γ-glutamyl transpeptidase, an ecto-enzyme regulator of intracellular redox potential, is a component of TM4 signal transduction complexes

Eur. J. Immunol. 1998. 28: 4123–4129

GGT is a component of TM4 protein complexes

4123

q -glutamyl transpeptidase, an ecto-enzyme regulator of intracellular redox potential, is a component of TM4 signal transduction complexes Timothy C. Nichols1, Joel M. Guthridge1, David R. Karp2, Hector Molina3, Dana R. Fletcher1 and V. Michael Holers1 1

Departments of Medicine and Immunology, Division of Rheumatology, University of Colorado Health Sciences Center, Denver, USA 2 Simmons Arthritis Research Center, University of Texas Southwestern Medical Center, Dallas, USA 3 Department of Medicine, Division of Rheumatology, Washington University School of Medicine, St. Louis, USA

CD21 (C3dg/EBV receptor) is physically associated on B cells with a complex of proteins that includes CD19 and the widely distributed tetraspan 4 (TM4) family protein CD81 as well as other TM4 proteins (CD53, CD37 and CD82). Monoclonal antibodies (mAb) were generated that blocked homotypic adhesion induced by CD21 ligands in the human B cell line Balm-1. One inhibitory mAb (3A8) was found to recognize the ecto-enzyme + -glutamyl transpeptidase (GGT), a membrane protein involved in recycling extracellular glutathione and regulating intracellular redox potential. Molecular associations between GGT and TM4 proteins CD81, CD53 and CD82, in addition to CD21 and CD19, were detected by co-precipitation and co-capping analysis. GGT is expressed on several B and T cell lines independently of CD21 expression. These results demonstrate that GGT is a component of widely distributed TM4 complexes, and that on B cells the GGT-containing TM4 complexes also contain CD19 and CD21. Key words: Tetraspan 4 / Complement / Redox regulation / CD19/CD21 / Lymphocyte

1 Introduction The complement receptor type 2 (CD21) is a 145-kDa glycoprotein expressed on mature B cells whose expression is required for the development of normal humoral immunity [1, 2]. CD21 was initially described as part of a complex which includes the signal-transducing molecule CD19, the IFN-inducible protein Leu 13, and the tetraspan 4 (TM4) family protein CD81 [3]. Recently, CD21 has been shown to be a component of even larger multicomponent complexes that also include MHC class II glycoproteins and the TM4 proteins CD37, CD53 and CD82 [4, 5]. TM4 proteins currently have no known physiological ligands, and their biologic roles are as yet ill defined [5]. However, mice in which CD81 expression

[I 18413] Abbreviations: TM4: Tetraspan GSH: SCR: Short consensus repeat

Glutathione

© WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998

Received Revised Accepted

6/5/98 24/8/98 2/9/98

has been eliminated demonstrate defects in Tdependent B cell responses, decreased CD19 expression and Ca2+ flux, and enhanced type II T-independent responses [6]. Homotypic adhesion is induced in B cell lines following ligation of CD21 with either polymerized C3dg or mAb OKB7 [7]. We screened hybridoma supernatants produced from the fusion of spleen cells from mice immunized with B cell lines based on their ability to inhibit ligand-induced adhesion. One of the mAb was found to recognize the ecto-enzyme + -glutamyl transpeptidase (GGT). GGT is a membrane ecto-enzyme that participates in the glutathione (GSH) recycling pathway and thereby plays an important role in the regulation of intracellular GSH levels and the cellular redox state [8]. Although GGT has been previously demonstrated on a number of hematopoietic cell lineages by enzymatic or histochemical techniques [9], it is not a commonly appreciated lymphocyte membrane protein and does not yet have a cluster designation (CD) number. 0014-2980/98/1212-4123$17.50 + .50/0

4124

T. C. Nichols et al.

Eur. J. Immunol. 1998. 28: 4123–4129

mAb 3A8 immunoprecipitated a heterodimer from Triton X-100 extracts of biotin-labeled (Fig. 1 A) or biosynthetically labeled (Fig. 1 C) Balm-1 cells. Each subunit had multiple isoforms varying in isoelectric point. To identify the heterodimer, the complex was purified by affinity chromatography. N-terminal sequencing of the large subunit yielded the 16-amino acid sequence MKKKLVVLGLLAVVLV, which is identical to the published sequence of GGT [10]. GGT is synthesized as a single-chain protein that is posttranslationally cleaved, resulting in a type 2 membranespanning protein composed of non-covalently associated large and small subunits similar to those identified by mAb 3A8 [8]. Two criteria were used to assure that mAb 3A8 directly binds to GGT. First, Balm-1 cell lysates solubilized with Triton X-100 that were immunoprecipitated with mAb 3A8 showed a specific decrease of approximately 70 % in the amount of GGT activity in comparison to controls (data not shown). Second, after transfection with a cDNA encoding human GGT [10], mouse fibroblasts expressed high levels of mAb 3A8reactive surface protein (data not shown).

