl) of tears. Their expansion could allow isola tion, identification and quantitation of individual tear components, enabling effective investigation of changes occurring in disease. © 2000 Harcourt Publishers Ltd
PROPER function of the tear film is essential for maintenance of ocular health. The normal aqueous tear layer in humans contains more than 60 protein components (Gachon et al 1979). The major tear proteins include lysozyme, lactoferrin, tear lipocalin and secretory IgA (sIgA), which are important in defending the ocular surface against infection (Sullivan 1994). Although human tears have been investigated, there has been little study of animal tears, particularly in relation to disease states. There are large variations in tear proteins among domestic animals (van Agtmaal et al 1985, Davidson et al 1994); thus accurate and sensitive characterisation of tear components in individual species is crucial for establishing normal tear profiles and for interpreting disease-induced changes. Characterisation of such changes should enhance understanding of the host response and lead to improvements in diagnosis, treatment and prognosis. A major hindrance to analysis of the tear film is the small quantity of fluid available. Separation techniques commonly require quantities of protein unavailable from collection of animal tears. This study demonstrates simple, reproducible methods for fractionating proteins of sub-microlitre and microlitre volumes of tear fluid from a range of animal species, including the koala. Analysis of koala tears has not been previously reported. Koalas (N = 32) were manually restrained for tear collection. Tears were collected using a 10 µL borosilicate glass microcapillary tube (Corning Inc, Corning, NY, USA). Tears were obtained from 27 koalas with clinically normal eyes and five koalas with conjunctivitis, four cases of which were known to be due to chlamydial infection. Corresponding author: Dr Susan Hemsley, Department of Veterinary Anatomy and Pathology, University of Sydney, NSW 2006, Australia, Fax: 61 2 9351 6880, Email: [email protected] 0034-5288/00/030207 + 03 $35.00/0
Mice (N = 33) were lightly anaesthetised (Avertin, 125 mg kg–1 intraperitoneal) before tear collection by gentle aspiration using soft polypropylene tubing (0·28-mm internal diameter). Tears were obtained from 30 clinically normal mice and from three mice 24 hours post induction of Pseudomonas aeruginosa keratitis as described by Cole et al (1998). Normal tears were collected from manually restrained dogs (N = 3), rats (N = 3) and cats (N = 3). Human tears (stimulated and unstimulated) (N = 9), were collected using 10 µl borosilicate glass microcapillary tubes. Tear samples from all species were stored at –70°C or –20°C until analysis. Size exclusion-high performance liquid chromatography (SE-HPLC) of all tears was carried out using a SMART integrated liquid chromatography system on a 2·4-ml Superose 12-microbore column (Pharmacia LKB, Uppsala, Sweden). Protein was eluted at 40 µl minute–1 with phosphate buffered saline (pH 7·2). Protein was detected at 280 nm with automated area integration of peaks. Samples were centrifuged for 5 minutes at 8000 g immediately prior to application of 0·75 µl volumes of samples to the column. Injections of human tear samples of 0·75 µl gave reproducible elution profiles directly comparable to those obtained with more conventional sample volumes (20 to 50 µl), but with improved resolution. Fractions collected from 20 µl samples of human tears were sufficient to allow identification by semi-dry Western immunoblotting of major proteins present in peaks of tear elution profiles (Fig 1a). Following SDS-PAGE of these fractions, protein was transferred electrophoretically on to a nitrocellulose membrane (Biorad, Richmond, CA, USA) using a semi-dry electroblotting apparatus (Gradipore, Sydney, Australia) according to the manufacturer’s intructions. Immunodetection of proteins was carried out as described by Bollag and Edelstein (1991). The following primary antibodies were used: rabbit anti-human lysozyme © 2000 Harcourt Publishers Ltd
S. Hemsley, N. Cole, P. Canfield, M. D. P. Willcox
208
a
0.018
LF
0.015
0.33
LZ
0.013 Absorbance 280 mm
b
0.37
0.28
0.010
0.23
0.008
0.18
0.005
0.13
0.003
0.08
A
0.000
0.03
0.50
1.00
1.50
2.00
2.50
mL
0.00
0.50 1.00
1.50 2.00 2.50
c
0.02
d 0.009 0.007 0.005
0.01
0.003 0.001 0.00
–0.001
0
1
2
0.50
1.00
1.50
2.00
e
2.50
f
0.060
0.010
0.040
0.020
0.000
0.000
–0.20
0.80
1.80
2.80 0.00
1.00
2.00
FIG 1: SE-HPLC protein elution profiles of 0·75-µl samples of koala, mouse, dog, rat and cat tears. Elution profiles for human tears are included for comparison. Variation among human tear profiles is due to individual variation and differing contributions of stimulated and unstimulated tears among samples. Axis identification and human peak assignments for figures a–f are indicated in figure a. (A = IgA-mucin; LF = lactoferrin; LZ = lysozyme) (a) Overlay of SEHPLC elution profiles of normal koala (–) and human (–) tears. (b) Overlay of SEHPLC elution profiles of normal koala tears (–; value for absorbance multiplied by 10) and conjunctivitis koala tears (–) demonstrating quantitative and qualitative differences between the samples. (c) Overlay of SE-HPLC elution profiles of normal mouse (–) and human (–) tears. (d) Overlay of SE-HPLC elution profiles of normal dog (–) and human (–) tears. (e) Overlay of SE-HPLC elution profiles of normal rat (–) and human (–) tears. (f) Overlay of SE-HPLC elution profiles of normal cat (lower trace) and human (higher trace) tears. Some human peaks off scale due to the relatively high protein concentration of the human tears compared to cat tears.
