Retinoic acid receptor-beta: immunodetection and phosphorylation on tyrosine residues.

Polyclonal (RP) and monoclonal (Ab) antibodies were raised against synthetic peptides (or fusion proteins) corresponding to amino acid sequences unique to human and mouse retinoic acid receptor-beta (RAR beta) isoforms. Antibodies directed against the A2 region [Ab6 beta 2(A2), Ab7 beta 2(A2), and RP beta 2(A2)], the D2 region [RP beta(D2)], or the F region [Ab8 beta(F)2, RP beta(F)1, and RP beta(F)2] were selected. The monoclonal and polyclonal antibodies directed against the D2 and F regions specifically immunoprecipitated and recognized by Western blotting all human and mouse RAR beta isoforms (mRAR beta 1, -beta 2, -beta 3, and -beta 4), produced in COS-1 cells transfected with expression vectors containing the corresponding RAR beta cDNA. Furthermore, in gel retardation assays, the monoclonal antibodies supershifted RAR beta protein-RA response element oligonucleotide complexes. Antibodies directed against the A2 region were specific for the RAR beta 2 isoform. The above antibodies allowed us to detect the presence of mRAR beta 2 proteins in mouse embryos and to show that their presence in embryonal carcinoma cells (F9 and P19 cell lines) is dependent upon RA treatment. The antibodies were also used to demonstrate that RAR beta proteins produced by transfection in COS-1 cells are phosphorylated. RAR beta 2 phosphorylation was not affected by RA treatment, whereas the phosphorylation of RAR beta 1 and RAR beta 3 isoforms was greatly enhanced by RA. We also show that, in contrast to RAR alpha 1 and RAR gamma 1, RAR beta 2 proteins contain phosphotyrosine residues and are only weakly phosphorylated in vitro by cAMP-dependent protein kinase. These results support our previous proposal that the various receptors have distinct functions in the RA-signaling pathway.     

Retinoic acid receptor gamma: specific immunodetection and phosphorylation

Synthetic peptides corresponding to cDNA-deduced amino acid sequences unique to the human and mouse retinoic acid receptor gamma 1 (hRAR-gamma 1 and mRAR-gamma 1, respectively) were used to generate anti-RAR-gamma 1 antibodies. Four mAbs were selected, which were directed against peptides found in region A1 (Ab1 gamma (A1)), region F (Ab2 gamma (mF) and Ab4 gamma (hF)) and region D2 (Ab5 gamma (D2)). These antibodies specifically immunoprecipitated and recognized by Western blotting RAR-gamma 1 proteins in COS-1 cells transfected with expression vectors containing the RAR-gamma 1 cDNAs. They all reacted with both human and mouse RAR-gamma 1 proteins, except Ab4 gamma (hF) that was specific for hRAR-gamma 1. Rabbit polyclonal antibodies, directed against a peptide from the mRAR-gamma 1 F region were also obtained (RP gamma (mF)) and found to be specific for mouse RAR-gamma 1 protein. Furthermore, in gel retardation/shift assays the antibodies specifically retarded the migration of complexes obtained with a RA response element (RARE). Antibodies raised against regions D2 and F also recognized the RAR-gamma 2 isoform which differs from RAR-gamma 1 only in the A region. On the other hand, antibodies directed against the A1 region of RAR-gamma 1 (Ab1 gamma (A1)) only reacted with the RAR-gamma 1 protein. The antibodies characterized here allowed us to detect the presence of mRAR-gamma 1 and gamma 2 isoforms in mouse embryos and F9 embryonal carcinoma cells nuclear extracts. They were also used to demonstrate that the mRAR-gamma 1 protein can be phosphorylated and that the phosphorylation occurs mainly in the NH2-terminal A/B region.     

The use of monoclonal antibodies to study the proteins specified by the transforming region of human adenoviruses.

