- Email: [email protected]
Anti-phosphoinositide auto-antibodies in sera of cancer patients: isotypic and immunochemical characterization
35
Cancer Letters, 66 (1992) 35-41 Elsevier Scientific Publishers Ireland Ltd.
Anti-phosphoinositide auto-antibodies in sera of cancer patients: isotypic and immunochemical characterization Sylvie Faiderbe,
Jean-Luc
Chagnaud
and Michel Geffard
Laboratoire d’lmmunologie et Pathologie, Uniuersitg de Bordeaux II, BP 66, 146, rue L&o Saignat, 33076 Bordeaux Cedex (France) (Received 26 May 1992) (Accepted 22 June 1992)
Keywords:
Summary
High levels of circulating anti-phosphatidylinositol (Ptdlns) auto-antibodies (auto-Ab) haue been previously found in sera of patients with malignant tumors. Using our ELlSA test, a high statistical leoel of immunological binding directed against Ptdlns was found in a large series of sera from patients with proliferative pathologies. In contrast to tumor markers, anti-Ptdlns auto-Ab did not vary whatever the histological grades, TMlV classification and patient’s age. These anti-Ptdlns auto-Ab were immunoglobulins (lg) of G class. Their specificity was evaluated by competition experiments in ELlSA and found to be rather high. An increase of these circulating auto-Ab reflected possible disturbances in Ptdlns turnouer and appearance of endogenous neoantigen with Ptdlns structure. Correspondence to: S. Faiderbe, Laboratoire d’lmmunologie et Pathologie, Universitk de Bordeaux II, BE 66, 146, rue Lko Saignat, 33076 Bordeaux Cedex, France. Abbreviations: Ab, antibodies; auto-Ab, auto-antibodies; Ag, immunoglobulin; Ins1,4,5P,, inositol-1,4,5 antigen; lg, t&phosphate; Ins, myo-inositol; OD, optical density; PBS, phosphatidylinositol; phosphate-buffered saline; Ptdlns, PdthwIP, phosphatidylinositol-4 monophosphate; PtdIns4.5Pz, phosphatidylinositol-4,5 b&phosphate; PtdChol, phosphatidylcholine; PtdEth, phosphatidylethanolamine; PdtSer, phosphatidylserine; PL, phospholipids; TH, thyroglobulin. 0304-3835/92/$05.00 Printed and Published
0 1992 Elsevier Scientific Publishers in Ireland
anti-phosphatidylinositol antibodies; cellular proliferation; avidity; specificity; cancers; hematological diseases Introduction
Increased levels of circulating antibodies (Ab) and auto-Ab have previously been reported in sera of patients with malignant tumors [ 1,4,15,17,18]. Auto-Ab were thus found to be directed against the neuronal nucleoprotein [l], nucleus, smooth muscle, actinomyosin-like component [ 17,181 and conjugated ‘benzo[a]pyrene-like’ structure [4]. Recently, circulating auto-Ab directed against a defined Ag, the PtdIns, in sera of patients with epithelial malignant tumors were found [5,6]. In the present report, using our immuno-enzymatic test, we have firstly extended our study to a large series of patients with epithelial cancers and hematologic diseases and secondly, characterized the antiPtdIns auto-Ab recognition site in sera of patients with proliferative pathologies using competition experiments. Materials
and Methods
Chemicals PtdIns, myo-inositol (Ins), inositol-1,4,5 trisphosphate (Ins 1 ,4,5P3), phosphatidylIreland Ltd
36
inositol-4 monophosphate (PtdIns 4P), phosphatidylinositol-4,5 bisphosphate (PtdIns 4,5P2) were purchased from Sigma Chemical Co (St. Louis, MO), as well as thyroglobulin (TH) and other phospholipids (PL): phosphatidylcholine (PtdChol) , phosphatidylethanolamine (PtdEth) and phosphatidylserine (PtdSer). Goat anti-human Ig labelled with horseradish peroxidase were obtained from Institut Pasteur Production. Patients Single serum specimen of patients with malignant (n = 332) or benign (n = 27) mammary tumors were supplied by the Fondation Bergoni$l of Bordeaux. Sera of patients with malignant (n = 115), non-malignant (n = 12) hematologic diseases or control sera (infectious hematologic diseases (n = 29)) were supplied by the Centre Franqois Magendie, H6pital Haut-LBvBque, CHU Bordeaux. Other control samples came from healthy blood donors (n = 66) and were obtained from the Centre de Transfusion Sanguine de Bordeaux and patients suffering from rheumatoid arthritis (n = 4) or neurologic diseases (Parkinson’s disease (n = 23), senile dementia (n = 15), epilepsy (n = 20), stroke (n = 4), Guillain-Barr6 syndrome (n = 6)) obtained from Hapital Pellegrin-Tripode, CHU Bordeaux. Immune-enzymatic test for detection and characterization of anti-Ptdlns auto-Ab in human sera Anti-PtdIns auto-Ab levels were checked using an ELISA assay as previously described [5,6] with minor modifications. Well-plates were coated with 200 ~1 of either PtdIns (20 pg/ml) or TH (20 pg/mI) in lo-’ M phosphate, 10m3 M CaC12 buffer, pH 7.4, for 16 h at 37OC. They were then rinsed 3 times with 10m2 M phosphate-buffered saline (PBS) and incubated for 2 h with 200 ~1 of human sera or purified Ig diluted at l/15 000 in PBS plus 10% glycerol and 0.01% lipid-free bovine serum albumin. Next, they were rinsed twice with PBS and incubated for 1 h in goat anti-