(Dako Australia, Sydney, Australia: A0099), rabbit antihuman lactoferrin (Sigma-Aldrich, Sydney, Australia: L3262) and goat anti-human IgA α chain-biotin (SigmaAldrich: A2691). Swine anti-rabbit immunoglobulin-peroxidase (Dako Australia: P0399) was used for development of rabbit primary antibodies and Extravidin peroxidase (Sigma-Aldrich: E2886) was used for development of the goat primary antibody. Elution profiles obtained for normal koala tears were reproducible and consisted of three major peaks. Overlay of the elution profile of normal koala tears with that of normal human tears showed marked differences (Fig 1a). The koala samples had fewer major peaks which eluted earlier than those of human tears. Tear elution profiles for koalas with conjunctivitis showed no consistent pattern when compared to those of normal tears; changes were predominantly quantitative although some qualitative differences occurred (Fig 1b). Profiles for normal mouse tears were reproducible and consisted of five major peaks (Fig 1c). Overlay of keratitis mouse tears and normal mouse tears showed only quantitative alterations in the profiles, with increased area under the curve in keratitis. Dog, rat and cat tears yielded unique and reproducible profiles. There were six major peaks in dog tears (Fig 1d) while in rat tears four peaks were identified
(Fig 1e). Tears from cats had low protein concentration but six peaks were resolved (Fig 1f). Sodium dodecylsulphate-polyacrylamide gel electrophoresis (SDS-PAGE) of reduced tear proteins was carried out using 4 to 20 per cent (w/v) polyacrylamide minigels (Novex, San Diego, CA, USA: EC6025 or Biorad: 161-1105). Standard molecular weight (MW) markers ranging from 2·5 kDa to 200 kDa (Mark 12, Novex: LC 5677) were run on each gel. Approximately 2-µl of tear sample diluted 1:4 was applied to each well. Samples were electrophoresed at 14 mA/gel constant current on a Hoefer Mighty Small slab gel electrophroesis unit (Hoefer Scientific Instruments, San Francisco, CA, USA: SE 250). Proteins were detected by staining with 0·025 per cent Coomassie Blue R250. However, there was insufficient protein present in normal mouse tears to be detected by this method. Silver staining was used for these samples. The distribution of protein bands following SDS-PAGE was reproducible and species specific (Fig 2). Tear samples from koalas with conjunctivitis separated by SDS-PAGE had a distribution of protein bands by MW similar to those of normal koala tears (Fig 2) but variable quantitative differences appeared to be present. Tear samples from normal mice and mice with P. aeruginosa keratitis separated by SDS-PAGE also showed little qualitative difference, but some proteins appeared altered in concentration. This study has established simple, rapid and reproducible techniques for fractionation of sub-microlitre and microlitre volumes of animal tear fluid. This overcomes a major hindrance to study of tear components, namely the limited volume of fluid usually available. In addition, this study has established normal protein fractionation profiles for all species examined. Preliminary information on tear proteins in diseased koala and mouse eyes was obtained. SE-HPLC resulted in highly repeatable elution profiles which varied among species. Resolution of individual
LF
97
97
55
55 31 LC
31
14 14
LZ 1
2
3
4
5
6
7
8
9
FIG 2: SDS-PAGE of reduced koala, mouse, cat, dog and rat tears. MW of standards for lanes 2, 3, 4 and 5 are indicated adjacent to lane 1 and those for lanes 6, 7 and 8 are indicated adjacent to lane 9. Lane 1 molecular weight standards. Lane 2 keratitis mouse tears. Lane 3 normal cat tears. Lane 4 normal dog tears. Lane 5 normal rat tears. Lane 6 conjunctivitis koala tears. Lane 7 normal koala tears. Lane 8 normal human tears. LF = lactoferrin; LC = lipocalin. LZ = lysozyme (Berta 1986). Lane 9 molecular weight standards.