2.2 Expression of GGT on B lymphocytes and B cell lines

Figure 1. (A) mAb 3A8 immunoprecipitates a complex of two proteins (arrows). Solubilized membrane preparations from biotinylated Balm-1 cells were immunoprecipitated with mAb UPC-10 (lane-1), mAb HB5 (anti-CD21) (lane 2) or mAb 3A8 (lane 3) and analyzed by reducing SDS-PAGE on a 12 % polyacrylamide gel. Weak bands at the top of lanes 1 and 3 represent gel loading artifacts. Solubilized membrane preparations were also prepared from Balm-1 cells labeled for 24 h with [35S] cysteine and [35S] methionine and immunoprecipitated with either control mAb UPC-10 (B) or mAb 3A8 (C) followed by two-dimensional IEF/SDS-PAGE analysis. Molecular weights are indicated to the left.

Table 1 summarizes the expression of GGT as compared to other relevant proteins on these cell lines. In addition to demonstrating variability in expression, these results

Table 1. Expression of GGT and other informative CD19/ CD21 complex members on cell linesa) GGT

CD21 CD19 Leu 13 CD81 CD53 CD82

Pre-B cells Nall 1

−

−

+

+

+

ND ND

Nalm 6

−

−

+

+

+

ND ND

+

+

+

+

Mature B cells

2 Results 2.1 Generation of mAb 3A8 and identification of the antigen recognized Three mAb were generated that inhibited CD21-induced homotypic adhesion. Two recognized a non-ligandbinding domain of CD21, while the third (3A8) did not react with a well-characterized protein. The effects of mAb 3A8 are specific for CD21, as homotypic adhesion induced by ligation of CD19 and CD81 is not altered (data not shown).

Balm-1

+

+

+

Daudi

−

+

+

+

+

+

−

Jiyoye

+

+

+

+

+

+

ND

Ramos

−

+

+

+

+

+

−

Raji

+

+

+

+

+

+

+

CEM

+

−

−

ND

+

+

+

Jurkat

+

+

−

ND

+

+

+

T cells

a) + indicates expression on G 5 % of the cells as assessed by flow cytometry.

Eur. J. Immunol. 1998. 28: 4123–4129

also show that the expression of CD21 is not dependent upon the expression of GGT, and vice versa. This point is illustrated by the observation that Daudi cells express CD21 but not GGT, and the T cell line CEM expresses GGT but not CD21. We confirmed the lack of expression of GGT on Daudi cells using both enzymatic assay and immunoprecipitation analysis (data not shown). Finally, 10 % of peripheral blood B cells from normal human donors express GGT at intermediate-high levels (data not shown).

2.3 Association of GGT with CD19/CD21 complex proteins A number of potential mechanisms could account for the inhibition of CD21-induced homotypic adhesion with mAb 3A8. Among the mechanisms examined was whether mAb 3A8 directly inhibited GGT enzyme activity or altered intracellular GSH concentrations; however, we found that the mAb was not capable of either function (data not shown).

GGT is a component of TM4 protein complexes

4125

We then considered whether GGT might be physically associated with the CD19/CD21/TM4 complex and that mAb 3A8 might alter homotypic adhesion by steric hindrance affecting CD21 ligands. To address this, the physical associations between GGT and CD19/CD21/ TM4 complex proteins were studied. In these experiments, Balm-1 cells were lysed with a mild detergent (1 % Brij 97) as no associations were apparent using Triton X-100 (data not shown). Anti-CD81 mAb was able to co-immunoprecipitate significant amounts of GGT activity (Fig. 2 A). mAb recognizing CD53 and CD82, both present in the CD19/CD21 complex, also specifically coimmunoprecipitated GGT activity (Fig. 2 B). The specificity of these TM4 interactions was confirmed by demonstrating that mAb specific for other TM4 proteins (CD9, CD37 and CD63) also expressed by Balm-1 cells and immunoprecipitated by the mAb, did not co-immunoprecipitate GGT activity (data not shown). We considered whether activation of B cells through CD21 might be required to demonstrate association of CD19 or CD21 with GGT as measured by immunopre-