human IgG(y), A(a), M(p) labelled with horseradish peroxidase diluted at l/20 000. Peroxidase was assayed as previously described [5,6]. Experimental values read on PtdIns were corrected by subtracting blank values read on TH. Human anti-PtdIns auto-Ab avidity and specificity were evaluated by competition experiments between Ptdlns coated on well-plates and PtdIns, PtdIns-related compounds or other PL, diluted to 10 -5 M to 10 -‘* M, previously incubated with a dilution of randomly chosen cancer sera (n = 4) or purified IgG for 16 h at 4OC. Determination of Zg levels in human sera IgG, A and M levels were evaluated by an immunonephelometric method using a BNA nephelometer (Behring) , as previously described [4]. Purification of human Ig and determination of anti-Ptdlns Ab isotype Sera of patients with malignant or benign tumors (n = 12)) malignant hematologic diseases (n = 8) and healthy blood donors (n = lo), randomly chosen from the above described groups, were G-200 chromatographed after a (NH&SO4 precipitation as previously described [4,9]. Each purified fraction was assayed in the ELISA test using PtdIns coated on well-plates. The human IgG, A or M binding on Ptdlns was revealed by addition of goat Ig anti-human IgG(y), A(a), or M(p) labelled with horseradish peroxidase. Statistical methods Levels of circulating anti-PtdIns auto-Ab were given as means of adsorbances with the standard deviation resulting from triplicate determination. Differences between means of groups were evaluated by analysis of variance followed by Student’s t-test (a, n 1 30) or by the Mann-Whitney test (U-test, n < 30). Results
Detection of anti-Ptdlns auto-Ab in large series of human sera Using PtdIns coated on well-plates, anti-
37
(OD) value of patients minus mean OD value of controls/mean OD value of controls) was found to be respectively: 1.8, 1.78, 2.42 and 1.35 for malignant, benign mammary tumors non-malignant hematologic and malignant, diseases. Conversely, no statistically significant difference between the mean OD from sera of control subgroups was found.
binding on Ptdlns and the TNM system, tumor grade and type
immunological
O.D.
Fig. 1. Levels of anti-Ptdlns auto-Ab in large series of human sera. OD represent the binding of sera (l/15 000) on Ptdlns coated on well-plates. Histograms show mean OD value (grey) with standard deviations (dashed lines) from sera of patients with malignant (1, n = 332) or benign (2, n = 27) mammary tumors, patients with malignant (3, n = 115) or non malignant (4, n = 12) hematoiogic diseases and control group (5, n = 167 (66 healthy blood donors, 29 patients with infectious hematologic diseases, 72 suffering from neurologic or auto-immune diseases)).
Ptdlns auto-Ab levels were assayed (Fig. 1). Mean adsorbance values with standard deviations from sera of 332 patients with malignant mammary tumors (Fig. 1, histogram l), 27 patients with benign mammary tumors (Fig. 1, histogram 2), 115 patients with malignant hematologic diseases (Fig. 1, histogram 3), 12 patients with non-malignant hematologic diseases (Fig. 1, histogram 4) and 167 controls (66 healthy blood donors, 29 patients with infectious hematologic diseases, 72 suffering from neurologic or autoimmune diseases (Fig. 1, histogram 5)) were, respectively 0.40 l 0.12, 0.39 f 0.12, 0.48 f 0.09, 0.33 f 0.05 and 0.140 l 0.06. The differences between mean absorbances of patients and controls were highly significant for all patient groups with malignant (a = 0.001) or benign (ar = 0.001) mammary tumors and malignant (01 = 0.001) or non-malignant ((IL= 0.001) hematologic diseases. The ratio of means (r) defined as (mean optical density
The relation between the presence of antiPtdIns auto-Ab and the clinical stage according to the TNM system was investigated (Table I). This study was performed on sera of patients with malignant mammary tumors (n = 303). No relation was found between the amount of circulating anti-PtdIns auto-Ab in human sera and tumor size, local lymphatic ganglion metastasis or distant metastasis invasion. A post-surgery approach taking into account the relation between anti-PtdIns auto-Ab and the tumor histological type and grade was performed on breast cancer patients with operable primary tumors (n = 261). No relation was
Table 1. Immunological binding on Ptdlns patients sera and the TNM classification. Category
T
N
M
a
in cancer
Clinical classification
n
Mean OD zt S.D.b
Tl T2 T3 T4 NO Nl N2 N3 Ml
58 150 25 53 157 105 11 12 18
0.38 0.36 0.37 0.37 0.37 0.37 0.36 0.36 0.38
zt 0.12 f 0.09 f 0.11 zt 0.16 ztz 0.06 f 0.15 zt 0.09 zt 0.13 f 0.18
The differences between mean absorbances of Tl and T4 groups or Nl and N3 groups were not significant (a! > 0.05, U-test). aTNM, classification of malignant tumors (UICC. Geneva). bS.D., standard deviation.
38
Table II.
found between anti-PtdIns auto-Ab levels and tumour type or histological grades (Table II).
and assayed with our ELISA test (l/15 000) on PtdIns. Anti-PtdIns auto-Ab binding was revealed by different anti-human Ig isotypes labelled with peroxidase. For sera of patients with proliferative pathologies, immunological binding was observed for fractions corresponding to the IgG elution peak since Ig isotype was revealed with goat anti-human IgG(y) Ab. No immunological binding was seen at the IgA and IgM elution peak since Ig were revealed with goat anti-human IgA(cr), or IgM(p). At the same Ig concentration for sera of healthy blood donors, the immunological response of the IgG elution peak was lo-50-times less than sera from cancer patients. These results strongly indicate that anti-PtdIns auto-Ab found in sera of patients with proliferative pathologies, are Ig of class G.
fsotypic characterization of circulating human anti-Ptdlns auto-Ab Sera of patients with malignant or benign mammary tumours (n = 12), malignant hematologic diseases (n = 8) and healthy blood donors (n = 10) randomly chosen in the above described groups were G-ZOO chromatographed as described in Materials and Methods. After evaluation of Ig concentration, each fraction was diluted at 0.5 mg/mI
Relation between anti-Ptdlns auto-Ab leoels and ZgG amounts in human sera IgA, IgM and IgG amounts in sera of patients with malignant or benign mammary tumors (n = 51)) malignant hematologic diseases (n = 30) and controls (healthy blood donors (n = 6), infectious hematologic diseases (n = 5)) Parkinson’s diseases (n = 5) and stroke (n = 4)) were evaluated by nephelometric procedure. No significant
Immunological binding on Ptdlns found in cancer patients sera and tumor histological grades.