Tear protein microanalysis
proteins was not possible, however, as under native conditions proteins in tear fluid are found as non-covalently associated complexes (Boonstra et al 1988). Protein purification by conventional HPLC requires specialised biochemical techniques and training. The SMART integrated chromatography micropurification system is accessible to users not specifically trained in protein fractionation techniques, as it is controlled by user-friendly software. In addition, the low dead volume of the system means that there is little buffer to interfere with subsequent protein chemistry techniques and enables very small sample volumes to be effectively separated (Gooley et al 1994). The major application of this technique to the study of tears is likely to be as the initial step in the isolation, identification and purification of individual proteins. However, with further characterisation of the proteins represented by peaks in tear elution profiles, applications to direct studies of ocular disease may be possible. SDS-PAGE gave greater resolution of individual proteins in tears than SE-HPLC. Animal tear profiles varied according to species and did not correspond in all respects to human tear profiles. Previous studies have shown similar variation among different species (van Agtmaal et al 1985, Davidson et al 1994). The variation may be because of presence or absence of individual protein components, variation in their concentrations, differing degrees of protein glycosylation or differing MW of individual proteins. Comparison of normal koala and mouse tears with tears from animals affected by conjunctivitis or keratitis revealed variable quantitative differences by both SE-HPLC and SDSPAGE. In conjunctivitis, increased protein concentrations are likely to have resulted from serum transudation or increases in proteins of lacrimal gland origin such as sIgA. In contrast, lowered protein concentrations may have been due to exhaustion of lacrimal gland secretory cells because of prolonged secretion of high volumes of tear fluid (Berta 1986). This study indicates a number of possible avenues for further investigation of tears of species encountered in veterinary practice and research. Extension of the techniques developed could allow identification and quantitation of individual tear components. With the availability of appropriate antibodies, chromatographic techniques and immunoblotting could be utilised to isolate, identify and purify individual proteins. Procedures such as ELISA could then be used to determine normal levels of these proteins and alterations with disease. Therefore, with expansion,
209
SE-HPLC
and SDS-PAGE of sub-microlitre and microlitre tear samples will enable rapid and reliable investigation of qualitative and quantitative changes in tear protein components in disease or drug therapy.
ACKNOWLEDGEMENTS We thank the Koala Preservation Society of NSW Inc., The Australian Wildlife Park, Featherdale Wildlife Park and Patricia Martin for their assistance in obtaining tear samples, Jacqui Norris and Daria Love for assistance with electrophoresis, Sydney University Protein and Macromolecular Analysis Centre for assistance with the SMART system, and Bozena Jantulik for assistance with graphs. This work was supported by a University of Sydney and Australian Research Council Institutional Grant, an Australian Federal Government Grant under the Cooperative Research Centres Programme and the Koala Preservation Society of NSW Inc.