Figure 2. GGT is physically associated with TM4 proteins CD81, CD53 and CD82. Balm-1 cells were lysed in GGT buffer containing 1 % Brij 97 and immunoprecipitated with specific mAb or UPC-10 isotype control as shown. The amount of GGT activity that co-immunoprecipitated with each was determined (A) and is expressed relative to the GGT activity immunoprecipitated by mAb 3A8 itself (approximately 80 % of total enzymatic activity in cell lysates). Shown are means ± SD of a representative experiment. The association of GGT with TM4 proteins CD53, CD81 and CD82 as compared to UPC-10 and CD46 controls was also examined in the same manner (B). These molecular associations were confirmed by immunoprecipitating Brij 97-solubilized Balm-1 cells with mAb 3A8 (GGT) as compared to positive (CD53 and CD81) and negative (UPC-10, CD58) controls, followed by Western blot analysis with either anti-CD81 (C) or anti-CD82 (D).

4126

T. C. Nichols et al.

Eur. J. Immunol. 1998. 28: 4123–4129

cipitation. However, pretreatment of Balm-1 cells with polymerized C3dg prior to lysis did not increase the strength of these associations (data not shown). The molecular associations of GGT with CD82 and CD81 were confirmed by immunoprecipitating Balm-1 cell lysates with mAb 3A8 or controls and Western blotting using either anti-CD81 (Fig. 2 C) or anti-CD82 (Fig. 2 D). As previously reported [4], CD53, CD82 and CD81 coimmunoprecipitate each other. In addition, CD82 as well as CD81 co-immunoprecipitate with GGT. Negative controls included mAb to CD58, which is also expressed at comparable levels. Association of GGT with CD53 could not be analyzed in this manner as mAb MEM-53 does not react by Western blot analysis. Daudi cells, which do not express GGT, were then transfected with the human GGT cDNA and sorted for cells expressing high levels for GGT. In these cells both antiCD81 and anti-CD53 were able to co-immunoprecipitate GGT activity (Fig. 3 A). Therefore, exogenously expressed GGT can interact with TM4 complexes that are normally expressed in the absence of GGT. AntiCD82 could not be tested as Daudi cells do not express this antigen. To determine if the association of GGT with TM4 proteins was specific to B cells, we also studied the CEM T cell line. GGT activity was also specifically coprecipitated with CD81, CD53 and CD82 (Fig. 3 B). Therefore, the association of GGT with TM4 proteins exists in both B and T cell lineages. Although the stoichiometry of the association appears relatively weak by Western blot analysis in Fig. 2 C and D, it is as high as 35 % when the ability to co-immunoprecipitate GGT activity is studied (Fig. 2 A, B). Therefore, a substantial portion of GGT protein is associated with these specific TM4 proteins. This level of detectable association is at least as great as that in many other studies of TM4 associations, in which often X 5 % of proteins co-fractionate using similar techniques [4, 5].

2.4 CD19 and CD21 co-cap with GGT Although the direct association of GGT with CD21 and CD19 could not be readily demonstrated by coimmunoprecipitation techniques, it was still possible that GGT was indirectly associated by their shared interactions with the same TM4 complexes. This issue was addressed by examining the ability of CD19/CD21 complex proteins to co-cap with GGT using confocal microscopy (Fig. 4). These experiments demonstrated specific co-capping of CD19 and CD21 in addition to CD82, CD81 and CD53. Thus, GGT apparently interacts only indirectly with CD19 and CD21 through TM4 proteins.

Figure 3. (A) Co-immunoprecipitation of GGT activity with CD81 and CD53 in Daudi B cells expressing recombinant GGT. The experimental conditions and calculations are as performed in Fig. 2 A and B. No GGT activity was present in untransfected Daudi cells (data not shown). (B) In the CEM T cell line, GGT enzyme activity also specifically coimmunoprecipitates with CD81, CD53 and CD82.