Histological grade
n
Mean OD f S.D.
1 2 3
41 123 96
0.38 f 0.11 0.36 * 0.14 0.36 +z 0.12
The study was performed on sera of breast cancer patients with operable primary tumor (n = 260). Each mean value represents the average of 3 experiments. The differences between mean absorbances of 1 to 3 histological grade groups were not significant (o > 0.05, U-test).
Table III.
Comparison of anti-Ptdlns auto-Ab levels and amounts of IgG in human sera.
Group”
n
Immunological binding on PtdIns (mean OD * S.D.)
IgG concentration (mg/ml)
Rb
Kt Controls Kl K2 K3
51 20 17 11 23
0.28 f 0.02 0.18 f 0.10 0.13 l 0.04 0.22 l 0.02 0.38 f 0.10
13.2 12.2 11.8 15.6 13.4
2.12 1.48 1.10 1.41 2.83
f 0.5 ztz0.6 ziz2.5 zt 3.9 f 3.5
“Kt, patients with malignant or benign mammary tumors randomly chosen in the population of 359 subjects tested; Kl, patients with malignant or benign mammary tumors having an anti-Ptdlns signal ~0.18 (control mean OD); K2, patients with malignant or benign mammary tumors having an anti-PtdIns signal comprised between 0.18 and 0.28; K3, patients with malignant or benign mammary tumors having an anti-PtdIns signal >0.28 (control mean OD + 2 x S.D.); Kt, Kl + K2 + K3; controls, healthy blood donors and patients having neurologic or auto-immune diseases. bR was defined as (mean OD value/mean IgG concentration value) x 100.
39
a
difference between patients with proliferative pathologies and controls was found for the Ig levels, whatever Ig isotype was considered. A complementary study was performed. Since anti-PtdIns auto-Ab were defined as Ig of G class, the IgG amounts found in human sera and the anti-PtdIns Ab levels were compared. A ratio R ((mean OD value/mean IgG concentration) x 100) was calculated for control and patient subgroups (Table III). Results show that R increased as OD and IgG amounts for all subgroups of patients with proliferative pathologies. Anti-PtdIns Ab levels increased more than IgG amounts, indicating a specificity of IgG binding on PtdIns and consequently an IgG-producing cell stimulation.
6.7 8
b
6 7. R
Human anti-Ptdlns auto-Ab aoidity and specificity Avidity of human anti-PtdIns Ab was performed by competition experiments as described in Materials and Methods. The displacement curves are plotted in Fig. 2. Avidity was defined as PtdIns concentration giving 50% inhibition of anti-PtdIns Ab/PtdIns binding. Best avidity was found with IgG from serum of breast cancer patient and occurred at a concentration of 4 x lo-’ M (Fig. 2a, curve 1). Identical displacements were observed with auto-Ab from sera of other patients with proliferative pathologies (breast cancer, gastrointestinal tract cancer, leukemia). AntiPtdIns Ab specificity was established by competition experiments between PtdIns coated on well-plates and PtdIns-related compounds or other PL, previously incubated with a given dilution of human IgG. Cross-reactivity ratios, defined as PtdIns concentration/PtdIns-related compounds concentration at half-displacement, showed that PtdIns 4P (Fig. 2a, curve 2) and PtdIns 4,5P2 (Fig. 2a, curve 5), Ins 1,4,5P3 (Fig. 2a, curve 3), and Ins (Fig. 2a, curve 4) were, respectively 17-, 23- and lOOfold less recognized than PtdIns. Other PL [PtdEth, PtdSer, PtdChol (curve 6, 7, 8, respectively)] were not recognized at all. These results showed that the glycerylinositol phosphate part of Ptdlns is immunodominant.
2.4 3.5 1 0
e m
0.5
o.oI -11
-10
-9
-8 log c
-7
-6
-5
Fi2.2. Displacement curves established from competition experiments between PtdIns coated on well-plates and Ptdlns (curve 1, 0), or Ptdlns-related compounds (PtdIns4P (curve 2, 0), Insl,4,5Ps (curve 3, Q, Ins (curve 4, A), Ptdlns4,5P2 (curve 5, A)) or other PL (PtdEth (curve 6, l), PtdSer (curve 7, x ), PtdChol (curve 8, 4) at various dilution, pre-incubated with purified IgG from breast cancer patient sera (a), or healthy blood donor sera (b) at l/15 000. B represents the OD value corresponding to the binding observed with different dilution of competitors. Bo represents the OD value corresponding to anti-PtdIns auto-Ab binding without competition.
When anti-PtdIns Ab of healthy blood donors were incubated with PtdIns (Fig. 2b, curve l), as well as other PtdIns-related compounds (Fig. 2b, curves 2, 3, 4, 5) or other PL (Fig. 2b, curves 6, 7, 8), only a small displacement was observed. This would be in relation with the polyspecificity of Ab detected in control sera, as described by other authors [2,4,16].
