ABSTRACT

Lacryglobin has been identified in human tears. This proteinhas high sequence homology to the mammaglobins, proteinsupregulated in breast cancer and in breast cancer metastasis.In order to investigate the utility of tear screening for cancer,tear samples were collected from patients with different typesof cancer and compared to controls.Tear samples were takenfrom five controls and eight breast, six lung, five colon, oneprostate and three ovary cancer patients.Tears were analysedusing 2-D gel electrophoresis (n = 25) and 1-D electro-phoresis (n = 3). Lacryglobin was present in the following percentage of patients: breast cancer (88%), lung (83%),colon (100%), ovary (33%), prostate (100%) and controls(60%). Two control patients with lacryglobin had a familyhistory of breast and prostate cancer. Lacryglobin wasdetected in some but not all tear samples and further studiesare warranted to investigate its potential as a marker forcancer.

Key words: cancer, lacryglobin, mammaglobin, tears.

INTRODUCTION

Tears may be a useful non-invasive way of detecting cancerand cancer metastasis. A protein called lacryglobin has beenidentified in human tears.1 Proteins with a close sequencematch to lacryglobin, called mammaglobin B and mamma-globin A, are sensitive and highly specific markers for breastcancer and breast cancer metastasis. In order to investigatethe use of tears to diagnose cancer and cancer metastasis, weexamined expression of lacryglobin in tears from cancerpatients and controls.

Lacryglobin has 68% sequence homology to mamma-globin A, a protein upregulated in breast cancer that is found in mammary gland tissue. Mammaglobin is a sensitive

and specific marker for breast cancer and breast cancermetastasis to lymph nodes.2,3 Lacryglobin is identical tomammaglobin B, a protein found in reproductive and gastro-intestinal tract tissues which is also a marker for cancermetastasis.4 Mammaglobin A and mammaglobin B belong tothe uteroglobin family of proteins.5 Mammaglobin A expres-sion in non-malignant tissue is limited to the mammarygland.2 In cancer, mammaglobin A mRNA has been identi-fied in circulating carcinoma cells in the peripheral blood-stream.6 In breast cancer, mammaglobin A is expressed at up to 10 times normal levels.2

Mammaglobin B has been identified in the breast, uterus,salivary gland, gastrointestinal tract, thymus, trachea,kidney and tears.5,7 The gene for mammaglobin B is locatedon the same chromosome as mammaglobin A (11q12 and11q13, respectively);8,9 however, their expression is not concordant.5 Mammaglobin B is a homologue of lipophilinC.8 Lipophilin C is the human homologue of prostatein, themajor protein secreted by the rat prostate gland. LipophilinC (mammaglobin B) binds androgens and other steroids. Itmay also bind estramusine which is a chemotherapeuticagent used for prostate cancer.10 Mammaglobin B is over-expressed in breast cancer and in malignant lymph nodesbut not in normal lymph nodes. As such, it may be a sensi-tive marker for metastasis of cancer cells to lymph nodes.7

METHODS

Patients were enrolled through the Prince of Wales HospitalOncology Day Centre. Procedures were in accordance withthe Declaration of Helsinki (2000). Institutional ethicsapproval and informed consent were obtained. Tears fromfive controls, eight breast cancer, six lung cancer, five coloncancer, one prostate cancer and three ovary cancer patientswere collected using glass microcapillary tubes and frozen at–80°C. Tear samples were analysed using 2-D gel electro-phoresis (n = 25) or 1-D gel electrophoresis (n = 3) accord-ing to methods previously described.1,11

Clinical and Experimental Ophthalmology (2001) 29, 161–163

Lens and Cornea

Lacryglobin in human tears, a potential marker for cancerVictoria Evans BOptom,1,2 Cassandra Vockler,3 Michael Friedlander PhD FRACP,4 Bradley Walsh PhD3 andMark DP Willcox PhD1

1Cooperative Research Centre for Eye Research and Technology, School of Optometry, University of New South Wales, 2School ofScience, University of Western Sydney Nepean, 3Australian Proteome Analysis Facility, Macquarie University and 4Oncology Day Centre,Prince of Wales Hospital, Sydney, New South Wales, Australia

� Correspondence: Ms Victoria Evans, Cooperative Research Centre for Eye Research and Technology, The University of New South Wales, Sydney, NSW 2052,

Australia. Email: v.evans@cclru.unsw.edu.au

The volume of lacryglobin and the volume of a referencelipocalin protein spot were calculated for all samples usingMelanie II 2-D PAGE image analysis software (Bio-Rad,Hercules, CA, USA). The level of lacryglobin was expressedas a percentage ratio of lacryglobin over the major isoformof tear lipocalin in order to allow comparison between gelsand tear samples.