The most straightforward explanation for the inhibition by 3A8 of CD21-induced homotypic adhesion, therefore, is that this particular mAb physically alters the multiprotein complex in a manner that specifically disrupts the CD21 association, resulting in a decrease in ligandinduced homotypic adhesion. Since the signaling pathway activated following CD21 ligation that is required for the induction of homotypic adhesion is unknown, we are currently unable to study this particular issue further.

3 Discussion GGT is an ecto-enzyme whose primary function is to cleave + -glutamyl groups from proteins and peptides and transfer these groups either to other amino acids or hydrolyze them to form glutamate [11]. The most abundant + -glutamyl-containing molecule is GSH, which is present in the cell in millimolar concentrations and plays a central role in setting the intracellular redox potential.

Eur. J. Immunol. 1998. 28: 4123–4129

GGT is a component of TM4 protein complexes

4127

cleaved by dipeptidase, and all three amino acids are transported back into the cell where they can then enter the GSH synthesis pathway. Although this GSH recycling pathway was originally described using biochemical and inhibitor studies [11], substantial support for this mechanism and the role of GGT has been provided by studies of mice in which the gene encoding GGT has been disrupted [12]. These mice demonstrate pronounced glutathionuria and significant (approximately 50 %) decreases in GSH in tissues including liver, kidney [12] and spleen (M. Lieberman, personal communication). We do not yet understand the functional significance of the association between GGT and TM4 proteins. However, based on previous studies there are several possible explanations for such an association. First, the association may be important in regulating local intracellular redox potential. In lymphocytes oxidative stress has been shown to affect signal transduction and the activation of transcription factors such as NF- O B [13]. An independent finding in support of this particular hypothesis is that signal transduction through the TM4 protein CD81 is controlled by the redox state of the cell [14]. Alternatively, GGT may be associated with the TM4 proteins because these proteins might play a role in the binding or recycling of the products of the GGT enzyme reactions, including small peptides, amino acids or other compounds. Others have previously proposed that TM4 proteins may function in nutrient transport roles [5]. For example, transfection of NIH3T3 cells with the cDNA encoding human GGT has been shown to confer the ability of these cells to utilize GSH as a cysteine source [15]. Since GGT is an enzyme involved in the generation of free cysteine, TM4 proteins may be important in the process of transporting cysteine for intracellular use.

Figure 4. GGT co-caps with CD19 and CD21 in addition to TM4 proteins that are part of the CD19/CD21 complex. Balm-1 cells were treated with either anti-CD46 control mAb (left) or anti-GGT mAb 3a8 (right) for 30 min at 37 °C. Texas Red-conjugated anti-mouse IgG secondary reagent was then added and cells were cultured for an additional 1 h at 37 °C. The cells were then washed in cold buffer and stained with FITC-labeled specific mAb as shown at the left of each column. Yellow regions indicate co-localization of red and green dye and demonstrate sites of association.

Although oxidized GSH is rapidly exported from cells, significant amounts of intact GSH cannot directly enter cells. GGT is the first enzyme involved in breaking down extracellular GSH into its two components, glutamate and cysteinyl-glycine. Cysteinyl-glycine is further

The TM4 proteins may also be involved in binding or recycling glutamate, another product of the GGTmediated cleavage of GSH. In this regard, TM4 proteins have been previously noted to have structural similarities with well-characterized gated ion channels such as central nervous system acetylcholine receptors [5], which are also similar to glutamate gated ion channels or glutamate receptors. Finally, GSH has been shown to be conjugated to other compounds such as leukotriene C4 (reviewed in [11]), and certain of these other compounds may also interact with TM4 proteins. In summary, we have demonstrated that GGT is physically associated with TM4 proteins and is also a component of a CD19/CD21-containing TM4 complex on human Bcells. This observation provides a potential link by which a GGTrelated mechanism, such as local regulation of redox potential, may influence the activities of TM4 complex proteins.

4128

T. C. Nichols et al.

Eur. J. Immunol. 1998. 28: 4123–4129

4 Materials and methods

4.5 Western blot analysis

4.1 Antibodies

Electrophoresed proteins were transferred to HyBond ECL nitrocellulose for 1 h at 50 V. The membranes were then blocked with PBS containing 0.25 % Tween-20 and 10 % dry milk, washed and incubated with anti-CD81 (biotinylated mAb 5A6) or anti-CD82 (mAb C33) in PBS with 1 % dry milk an 1 % BSA. These were then washed and incubated with avidin-peroxidase (Pierce) or horseradish peroxidaseconjugated goat anti-mouse IgG (Jackson Labs, West Grove, PA), respectively. After washing, the blots were incubated in ECL substrate (Amersham) and exposed to film.