40
Discussion
References 1
Our results described significantly increased anti-PtdIns auto-Ab levels in sera of patients with proliferative pathologies, although total amounts of Ig were not significantly increased (Table III). This specific immunological binding was related to the IgG isotype. Competition experiments clearly demonstrated Ab recognition for the glycerylinositol phosphate part of PtdIns (Fig. 2a). In contrast to tumor markers, anti-PtdIns auto-Ab levels did not vary histological grades,TNM whatever the classification (Tables I, II) and patient’s age [6]. Moreover, control purified IgG fractions did not significantly recognize PtdIns or related compounds (Fig. 2b), demonstrating their polyspecificity, as shown by other authors [2,4,16]. Thus, the increase in anti-PtdIns auto-Ab levels and their specificity in sera of patients with proliferative pathologies, points to the activation of auto-reactive cells. It could be due to a possible disturbance in PtdIns turn over. Authors agreed that during cell proliferation PtdIns turn-over was increased [19] and that this key molecule was also widely implicated in protein anchorage [ 10,111. Others showed that in malignant tumors, phosphoinositide metabolites levels have been altered [ES]. Moreover, C and D phospholipases activities with PtdIns or PtdIns 4,5P2 as substrates, are many fold higher in cancer cells [3,8,12 - 141. The neosynthesis of an endogenous ‘PtdIns-like’ structure and consequently the appearance of anti-‘PtdIns-like’ auto-Ab during cell proliferation, might be understood by considering these events.
2
3
4
5
6
7
8
9
10
11
12
Acknowledgements The authors thank Dr. J. W*afflart (FondaBordeaux) and Prof. A. tion BergoniQ, Broustet (Centre Francois Magendie, HautLeveque, Bordeaux) for supplying cancer patients’ sera. This work was supported by grants from the ‘Association pour la Recherche contre le Cancer’ (ARC), Villejuif, France and the ‘Conseil Regional d’Aquitaine’.
13
14
15
Anderson, R.E., Rosenblum, M.K., Grauss, F., Wilfy, R.G. and Posner, J.B. (1988) Autoantibodies in paraneoplastic syndromes associated with small-cell lung cancer. Neurology, 38, 1391- 1398. Avrameas, S. (1988) Natural antibodies and autoantibodies. In: ISI Atlas of Science/Immunology, pp. 111-115. Balsinde, J., Diez, E. and Mollinedo, F. (1988) A pathway for generation of second messenger. Biochem. Biophys. Res. Commun., 154, 502-508. Chagnaud, J.L., Faiderbe, S. and Geffard, M. (1992) Identification and immunochemical characterization of IgA in sera of patients with mammary tumors. Int. J. Cancer, 50, 395-401. Faiderbe, S., Chagnaud, J.L., Wafflart, J. and Geffard, M. (1990) Autoanticorps ding& contre un lipide membranaire dans les &rums de patients porteurs de tumeurs malignes. C. R. Acad. Sci. Parts, 310, 49- 52. Faiderbe, S., Chagnaud, J.L., Charrier, M.C., Peyron, M.A. and Geffard, M. (1991) Autoantibodies directed against lipid membrane components in sera of patients with malignant tumors. Cancer Detec. Prev., 15, 199-203. Foong, Y.T., Cheng, M.M., Sam, C.K., Dillinger, J., Hinderer, W. and Prasad, H. (1990) Serum and salivary IgA antibodies against a defined epitope of the EpsteinBarr-virus nuclear antigen (EBNA) are elevated in carcinoma. Int. J. Cancer, 45, nasopharyngeal 1061- 1064. Lalwani, N., Hylemon, P. and Stem, S. (1991) Altered levels of phosphoinositide metabolites and activation of guanine-nucleotide dependent phospholipase C in rat hepatic tumors. J. Cell. Biol., 147, 354-361. Loirand, G., Faiderbe, S., Baron, A., Geffard, M. and Mironneau, J. (1992) Auto anti-phosphatidylinositol antibodies specifically inhibit noradrenaline effects on Ca2+ and Cl - channels in rat portal vein myocytes. J. Biol. Chem., 267, 4312-4316. Low, M.G. and Saltiel, A.R. (1988) Structural and functional roles of glycosyl-phosphatidylinositol in membranes. Science, 239, 268 - 275. Low, M.G. (1989) Glycosyl phosphatidylinositol: a versatile anchor for ceil-surface proteins. The FASEB J., 3, 1600 - 1608. Pevella, F.. Jankiewieck, R. and Dandrouin, E. (1991) Phospholipase C from human melanoma: purification and characterization of a phosphatidylinositol-selective enzyme. Biochem. Biophys. Acta, 1076, 209-214. Rillema, J.A. (1986) Phospholipase C activity in normal rat mammary tissues and in DMBA-induced rat mammary tumors. Proc. Sot. Expt. Biol. Med., 181, 450-453. Rillema, J.A. (1989) Possible role of phospholipase C in the regulation of cell division in normal and neoplasic cells. Med. Hypotheses, 29, 1 - 4. Tannenberg, A.E.G., Miiller, H.K., Gauche, M.N. and Nairn, R.C. (1973) Incidence of auto-antibodies in cancer patients. Clin. Exp. Immunol., 15, 153- 156.
41
16
17
Terninck, T. and Avrameas, S. (1986) Murine natural their monoclonal auto-antibodies: a study of polyspecificittes and their affinities. fmmunol. Rev., 94, 99 - 112. Wasserman, J., Glas, V. and Blomgren, H. (1975) Autoantibodies in patients with carcinoma of the breast. Correlation with prognosis. Clin. Exp. fmmunol., 19, 417 -422.
18 19
Whitehouse, J.M.A. (1973) Circulating antibodies in human malignant diseases. Br. J. Cancer, 28, 170 - 174. Whitman, M. and Cantley, L. (1988) Phosphoinositide metabolism and the control of cell proliferation. Biochem. Biophys. Acta, 948, 327 - 345.