REFERENCES BERTA, A. (1986) Standardization of tear protein determinations: The effects of sampling, flow rate, and vascular permeability. In: The Preocular Tear Film in Health, Disease and Contact Lens Wear. Ed F. J. Holly. Lubbock: Dry Eye Institute, Lubbock, Texas. pp, 418–435 BOLLAG, D.M. & EDELSTEIN, S.J. (1991) Immunoblotting. In Protein Methods, pp 181–211. New York: Wiley-Liss BOONSTRA, A., BREEBAART, A.C., BRINKMAN, C.J.J., LUYENDIJK, L., KUIZENGA, A. & KIJLSTRA, A. (1988) Factors influencing the quantitative determination of tear proteins by high performance liquid chromatography. Current Eye Research 7, 893–901 COLE, N., WILLCOX, M.D.P., FLEISZIG, S.M.J., STAPLETON, F., BAO, B., TOUT, S. & HUSBAND, A. (1998) Different strains of Pseudomonas aeruginosa isolated from ocular infections or inflammation display distinct corneal pathologies in an animal model. Current Eye Research 17, 730–735 DAVIDSON, H.J., BLANCHARD, G.L. & MONTGOMERY, P.C. (1994) Comparisons of tear proteins in the cow, horse, dog and rabbit. In: Lacrimal Gland, Tear Film and Dry Eye Syndromes. Advances in Experimental Medicine and Biology No.350. Ed D. A. Sullivan. New York: Plenum Press, pp. 331–334 GACHON, A.M., VERRELLE, P., BETAIL, G. & DASTUGUE, B. (1979) Immunological and electrophoretic studies of human tear proteins. Experimental Eye Research 29, 539–553 GOOLEY, A.A., ZHOU CHOU, A. WILLIAMS, K.L. (1994) Integration makes protein purification easier. Todays Life Sciences June, 32–36 SULLIVAN, D.A. (1994) Ocular mucosal immunity. In: Handbook of Mucosal Immunology. Eds P. L. Ogra, J. Mestecky, M. E. Lamm, W. Strober, J. R. McGhee & J. Bienenstock. San Diego: Academic Press, pp. 569–597 VAN AGTMAAL, E.J., THÖRIG, L. & VAN HAERINGEN, N.J. (1985) Comparative protein patterns in tears of several species. In: Proteins of the Biological Fluids. Vol. 32. Ed H. Peeters. Oxford: Pergamon Press, pp. 395–397 Accepted November 20, 1999
Protein microanalysis of animal tears S. HEMSLEY*, N. COLE*†, P. CANFIELD*, M. D. P. WILLCOX† *Department of Veterinary Anatomy and Pathology, University of Sydney, Sydney, NSW 2006, Australia, †Cooperative Research Centre for Eye Research and Technology and Corneal and Contact Lens Research Unit, School of Optometry, University of New South Wales, Sydney, NSW 2052, Australia SUMMARY Sub-microlitre volumes of normal koala, mouse, dog, rat and cat tears were fractionated using size exclusion-high performance liquid chromatography (SE-HPLC), giving reproducible profiles which were different for each species. Microlitre volumes of tears were also fractionated using sodium dodecylsulphate-polyacrylamide gel electrophoresis (SDS-PAGE), resulting in good separation of individual tear proteins with a species specific distribution. Tears from koalas with conjunctivitis and mice with keratitis were similarly examined and showed mostly quantitative changes. These simple, rapid techniques gave reproducible results and, in contrast to conventional separation techniques, used easily obtainable volumes (as little as 0·75 µl) of tears. Their expansion could allow isola tion, identification and quantitation of individual tear components, enabling effective investigation of changes occurring in disease. © 2000 Harcourt Publishers Ltd
PROPER function of the tear film is essential for maintenance of ocular health. The normal aqueous tear layer in humans contains more than 60 protein components (Gachon et al 1979). The major tear proteins include lysozyme, lactoferrin, tear lipocalin and secretory IgA (sIgA), which are important in defending the ocular surface against infection (Sullivan 1994). Although human tears have been investigated, there has been little study of animal tears, particularly in relation to disease states. There are large variations in tear proteins among domestic animals (van Agtmaal et al 1985, Davidson et al 1994); thus accurate and sensitive characterisation of tear components in individual species is crucial for establishing normal tear profiles and for interpreting disease-induced changes. Characterisation of such changes should enhance understanding of the host response and lead to improvements in diagnosis, treatment and prognosis. A major hindrance to analysis of the tear film is the small quantity of fluid available. Separation techniques commonly require quantities of protein unavailable from collection