The anti-GGT mAb 3A8 (IgG2a) was generated by immunizing BALB/c mice with Balm-1 cells that had been treated in vitro at 37° with the anti-CD21 mAb OKB7 (Ortho Diagnostics, Raritan, NY) at 10 ? g/ml for 4 h. The hybridomas resulting from the fusion of splenocytes with the murine myeloma P3X63 Ag8.6.5.3 were screened quantitatively based on the ability of the mAb-containing supernatant to inhibit CD21induced homotypic adhesion. mAb 5A6 (anti-CD81), C33 (anti-CD82) and MEM-53 (anti-CD53) were kindly provided by their creators. Other mAb used included B4 (anti-CD19) (Coulter, Hialeah, FL), HB5 (anti-CD21) (ATCC, Rockville, MD), HD-37 (anti-CD19) (Dako, Carpintera, CA), anti-CD58 (Serotech USA, Washington, D.C.), anti-CD37 (Dako), antiCD9 (Dako), anti-CD63 (AMAC, Inc., Westbrook, ME) and UPC-10 (IgG2a isotype control) (Sigma, St. Louis, MO).

4.2 Cell surface protein biotinylation and SDS-PAGE analysis Balm-1 cells were surface labeled using sulfo-NHS-biotin (Pierce, Rockford, IL), solubilized in 1 % Triton X-100 lysis buffer, immunoprecipated as described below, transferred to nitrocellulose paper probed with avidin-HRP (Pierce, Rockford, IL) and analyzed by enhanced chemiluminescence ECL (Amersham, Arlington Heights, IL).

4.3 Two-dimensional IEF/SDS-PAGE analysis Balm-1 cells were biosynthetically labeled with Tran35Slabel (ICN, Costa Mesa, CA) for 24 h. Membrane preparations were solubilized in Triton X-100 lysis buffer (1 % Triton-100, 50 mM Tris, pH 7.4, 150 mM NaCl, and 1 mM EDTA) and immunoprecipitated overnight with mAb UPC-10 or mAb 3A8 and protein G-Sepharose (Pharmacia, Piscataway, NJ). Following washing, samples were eluted using IEF buffer and run on IEF gels. IEF gels were then saturated with reducing SDS-PAGE sample buffer, layered and electrophoresed on SDS-PAGE (12 % polyacrylamide) gels. The gels were soaked in Fluoro-hance (RPI, Mt Prospect, IL), dried and exposed to film.

4.4 GGT enzyme assays

The GGT enzyme assay was performed as described [16] using Balm-1 cells lysed in 50 mM Tris pH 8.5, 1 mM MgCl2, 1 mM CaCl2 and 150 mM NaCl (GGT buffer) containing as detergents either 1 % Triton X-100, Brij 58, or Brij 97.

4.6 Co-capping and confocal microscopic analysis Balm-1 cells were washed and resuspended at 1 × 107 cells/ ml in PBS with 2 % BSA (capping buffer). mAb 3A8 or antiCD46 (1 ? g) was added to 1 × 106 cells in 100 ? l and incubated for 30 min at 37 °C and then for 1 h with 1 ? g/ml antimouse IgG antibody conjugated to Texas Red (Molecular Probes, Eugene, OR). Cells were then washed and resuspended in ice-cold buffer with 0.02 % azide. The cells were incubated for 20 min at 4 °C with 1 ? g/ml of mAb either biotinylated or conjugated directly to FITC and either washed and fixed with 4 % paraformaldehyde, or washed and incubated one additional time with streptavidin-FITC before washing and fixation. Samples were analyzed using a BioRad MRC-1000 confocal microscope (Bio-Rad Laboratories, Hercules, CA). Data were analyzed using Confocal Assistant version 3.10 (Bio-Rad).

Acknowledgement: The authors thank Carol Brenner Webster for excellent technical assistance and Shad Eubanks and John Freed at the NJC Protein Facility for protein sequence analysis. These studies were supported by a grant from the Rocky Mountain Chapter of the Arthritis Foundation (VMH) and a Biomedical Sciences Grant from the Arthritis Foundation (DRK).