Cancer Letters, 66 (1992) 35-41 Elsevier Scientific Publishers Ireland Ltd.
Anti-phosphoinositide auto-antibodies in sera of cancer patients: isotypic and immunochemical characterization Sylvie Faiderbe,
Jean-Luc
Chagnaud
and Michel Geffard
Laboratoire d’lmmunologie et Pathologie, Uniuersitg de Bordeaux II, BP 66, 146, rue L&o Saignat, 33076 Bordeaux Cedex (France) (Received 26 May 1992) (Accepted 22 June 1992)
Keywords:
Summary
High levels of circulating anti-phosphatidylinositol (Ptdlns) auto-antibodies (auto-Ab) haue been previously found in sera of patients with malignant tumors. Using our ELlSA test, a high statistical leoel of immunological binding directed against Ptdlns was found in a large series of sera from patients with proliferative pathologies. In contrast to tumor markers, anti-Ptdlns auto-Ab did not vary whatever the histological grades, TMlV classification and patient’s age. These anti-Ptdlns auto-Ab were immunoglobulins (lg) of G class. Their specificity was evaluated by competition experiments in ELlSA and found to be rather high. An increase of these circulating auto-Ab reflected possible disturbances in Ptdlns turnouer and appearance of endogenous neoantigen with Ptdlns structure. Correspondence to: S. Faiderbe, Laboratoire d’lmmunologie et Pathologie, Universitk de Bordeaux II, BE 66, 146, rue Lko Saignat, 33076 Bordeaux Cedex, France. Abbreviations: Ab, antibodies; auto-Ab, auto-antibodies; Ag, immunoglobulin; Ins1,4,5P,, inositol-1,4,5 antigen; lg, t&phosphate; Ins, myo-inositol; OD, optical density; PBS, phosphatidylinositol; phosphate-buffered saline; Ptdlns, PdthwIP, phosphatidylinositol-4 monophosphate; PtdIns4.5Pz, phosphatidylinositol-4,5 b&phosphate; PtdChol, phosphatidylcholine; PtdEth, phosphatidylethanolamine; PdtSer, phosphatidylserine; PL, phospholipids; TH, thyroglobulin. 0304-3835/92/$05.00 Printed and Published
0 1992 Elsevier Scientific Publishers in Ireland
anti-phosphatidylinositol antibodies; cellular proliferation; avidity; specificity; cancers; hematological diseases Introduction
Increased levels of circulating antibodies (Ab) and auto-Ab have previously been reported in sera of patients with malignant tumors [ 1,4,15,17,18]. Auto-Ab were thus found to be directed against the neuronal nucleoprotein [l], nucleus, smooth muscle, actinomyosin-like component [ 17,181 and conjugated ‘benzo[a]pyrene-like’ structure [4]. Recently, circulating auto-Ab directed against a defined Ag, the PtdIns, in sera of patients with epithelial malignant tumors were found [5,6]. In the present report, using our immuno-enzymatic test, we have firstly extended our study to a large series of patients with epithelial cancers and hematologic diseases and secondly, characterized the antiPtdIns auto-Ab recognition site in sera of patients with proliferative pathologies using competition experiments. Materials
and Methods
Chemicals PtdIns, myo-inositol (Ins), inositol-1,4,5 trisphosphate (Ins 1 ,4,5P3), phosphatidylIreland Ltd
36
inositol-4 monophosphate (PtdIns 4P), phosphatidylinositol-4,5 bisphosphate (PtdIns 4,5P2) were purchased from Sigma Chemical Co (St. Louis, MO), as well as thyroglobulin (TH) and other phospholipids (PL): phosphatidylcholine (PtdChol) , phosphatidylethanolamine (PtdEth) and phosphatidylserine (PtdSer). Goat anti-human Ig labelled with horseradish peroxidase were obtained from Institut Pasteur Production. Patients Single serum specimen of patients with malignant (n = 332) or benign (n = 27) mammary tumors were supplied by the Fondation Bergoni$l of Bordeaux. Sera of patients with malignant (n = 115), non-malignant (n = 12) hematologic diseases or control sera (infectious hematologic diseases (n = 29)) were supplied by the Centre Franqois Magendie, H6pital Haut-LBvBque, CHU Bordeaux. Other control samples came from healthy blood donors (n = 66) and were obtained from the Centre de Transfusion Sanguine de Bordeaux and patients suffering from rheumatoid arthritis (n = 4) or neurologic diseases (Parkinson’s disease (n = 23), senile dementia (n = 15), epilepsy (n = 20), stroke (n = 4), Guillain-Barr6 syndrome (n = 6)) obtained from Hapital Pellegrin-Tripode, CHU Bordeaux. Immune-enzymatic test for detection and characterization of anti-Ptdlns auto-Ab in human sera Anti-PtdIns auto-Ab levels were checked using an ELISA assay as previously described [5,6] with minor modifications. Well-plates were coated with 200 ~1 of either PtdIns (20 pg/ml) or TH (20 pg/mI) in lo-’ M phosphate, 10m3 M CaC12 buffer, pH 7.4, for 16 h at 37OC. They were then rinsed 3 times with 10m2 M phosphate-buffered saline (PBS) and incubated for 2 h with 200 ~1 of human sera or purified Ig diluted at l/15 000 in PBS plus 10% glycerol and 0.01% lipid-free bovine serum albumin. Next, they were rinsed twice with PBS and incubated for 1 h in goat anti-
human IgG(y), A(a), M(p) labelled with horseradish peroxidase diluted at l/20 000. Peroxidase was assayed as previously described [5,6]. Experimental values read on PtdIns were corrected by subtracting blank values read on TH. Human anti-PtdIns auto-Ab avidity and specificity were evaluated by competition experiments between Ptdlns coated on well-plates and PtdIns, PtdIns-related compounds or other PL, diluted to 10 -5 M to 10 -‘* M, previously incubated with a dilution of randomly chosen cancer sera (n = 4) or purified IgG for 16 h at 4OC. Determination of Zg levels in human sera IgG, A and M levels were evaluated by an immunonephelometric method using a BNA nephelometer (Behring) , as previously described [4]. Purification of human Ig and determination of anti-Ptdlns Ab isotype Sera of patients with malignant or benign tumors (n = 12)) malignant hematologic diseases (n = 8) and healthy blood donors (n = lo), randomly chosen from the above described groups, were G-200 chromatographed after a (NH&SO4 precipitation as previously described [4,9]. Each purified fraction was assayed in the ELISA test using PtdIns coated on well-plates. The human IgG, A or M binding on Ptdlns was revealed by addition of goat Ig anti-human IgG(y), A(a), or M(p) labelled with horseradish peroxidase. Statistical methods Levels of circulating anti-PtdIns auto-Ab were given as means of adsorbances with the standard deviation resulting from triplicate determination. Differences between means of groups were evaluated by analysis of variance followed by Student’s t-test (a, n 1 30) or by the Mann-Whitney test (U-test, n < 30). Results