of animal tears. This study demonstrates simple, reproducible methods for fractionating proteins of sub-microlitre and microlitre volumes of tear fluid from a range of animal species, including the koala. Analysis of koala tears has not been previously reported. Koalas (N = 32) were manually restrained for tear collection. Tears were collected using a 10 µL borosilicate glass microcapillary tube (Corning Inc, Corning, NY, USA). Tears were obtained from 27 koalas with clinically normal eyes and five koalas with conjunctivitis, four cases of which were known to be due to chlamydial infection. Corresponding author: Dr Susan Hemsley, Department of Veterinary Anatomy and Pathology, University of Sydney, NSW 2006, Australia, Fax: 61 2 9351 6880, Email: [email protected] 0034-5288/00/030207 + 03 $35.00/0
Mice (N = 33) were lightly anaesthetised (Avertin, 125 mg kg–1 intraperitoneal) before tear collection by gentle aspiration using soft polypropylene tubing (0·28-mm internal diameter). Tears were obtained from 30 clinically normal mice and from three mice 24 hours post induction of Pseudomonas aeruginosa keratitis as described by Cole et al (1998). Normal tears were collected from manually restrained dogs (N = 3), rats (N = 3) and cats (N = 3). Human tears (stimulated and unstimulated) (N = 9), were collected using 10 µl borosilicate glass microcapillary tubes. Tear samples from all species were stored at –70°C or –20°C until analysis. Size exclusion-high performance liquid chromatography (SE-HPLC) of all tears was carried out using a SMART integrated liquid chromatography system on a 2·4-ml Superose 12-microbore column (Pharmacia LKB, Uppsala, Sweden). Protein was eluted at 40 µl minute–1 with phosphate buffered saline (pH 7·2). Protein was detected at 280 nm with automated area integration of peaks. Samples were centrifuged for 5 minutes at 8000 g immediately prior to application of 0·75 µl volumes of samples to the column. Injections of human tear samples of 0·75 µl gave reproducible elution profiles directly comparable to those obtained with more conventional sample volumes (20 to 50 µl), but with improved resolution. Fractions collected from 20 µl samples of human tears were sufficient to allow identification by semi-dry Western immunoblotting of major proteins present in peaks of tear elution profiles (Fig 1a). Following SDS-PAGE of these fractions, protein was transferred electrophoretically on to a nitrocellulose membrane (Biorad, Richmond, CA, USA) using a semi-dry electroblotting apparatus (Gradipore, Sydney, Australia) according to the manufacturer’s intructions. Immunodetection of proteins was carried out as described by Bollag and Edelstein (1991). The following primary antibodies were used: rabbit anti-human lysozyme © 2000 Harcourt Publishers Ltd
S. Hemsley, N. Cole, P. Canfield, M. D. P. Willcox
208
a
0.018
LF
0.015
0.33
LZ
0.013 Absorbance 280 mm
b
0.37
0.28
0.010
0.23
0.008
0.18
0.005
0.13
0.003
0.08
A
0.000
0.03
0.50
1.00
1.50
2.00
2.50
mL
0.00
0.50 1.00
1.50 2.00 2.50
c
0.02
d 0.009 0.007 0.005
0.01
0.003 0.001 0.00
–0.001
0
1
2
0.50
1.00
1.50
2.00
e
2.50
f
0.060
0.010
0.040
0.020
0.000
0.000
–0.20
0.80
1.80
2.80 0.00
1.00
2.00
FIG 1: SE-HPLC protein elution profiles of 0·75-µl samples of koala, mouse, dog, rat and cat tears. Elution profiles for human tears are included for comparison. Variation among human tear profiles is due to individual variation and differing contributions of stimulated and unstimulated tears among samples. Axis identification and human peak assignments for figures a–f are indicated in figure a. (A = IgA-mucin; LF = lactoferrin; LZ = lysozyme) (a) Overlay of SEHPLC elution profiles of normal koala (–) and human (–) tears. (b) Overlay of SEHPLC elution profiles of normal koala tears (–; value for absorbance multiplied by 10) and conjunctivitis koala tears (–) demonstrating quantitative and qualitative differences between the samples. (c) Overlay of SE-HPLC elution profiles of normal mouse (–) and human (–) tears. (d) Overlay of SE-HPLC elution profiles of normal dog (–) and human (–) tears. (e) Overlay of SE-HPLC elution profiles of normal rat (–) and human (–) tears. (f) Overlay of SE-HPLC elution profiles of normal cat (lower trace) and human (higher trace) tears. Some human peaks off scale due to the relatively high protein concentration of the human tears compared to cat tears.