Lacryglobin was identified using matrix-assisted laserdesorption–time of flight (MALDI–TOF) mass spectro-metry and peptide mass fingerprint matching on proteinsequence databases.12,13 Saliva samples were examined with1-D sodium dodecylsulfate polyacrylamide gel electro-phoresis (SDS-PAGE) to see if an 8-kDa band could be visualized in a subject who had lacryglobin in their tears.

RESULTS

Lacryglobin was identified using peptide mass finger-printing. On 2-D gels it had a relative molecular weight of8 kDa and an isoelectric point of approximately 5 (Fig. 1).The frequency of expression of lacryglobin varied amongpatients. There was a high frequency of lacryglobin detec-tion in breast, lung, colon and prostate cancer patients andin controls with a family history of breast and prostatecancer. The frequency of detection was low in ovariancancer patients and in controls with no family history ofcancer (Table 1). In this small sample size, no relationshipbetween the ratio of lacryglobin to lipocalin volume withmedical history, age, cancer type or disease status was identified (Table 2). Lacryglobin could not be detected insaliva using SDS-PAGE.

DISCUSSION

Lacryglobin was detected in some but not all tear samples.Lacryglobin was detected in a high number of breast cancer,lung cancer and colon cancer patients. Expression of lacry-globin was low in ovary cancer patients. Expression in con-trols varied, possibly depending on whether or not there wasa family history. This needs to be examined further in alarger number of controls in order to establish normal rangesand frequency of expression in tears and to establish therelationship of lacryglobin expression with family history.

The high frequency of expression of lacryglobin in coloncancer patients and the increased expression of lacryglobinin patients with a family history of cancer may point to apotential utility for lacryglobin screening. Colon cancer isfrequently difficult to detect by invasive techniques so anon-invasive technique that could measure cancer or cancerprogression would be very valuable.

Lacryglobin is a low molecular weight protein. In saliva,the presence of high molecular weight mucins interfereswith detection of small proteins like lacryglobin. The use oftear samples is advantageous as they do not require purifica-tion steps before analysis. Development of an immuno-chemical technique such as enzyme-linked immunosorbentassay (ELISA) may allow more accurate determination oflacryglobin protein levels in tear samples.

The volume of lacryglobin expressed was not correlatedwith cancer type in this small sample size. The patients whodonated tears for this study were generally at an advancedstage of cancer development with cancer metastasis to theliver and bone. This makes the determination of a relation-ship between lacryglobin expression and disease status moredifficult.

Identification of the source of lacryglobin will providemore information about its expression in tears and its rela-tionship to cancer. Lacryglobin is probably synthesized inthe lacrimal gland; however, it is possible that it has leakedfrom conjunctival blood vessels.

Lacryglobin was detected in tear samples of the majorityof patients with breast, colon and lung cancer types. Furtherinvestigation of lacryglobin expression in cancer patientsand age-matched controls is underway.

162 Evans et al.

Figure 1. Two-dimensional gel electrophoresis of human tearsfrom a patient with breast cancer. Proteins are separated accordingto isoelectric point (pI) along the horizontal axis (pH 4–7) andaccording to molecular weight (MW) along the vertical axis.Lacryglobin is marked with a circle. Tear lipocalins can be seen inthe centre of the gel at about 20 kDa.

Table 1. Frequency of lacryglobin detection in tear samples

Patient group n Frequency of detection in tears(% of subjects)

Breast cancer 8 88Lung cancer 6 83Colon cancer 5 100Ovary cancer 3 33Prostate cancer 1 100Control

No family history 3 33Family history 2 100

ACKNOWLEDGEMENTS

The authors would like to thank everyone who donatedtears for this study and to Research Nurse Jane Magnay andthe staff of the Oncology Day Centre for tear collection andpatient recruitment. The authors would also like to thank DrCraig Lewis and Dr David Goldstein for access to theirpatients, and Dr David Basseal for gel image analysis.