Detection of anti-Ptdlns auto-Ab in large series of human sera Using PtdIns coated on well-plates, anti-
37
(OD) value of patients minus mean OD value of controls/mean OD value of controls) was found to be respectively: 1.8, 1.78, 2.42 and 1.35 for malignant, benign mammary tumors non-malignant hematologic and malignant, diseases. Conversely, no statistically significant difference between the mean OD from sera of control subgroups was found.
binding on Ptdlns and the TNM system, tumor grade and type
immunological
O.D.
Fig. 1. Levels of anti-Ptdlns auto-Ab in large series of human sera. OD represent the binding of sera (l/15 000) on Ptdlns coated on well-plates. Histograms show mean OD value (grey) with standard deviations (dashed lines) from sera of patients with malignant (1, n = 332) or benign (2, n = 27) mammary tumors, patients with malignant (3, n = 115) or non malignant (4, n = 12) hematoiogic diseases and control group (5, n = 167 (66 healthy blood donors, 29 patients with infectious hematologic diseases, 72 suffering from neurologic or auto-immune diseases)).
Ptdlns auto-Ab levels were assayed (Fig. 1). Mean adsorbance values with standard deviations from sera of 332 patients with malignant mammary tumors (Fig. 1, histogram l), 27 patients with benign mammary tumors (Fig. 1, histogram 2), 115 patients with malignant hematologic diseases (Fig. 1, histogram 3), 12 patients with non-malignant hematologic diseases (Fig. 1, histogram 4) and 167 controls (66 healthy blood donors, 29 patients with infectious hematologic diseases, 72 suffering from neurologic or autoimmune diseases (Fig. 1, histogram 5)) were, respectively 0.40 l 0.12, 0.39 f 0.12, 0.48 f 0.09, 0.33 f 0.05 and 0.140 l 0.06. The differences between mean absorbances of patients and controls were highly significant for all patient groups with malignant (a = 0.001) or benign (ar = 0.001) mammary tumors and malignant (01 = 0.001) or non-malignant ((IL= 0.001) hematologic diseases. The ratio of means (r) defined as (mean optical density
The relation between the presence of antiPtdIns auto-Ab and the clinical stage according to the TNM system was investigated (Table I). This study was performed on sera of patients with malignant mammary tumors (n = 303). No relation was found between the amount of circulating anti-PtdIns auto-Ab in human sera and tumor size, local lymphatic ganglion metastasis or distant metastasis invasion. A post-surgery approach taking into account the relation between anti-PtdIns auto-Ab and the tumor histological type and grade was performed on breast cancer patients with operable primary tumors (n = 261). No relation was
Table 1. Immunological binding on Ptdlns patients sera and the TNM classification. Category
T
N
M
a
in cancer
Clinical classification
n
Mean OD zt S.D.b
Tl T2 T3 T4 NO Nl N2 N3 Ml
58 150 25 53 157 105 11 12 18
0.38 0.36 0.37 0.37 0.37 0.37 0.36 0.36 0.38
zt 0.12 f 0.09 f 0.11 zt 0.16 ztz 0.06 f 0.15 zt 0.09 zt 0.13 f 0.18
The differences between mean absorbances of Tl and T4 groups or Nl and N3 groups were not significant (a! > 0.05, U-test). aTNM, classification of malignant tumors (UICC. Geneva). bS.D., standard deviation.
38
Table II.
found between anti-PtdIns auto-Ab levels and tumour type or histological grades (Table II).
and assayed with our ELISA test (l/15 000) on PtdIns. Anti-PtdIns auto-Ab binding was revealed by different anti-human Ig isotypes labelled with peroxidase. For sera of patients with proliferative pathologies, immunological binding was observed for fractions corresponding to the IgG elution peak since Ig isotype was revealed with goat anti-human IgG(y) Ab. No immunological binding was seen at the IgA and IgM elution peak since Ig were revealed with goat anti-human IgA(cr), or IgM(p). At the same Ig concentration for sera of healthy blood donors, the immunological response of the IgG elution peak was lo-50-times less than sera from cancer patients. These results strongly indicate that anti-PtdIns auto-Ab found in sera of patients with proliferative pathologies, are Ig of class G.
fsotypic characterization of circulating human anti-Ptdlns auto-Ab Sera of patients with malignant or benign mammary tumours (n = 12), malignant hematologic diseases (n = 8) and healthy blood donors (n = 10) randomly chosen in the above described groups were G-ZOO chromatographed as described in Materials and Methods. After evaluation of Ig concentration, each fraction was diluted at 0.5 mg/mI
Relation between anti-Ptdlns auto-Ab leoels and ZgG amounts in human sera IgA, IgM and IgG amounts in sera of patients with malignant or benign mammary tumors (n = 51)) malignant hematologic diseases (n = 30) and controls (healthy blood donors (n = 6), infectious hematologic diseases (n = 5)) Parkinson’s diseases (n = 5) and stroke (n = 4)) were evaluated by nephelometric procedure. No significant
Immunological binding on Ptdlns found in cancer patients sera and tumor histological grades.