(Dako Australia, Sydney, Australia: A0099), rabbit antihuman lactoferrin (Sigma-Aldrich, Sydney, Australia: L3262) and goat anti-human IgA α chain-biotin (SigmaAldrich: A2691). Swine anti-rabbit immunoglobulin-peroxidase (Dako Australia: P0399) was used for development of rabbit primary antibodies and Extravidin peroxidase (Sigma-Aldrich: E2886) was used for development of the goat primary antibody. Elution profiles obtained for normal koala tears were reproducible and consisted of three major peaks. Overlay of the elution profile of normal koala tears with that of normal human tears showed marked differences (Fig 1a). The koala samples had fewer major peaks which eluted earlier than those of human tears. Tear elution profiles for koalas with conjunctivitis showed no consistent pattern when compared to those of normal tears; changes were predominantly quantitative although some qualitative differences occurred (Fig 1b). Profiles for normal mouse tears were reproducible and consisted of five major peaks (Fig 1c). Overlay of keratitis mouse tears and normal mouse tears showed only quantitative alterations in the profiles, with increased area under the curve in keratitis. Dog, rat and cat tears yielded unique and reproducible profiles. There were six major peaks in dog tears (Fig 1d) while in rat tears four peaks were identified
(Fig 1e). Tears from cats had low protein concentration but six peaks were resolved (Fig 1f). Sodium dodecylsulphate-polyacrylamide gel electrophoresis (SDS-PAGE) of reduced tear proteins was carried out using 4 to 20 per cent (w/v) polyacrylamide minigels (Novex, San Diego, CA, USA: EC6025 or Biorad: 161-1105). Standard molecular weight (MW) markers ranging from 2·5 kDa to 200 kDa (Mark 12, Novex: LC 5677) were run on each gel. Approximately 2-µl of tear sample diluted 1:4 was applied to each well. Samples were electrophoresed at 14 mA/gel constant current on a Hoefer Mighty Small slab gel electrophroesis unit (Hoefer Scientific Instruments, San Francisco, CA, USA: SE 250). Proteins were detected by staining with 0·025 per cent Coomassie Blue R250. However, there was insufficient protein present in normal mouse tears to be detected by this method. Silver staining was used for these samples. The distribution of protein bands following SDS-PAGE was reproducible and species specific (Fig 2). Tear samples from koalas with conjunctivitis separated by SDS-PAGE had a distribution of protein bands by MW similar to those of normal koala tears (Fig 2) but variable quantitative differences appeared to be present. Tear samples from normal mice and mice with P. aeruginosa keratitis separated by SDS-PAGE also showed little qualitative difference, but some proteins appeared altered in concentration. This study has established simple, rapid and reproducible techniques for fractionation of sub-microlitre and microlitre volumes of animal tear fluid. This overcomes a major hindrance to study of tear components, namely the limited volume of fluid usually available. In addition, this study has established normal protein fractionation profiles for all species examined. Preliminary information on tear proteins in diseased koala and mouse eyes was obtained. SE-HPLC resulted in highly repeatable elution profiles which varied among species. Resolution of individual
LF
97
97
55
55 31 LC
31
14 14
LZ 1
2
3
4
5
6
7
8
9
FIG 2: SDS-PAGE of reduced koala, mouse, cat, dog and rat tears. MW of standards for lanes 2, 3, 4 and 5 are indicated adjacent to lane 1 and those for lanes 6, 7 and 8 are indicated adjacent to lane 9. Lane 1 molecular weight standards. Lane 2 keratitis mouse tears. Lane 3 normal cat tears. Lane 4 normal dog tears. Lane 5 normal rat tears. Lane 6 conjunctivitis koala tears. Lane 7 normal koala tears. Lane 8 normal human tears. LF = lactoferrin; LC = lipocalin. LZ = lysozyme (Berta 1986). Lane 9 molecular weight standards.