5 References 1 Ahearn, J. M. and Fearon, D. T., Structure and function of the complement receptors, CR1 (CD35) and CR2 (CD21). Adv. Immunol. 1989. 46: 183–219. 2 Molina, H., Holers, V. M., Li, B., Fang, Y.-F., Mariathasan, S., Goellner, F., Strauss-Schoenberger, J., Karr, R. W. and Chaplin, D. D., Markedly impaired humoral immune response in mice deficient in complement receptors 1 and 2. Proc. Natl. Acad. Sci. USA 1996. 93: 3357–3361. 3 Bradbury, L. E., Kansas, G. S., Levy, S., Evans, R. L. and Tedder, T. F., The CD19/CD21 signal transducing complex of human B lymphocytes includes the target of

Eur. J. Immunol. 1998. 28: 4123–4129 antiproliferative antibody-1 and Leu-13 molecules. J. Immunol. 1992. 149: 2841–2850. 4 Angelisova, P. and Horejsi, V., Association of four antigens of the tetraspan family (CD37, CD53, Tapa-1 and R2/C33) with MHC class II glycoproteins. Immunogenetics 1994. 39: 249–256. 5 Wright, M. D. and Tomlinson, M. G., The ins and outs of the transmembrane 4 superfamily. Immunol. Today 1994. 15: 588–594. 6 Maecher, H. T. and Levy, S., Normal lymphocyte development but delayed humoral immune response in CD81-null mice. J. Exp. Med. 1997. 185: 1505–1510. 7 Luxembourg, A. T. and Cooper, N. R., Modulation of signaling via the B cell antigen receptor by CD21, the receptor for C3dg and EBV. J. Immunol. 1994. 153: 4448–4457. 8 Lieberman, M. W., Barrios, R., Habib, G. M., Lebovitz, R. M., Rajagopalan, S., Sepulveda, A. R., Shi, Z. and Wan, D., Gamma-glutamyl transpeptidase: What does the organization and expression of a multipromotor gene tell us about its function? Am. J. Pathol. 1997. 147: 1175–1185. 9 Grisk, O., Kuster, U. and Ansorge, S., The activity of gamma-glutamyl transpeptidase in populations of mononuclear cells from human peripheral blood. Biol. Chem. Hoppe-Seyler 1993. 374: 287–290. 10 Goodspeed, D. C., Dunn, T. J., Miller, C. D. and Pitot, H. C., Human gamma-glutamyl transpeptidase cDNA: Comparison of hepatoma and kidney mRNA in the human and rat. Gene 1987. 76: 1–9. 11 Meister, A., Tate, S. S. and Griffith, O. W., Gammaglutamyl transpeptidase. Methods Enzymol. 1997. 77: 237–253.

GGT is a component of TM4 protein complexes

4129

12 Lieberman, M. W., Wiseman, A. L., Shi, Z.-Z., Carter, B. Z., Barrios, R., Ou, C.-N., Chevez-Barrios, P., Wang, Y., Habib, G. M., Goodman, J. C., Huang, S. L., Lebovitz, R. M. and Matzuk, M. M., Growth retardation and cysteine deficiency in + -glutamyl transpeptidasedeficient mice. Proc. Natl. Acad. Sci. USA 1996. 93: 7923–7926. 13 Schreck, R., Rieber, P. and Baeuerle, P. A., Reactive oxygen intermediates as apparently widely used messengers in the activation of the NF- O B transcription factor and HIV-1. EMBO J. 1991. 10: 2247–2258. 14 Schick, M. R., Nguyen, V. Q. and Levy, S., Anti-TAPA-1 antibodies induce protein tyrosine phosphorylation that is prevented by increasing intracellular thiol levels. J. Immunol. 1993. 151: 1918–1925. 15 Hanigan, M. H. and Ricketts, W. A., Extracellular glutathione is a source of cysteine for cells that express gamma-glutamyl transpeptidase. Biochemistry 1993. 32: 6302–6306. 16 Sepulveda, F. V., Burton, K. A. and Pearson, J. D., The development of gamma-glutamyl transferase in a pig renal-epithelial cell line in vitro. Biochem. J. 1982. 208: 509–512.

Correspondence: V. Michael Holers, Box B-115, University of Colorado Health Sciences Center, 4200 East Ninth St., Denver, CO 80262, USA Fax: +1-303 315 5540 e-mail: Michael.Holers — uchsc.edu