Histological grade
n
Mean OD f S.D.
1 2 3
41 123 96
0.38 f 0.11 0.36 * 0.14 0.36 +z 0.12
The study was performed on sera of breast cancer patients with operable primary tumor (n = 260). Each mean value represents the average of 3 experiments. The differences between mean absorbances of 1 to 3 histological grade groups were not significant (o > 0.05, U-test).
Table III.
Comparison of anti-Ptdlns auto-Ab levels and amounts of IgG in human sera.
Group”
n
Immunological binding on PtdIns (mean OD * S.D.)
IgG concentration (mg/ml)
Rb
Kt Controls Kl K2 K3
51 20 17 11 23
0.28 f 0.02 0.18 f 0.10 0.13 l 0.04 0.22 l 0.02 0.38 f 0.10
13.2 12.2 11.8 15.6 13.4
2.12 1.48 1.10 1.41 2.83
f 0.5 ztz0.6 ziz2.5 zt 3.9 f 3.5
“Kt, patients with malignant or benign mammary tumors randomly chosen in the population of 359 subjects tested; Kl, patients with malignant or benign mammary tumors having an anti-Ptdlns signal ~0.18 (control mean OD); K2, patients with malignant or benign mammary tumors having an anti-PtdIns signal comprised between 0.18 and 0.28; K3, patients with malignant or benign mammary tumors having an anti-PtdIns signal >0.28 (control mean OD + 2 x S.D.); Kt, Kl + K2 + K3; controls, healthy blood donors and patients having neurologic or auto-immune diseases. bR was defined as (mean OD value/mean IgG concentration value) x 100.
39
a
difference between patients with proliferative pathologies and controls was found for the Ig levels, whatever Ig isotype was considered. A complementary study was performed. Since anti-PtdIns auto-Ab were defined as Ig of G class, the IgG amounts found in human sera and the anti-PtdIns Ab levels were compared. A ratio R ((mean OD value/mean IgG concentration) x 100) was calculated for control and patient subgroups (Table III). Results show that R increased as OD and IgG amounts for all subgroups of patients with proliferative pathologies. Anti-PtdIns Ab levels increased more than IgG amounts, indicating a specificity of IgG binding on PtdIns and consequently an IgG-producing cell stimulation.
6.7 8
b
6 7. R
Human anti-Ptdlns auto-Ab aoidity and specificity Avidity of human anti-PtdIns Ab was performed by competition experiments as described in Materials and Methods. The displacement curves are plotted in Fig. 2. Avidity was defined as PtdIns concentration giving 50% inhibition of anti-PtdIns Ab/PtdIns binding. Best avidity was found with IgG from serum of breast cancer patient and occurred at a concentration of 4 x lo-’ M (Fig. 2a, curve 1). Identical displacements were observed with auto-Ab from sera of other patients with proliferative pathologies (breast cancer, gastrointestinal tract cancer, leukemia). AntiPtdIns Ab specificity was established by competition experiments between PtdIns coated on well-plates and PtdIns-related compounds or other PL, previously incubated with a given dilution of human IgG. Cross-reactivity ratios, defined as PtdIns concentration/PtdIns-related compounds concentration at half-displacement, showed that PtdIns 4P (Fig. 2a, curve 2) and PtdIns 4,5P2 (Fig. 2a, curve 5), Ins 1,4,5P3 (Fig. 2a, curve 3), and Ins (Fig. 2a, curve 4) were, respectively 17-, 23- and lOOfold less recognized than PtdIns. Other PL [PtdEth, PtdSer, PtdChol (curve 6, 7, 8, respectively)] were not recognized at all. These results showed that the glycerylinositol phosphate part of Ptdlns is immunodominant.
2.4 3.5 1 0
e m
0.5
o.oI -11
-10
-9
-8 log c
-7
-6
-5
Fi2.2. Displacement curves established from competition experiments between PtdIns coated on well-plates and Ptdlns (curve 1, 0), or Ptdlns-related compounds (PtdIns4P (curve 2, 0), Insl,4,5Ps (curve 3, Q, Ins (curve 4, A), Ptdlns4,5P2 (curve 5, A)) or other PL (PtdEth (curve 6, l), PtdSer (curve 7, x ), PtdChol (curve 8, 4) at various dilution, pre-incubated with purified IgG from breast cancer patient sera (a), or healthy blood donor sera (b) at l/15 000. B represents the OD value corresponding to the binding observed with different dilution of competitors. Bo represents the OD value corresponding to anti-PtdIns auto-Ab binding without competition.
When anti-PtdIns Ab of healthy blood donors were incubated with PtdIns (Fig. 2b, curve l), as well as other PtdIns-related compounds (Fig. 2b, curves 2, 3, 4, 5) or other PL (Fig. 2b, curves 6, 7, 8), only a small displacement was observed. This would be in relation with the polyspecificity of Ab detected in control sera, as described by other authors [2,4,16].