Tear protein microanalysis
proteins was not possible, however, as under native conditions proteins in tear fluid are found as non-covalently associated complexes (Boonstra et al 1988). Protein purification by conventional HPLC requires specialised biochemical techniques and training. The SMART integrated chromatography micropurification system is accessible to users not specifically trained in protein fractionation techniques, as it is controlled by user-friendly software. In addition, the low dead volume of the system means that there is little buffer to interfere with subsequent protein chemistry techniques and enables very small sample volumes to be effectively separated (Gooley et al 1994). The major application of this technique to the study of tears is likely to be as the initial step in the isolation, identification and purification of individual proteins. However, with further characterisation of the proteins represented by peaks in tear elution profiles, applications to direct studies of ocular disease may be possible. SDS-PAGE gave greater resolution of individual proteins in tears than SE-HPLC. Animal tear profiles varied according to species and did not correspond in all respects to human tear profiles. Previous studies have shown similar variation among different species (van Agtmaal et al 1985, Davidson et al 1994). The variation may be because of presence or absence of individual protein components, variation in their concentrations, differing degrees of protein glycosylation or differing MW of individual proteins. Comparison of normal koala and mouse tears with tears from animals affected by conjunctivitis or keratitis revealed variable quantitative differences by both SE-HPLC and SDSPAGE. In conjunctivitis, increased protein concentrations are likely to have resulted from serum transudation or increases in proteins of lacrimal gland origin such as sIgA. In contrast, lowered protein concentrations may have been due to exhaustion of lacrimal gland secretory cells because of prolonged secretion of high volumes of tear fluid (Berta 1986). This study indicates a number of possible avenues for further investigation of tears of species encountered in veterinary practice and research. Extension of the techniques developed could allow identification and quantitation of individual tear components. With the availability of appropriate antibodies, chromatographic techniques and immunoblotting could be utilised to isolate, identify and purify individual proteins. Procedures such as ELISA could then be used to determine normal levels of these proteins and alterations with disease. Therefore, with expansion,
209
SE-HPLC
and SDS-PAGE of sub-microlitre and microlitre tear samples will enable rapid and reliable investigation of qualitative and quantitative changes in tear protein components in disease or drug therapy.
ACKNOWLEDGEMENTS We thank the Koala Preservation Society of NSW Inc., The Australian Wildlife Park, Featherdale Wildlife Park and Patricia Martin for their assistance in obtaining tear samples, Jacqui Norris and Daria Love for assistance with electrophoresis, Sydney University Protein and Macromolecular Analysis Centre for assistance with the SMART system, and Bozena Jantulik for assistance with graphs. This work was supported by a University of Sydney and Australian Research Council Institutional Grant, an Australian Federal Government Grant under the Cooperative Research Centres Programme and the Koala Preservation Society of NSW Inc.
REFERENCES BERTA, A. (1986) Standardization of tear protein determinations: The effects of sampling, flow rate, and vascular permeability. In: The Preocular Tear Film in Health, Disease and Contact Lens Wear. Ed F. J. Holly. Lubbock: Dry Eye Institute, Lubbock, Texas. pp, 418–435 BOLLAG, D.M. & EDELSTEIN, S.J. (1991) Immunoblotting. In Protein Methods, pp 181–211. New York: Wiley-Liss BOONSTRA, A., BREEBAART, A.C., BRINKMAN, C.J.J., LUYENDIJK, L., KUIZENGA, A. & KIJLSTRA, A. (1988) Factors influencing the quantitative determination of tear proteins by high performance liquid chromatography. Current Eye Research 7, 893–901 COLE, N., WILLCOX, M.D.P., FLEISZIG, S.M.J., STAPLETON, F., BAO, B., TOUT, S. & HUSBAND, A. (1998) Different strains of Pseudomonas aeruginosa isolated from ocular infections or inflammation display distinct corneal pathologies in an animal model. Current Eye Research 17, 730–735 DAVIDSON, H.J., BLANCHARD, G.L. & MONTGOMERY, P.C. (1994) Comparisons of tear proteins in the cow, horse, dog and rabbit. In: Lacrimal Gland, Tear Film and Dry Eye Syndromes. Advances in Experimental Medicine and Biology No.350. Ed D. A. Sullivan. New York: Plenum Press, pp. 331–334 GACHON, A.M., VERRELLE, P., BETAIL, G. & DASTUGUE, B. (1979) Immunological and electrophoretic studies of human tear proteins. Experimental Eye Research 29, 539–553 GOOLEY, A.A., ZHOU CHOU, A. WILLIAMS, K.L. (1994) Integration makes protein purification easier. Todays Life Sciences June, 32–36 SULLIVAN, D.A. (1994) Ocular mucosal immunity. In: Handbook of Mucosal Immunology. Eds P. L. Ogra, J. Mestecky, M. E. Lamm, W. Strober, J. R. McGhee & J. Bienenstock. San Diego: Academic Press, pp. 569–597 VAN AGTMAAL, E.J., THÖRIG, L. & VAN HAERINGEN, N.J. (1985) Comparative protein patterns in tears of several species. In: Proteins of the Biological Fluids. Vol. 32. Ed H. Peeters. Oxford: Pergamon Press, pp. 395–397 Accepted November 20, 1999