Victoria Evans was supported by a CRCERT/UWSNepean scholarship. This work was partly supported by theAustralian Government CRC scheme and has been facili-tated by access to the Australian Proteome Analysis FacilityEstablished under the Australian Government’s MajorNational Research Facilities Program.

REFERENCES

1. Molloy M et al. Electrophoresis 1997; 18: 2811–15.2. Watson M et al. Cancer Res. 1996; 56: 860–65.3. Min CJ. Cancer Res. 1998; 58: 4581–4.4. Aihara T et al. Breast Cancer Res. Treat. 1999; 58: 137–40.5. Becker R et al. Genomics 1998; 54: 70–74.6. Grunewald K et al. Lab. Invest. 2000; 80: 1071–7.7. Aihara T et al. Cancer Lett. 2000; 150: 79–84.8. Zhao C et al. Biochem. Biophys. Res. Commun. 1999; 256: 147–55.9. Watson MA et al. Oncogene 1998; 16: 817–24.

10. Lehrer R et al. FEBS Lett. 1998; 432: 163–7.11. Evans V et al. Clin. Exp. Ophthalmol. 2000; 28: 208–11.12. Yates JR. J. Mass Spectrom. 1998; 33: 1–19.13. Cottrell J. Pept. Res. 1994; 7: 115–24.

Tear lacryglobin screening for cancer 163

Table 2. Medical history of each tear sample donor

Patient Age Sex Lacryglobin/ Other sites Hormonal Current Treatmentno. (years) lipocalin volume therapy chemotherapy response

Breast cancer2 61 F 18.3 Bone, liver Nil Capecitabine Progressive disease3 58 F 18.6 Lung, other Nil Navelbine Progressive disease7 65 F 74.1 Bone Amridex Nil Progressive disease11 46 F 4.0 Bone Nil Nil Progressive disease12 50 F ND Lung, liver Nil Navelbine Stabilized15 75 F 1.6 Liver Nil Navelbine Progressive disease18 66 F 29.1 Bone Amridex Nilxaliplatin Stabilized28 NA F 33.6 NA NA NA NA

Colon cancer1 NA M 52.3 Liver Nil Oxaliplatin Nil5 69 M 66.4 Liver, lung Nil Nil Progressive disease13 56 M 3.7 Liver Nil Oxaliplatin Progressive disease16 48 M 6.9 Liver Nil Oxaliplatin Progressive disease17 70 F 29.1 Liver Nil Oxaliplatin Stabilized

Lung cancer4 54 M 20.7 Other Nil Cisplatin, navelbine Progressive disease6 77 M 53.3 Nil Nil Gemcitabine Progressive disease8 68 M 6.5 Nil Nil Nil Progressive disease9 40 M 25.4 Other Nil Gemcitabine, cisplatin Progressive disease10 NA F 34.4 Kidney Nil Nil Progressive disease14 75 F ND Lung Nil Navelbine Progressive disease

Ovarian cancer19 63 F ND Local Nil Carboplatin Stabilized20 44 F ND Local Nil Etoposide NA21 58 F 12.3 Local Tamoxifen Nil Progressive disease

Prostate cancer22 73 M 19.1 Bone Zoladex Nil Stabilized

Control (family history)23 39 M 11.3 Family history of breast cancer (mother) –24* 24 F 30.8* Family history of breast cancer (sister) –

Control (no family history)25 40 F 14.6 – – – –26* 24 F ND* – – – –27* 36 F ND* – – – –

ND, not detected; NA, not available; Stabilized, while metastatic cancer had previously been diagnosed in these patients, treatment hashalted the progression of the metastasis; Other, metastases to sites other than liver, lung, bone and kidney.

*Tear samples analysed using 1-D sodium dodecylsulfate polyacrylamide gel electrophoresis and matrix-assisted laser desorption–time of flight mass spectrometry. Comparison of lacryglobin/lipocalin volume ratio with other patients is not advised due to the difference in technique.