40
Discussion
References 1
Our results described significantly increased anti-PtdIns auto-Ab levels in sera of patients with proliferative pathologies, although total amounts of Ig were not significantly increased (Table III). This specific immunological binding was related to the IgG isotype. Competition experiments clearly demonstrated Ab recognition for the glycerylinositol phosphate part of PtdIns (Fig. 2a). In contrast to tumor markers, anti-PtdIns auto-Ab levels did not vary histological grades,TNM whatever the classification (Tables I, II) and patient’s age [6]. Moreover, control purified IgG fractions did not significantly recognize PtdIns or related compounds (Fig. 2b), demonstrating their polyspecificity, as shown by other authors [2,4,16]. Thus, the increase in anti-PtdIns auto-Ab levels and their specificity in sera of patients with proliferative pathologies, points to the activation of auto-reactive cells. It could be due to a possible disturbance in PtdIns turn over. Authors agreed that during cell proliferation PtdIns turn-over was increased [19] and that this key molecule was also widely implicated in protein anchorage [ 10,111. Others showed that in malignant tumors, phosphoinositide metabolites levels have been altered [ES]. Moreover, C and D phospholipases activities with PtdIns or PtdIns 4,5P2 as substrates, are many fold higher in cancer cells [3,8,12 - 141. The neosynthesis of an endogenous ‘PtdIns-like’ structure and consequently the appearance of anti-‘PtdIns-like’ auto-Ab during cell proliferation, might be understood by considering these events.
2
3
4
5
6
7
8
9
10
11
12
Acknowledgements The authors thank Dr. J. W*afflart (FondaBordeaux) and Prof. A. tion BergoniQ, Broustet (Centre Francois Magendie, HautLeveque, Bordeaux) for supplying cancer patients’ sera. This work was supported by grants from the ‘Association pour la Recherche contre le Cancer’ (ARC), Villejuif, France and the ‘Conseil Regional d’Aquitaine’.
13
14
15
Anderson, R.E., Rosenblum, M.K., Grauss, F., Wilfy, R.G. and Posner, J.B. (1988) Autoantibodies in paraneoplastic syndromes associated with small-cell lung cancer. Neurology, 38, 1391- 1398. Avrameas, S. (1988) Natural antibodies and autoantibodies. In: ISI Atlas of Science/Immunology, pp. 111-115. Balsinde, J., Diez, E. and Mollinedo, F. (1988) A pathway for generation of second messenger. Biochem. Biophys. Res. Commun., 154, 502-508. Chagnaud, J.L., Faiderbe, S. and Geffard, M. (1992) Identification and immunochemical characterization of IgA in sera of patients with mammary tumors. Int. J. Cancer, 50, 395-401. Faiderbe, S., Chagnaud, J.L., Wafflart, J. and Geffard, M. (1990) Autoanticorps ding& contre un lipide membranaire dans les &rums de patients porteurs de tumeurs malignes. C. R. Acad. Sci. Parts, 310, 49- 52. Faiderbe, S., Chagnaud, J.L., Charrier, M.C., Peyron, M.A. and Geffard, M. (1991) Autoantibodies directed against lipid membrane components in sera of patients with malignant tumors. Cancer Detec. Prev., 15, 199-203. Foong, Y.T., Cheng, M.M., Sam, C.K., Dillinger, J., Hinderer, W. and Prasad, H. (1990) Serum and salivary IgA antibodies against a defined epitope of the EpsteinBarr-virus nuclear antigen (EBNA) are elevated in carcinoma. Int. J. Cancer, 45, nasopharyngeal 1061- 1064. Lalwani, N., Hylemon, P. and Stem, S. (1991) Altered levels of phosphoinositide metabolites and activation of guanine-nucleotide dependent phospholipase C in rat hepatic tumors. J. Cell. Biol., 147, 354-361. Loirand, G., Faiderbe, S., Baron, A., Geffard, M. and Mironneau, J. (1992) Auto anti-phosphatidylinositol antibodies specifically inhibit noradrenaline effects on Ca2+ and Cl - channels in rat portal vein myocytes. J. Biol. Chem., 267, 4312-4316. Low, M.G. and Saltiel, A.R. (1988) Structural and functional roles of glycosyl-phosphatidylinositol in membranes. Science, 239, 268 - 275. Low, M.G. (1989) Glycosyl phosphatidylinositol: a versatile anchor for ceil-surface proteins. The FASEB J., 3, 1600 - 1608. Pevella, F.. Jankiewieck, R. and Dandrouin, E. (1991) Phospholipase C from human melanoma: purification and characterization of a phosphatidylinositol-selective enzyme. Biochem. Biophys. Acta, 1076, 209-214. Rillema, J.A. (1986) Phospholipase C activity in normal rat mammary tissues and in DMBA-induced rat mammary tumors. Proc. Sot. Expt. Biol. Med., 181, 450-453. Rillema, J.A. (1989) Possible role of phospholipase C in the regulation of cell division in normal and neoplasic cells. Med. Hypotheses, 29, 1 - 4. Tannenberg, A.E.G., Miiller, H.K., Gauche, M.N. and Nairn, R.C. (1973) Incidence of auto-antibodies in cancer patients. Clin. Exp. Immunol., 15, 153- 156.
41
16
17
Terninck, T. and Avrameas, S. (1986) Murine natural their monoclonal auto-antibodies: a study of polyspecificittes and their affinities. fmmunol. Rev., 94, 99 - 112. Wasserman, J., Glas, V. and Blomgren, H. (1975) Autoantibodies in patients with carcinoma of the breast. Correlation with prognosis. Clin. Exp. fmmunol., 19, 417 -422.
18 19
Whitehouse, J.M.A. (1973) Circulating antibodies in human malignant diseases. Br. J. Cancer, 28, 170 - 174. Whitman, M. and Cantley, L. (1988) Phosphoinositide metabolism and the control of cell proliferation. Biochem. Biophys. Acta, 948, 327 - 345.