Monoclonal antibodies against two of the proteins specified by one of the transforming genes (early region 1B) of human adenovirus type 2 have been produced and characterized. Two clones (RA1 and PA6), generated by fusion of mouse myeloma NSO cells with splenocytes from rats immunized with whole-cell lysates of an adenovirus-transformed rat cell line (F19), secreted antibodies against a 58 kDa protein. Another clone (DC1) produced antibodies against the same protein, and resulted from fusion of immune rat splenocytes with the rat myeloma Y3.Ag.1.2.3. Immunoprecipitation studies showed that all three antibodies recognized [35S]-methionine-labelled 58 kDa protein, and phosphorylated derivatives of the 58 kDa protein labelled with [32P]orthophosphate present in infected human cells. One clone (EC3) produced antibody against a 19 kDa protein also encoded by early region 1B, but not sharing sequence homology with 58 kDa. The identity of the 19 kDa protein recognized by the EC3 antibody was established by immunoprecipitation from lysates of labelled-infected cells and from products of cell-free translation directed by mRNA isolated from adenovirus 2-infected cells. Indirect immunofluorescent-antibody staining of infected human cells using the RA1 and EC3 antibodies revealed a nuclear location of the 58 kDa protein and a mainly cytoplasmic location of the 19 kDa protein.     

Functional analysis of type 1alpha cGMP-dependent protein kinase using green fluorescent fusion proteins

The cGMP-dependent protein kinases (PKGs) are ubiquitous effector enzymes that regulate a variety of physiological processes in response to nitric oxide and natriuretic agonists. We have constructed green fluorescent fusion proteins (GFP) using full-length (PKG-GFP) and truncations encoding either the regulatory domain of PKG1alpha (G1alphaR-GFP) or the catalytic domains of PKG1alpha (GFP-G1C) to examine the enzymatic properties and intracellular location. When transiently transfected into mammalian cells, these constructs were detected on Western blots at the expected sizes using anti-GFP antibodies. The GFP-G1C and the full-length PKG1alpha-GFP fusion proteins were found to have constitutive activity both in vivo and in vitro. The G1alphaR-GFP protein was found to dimerize with endogenous type 1 PKG and behaved in a dominant negative manner both in vivo and in vitro. When expressed transiently in either HEK-293 or A549 epithelial cells, the fusion proteins encoding the amino-terminal regulatory domains (PKG-GFP, G1alphaR-GFP) were present in the cytosol and were rarely observed in the nucleus. In contrast, the GFP-G1C (lacking regulatory domains) concentrated in the nucleus. Of the fusion proteins containing the regulatory region, the constitutive PKG-GFP protein was present in a more centralized location, whereas the G1alphaR-GFP protein colocalized with F-actin on stress fibers and in dynamic regions of the plasma membrane. Microscopic and immunoprecipitation studies indicated that both the G1alphaR-GFP and the PKG-GFP fusion proteins colocalized with vasodilator-stimulated phosphoprotein (VASP). These constructs thus represent novel tools with which to visualize inactive, and activated, PKG1alpha in vivo, and we have used them to demonstrate two functionally independent domains. In addition, we show for the first time in living cells that PKG is found in dynamic membrane regions in association with VASP.     

Immunoaffinity purification of SRT-tagged human creatine kinase by peptide elution

The mouse monoclonal antibody (Mab), SRT10, recognizes a linear epitope of 10 amino acids (ThrPheIleGlyAlaIleAlaThrAspThr). When these epitope-tagged fusion proteins are expressed in mammalian cells, the Mab can detect the tagged proteins by immunoblotting, immunocytochemistry and immunoprecipitation. Here, we describe an efficient method for the purification of SRT-tagged recombinant human creatine kinase (CK) transiently expressed in mammalian cells. This method utilizes the expression of the N-terminal- or C-terminal-tagged CK in transiently transfected HEK293 cells followed by binding to anti-SRT-agarose affinity resin and competitive elution with SRT peptide. Recombinant CK was purified near homogeneity as judged by SDS-PAGE.     

Functional redundancy of the DE-1 and alpha A-CRYBP1 regulatory sites of the mouse alpha A-crystallin promoter.

Previous studies have implicated the DE-1 (-111/-106) and alpha A-CRYBP1 (-66/-57) sites for activity of the mouse alpha A-crystallin promoter in transiently transfected lens cells. Here we have used the bacterial chloramphenicol acetyltransferase (CAT) reporter gene to test the functional importance of the putative DE-1 and alpha A-CRYBP1 regulatory elements by site-specific and deletion mutagenesis in stably transformed alpha TN4-1 lens cells and in transgenic mice. FVB/N and C57BL/6 x SJL F2 hybrid transgenic mice were assayed for CAT activity in the lens, heart, lung, kidney, spleen, liver, cerebrum, and muscle. F0, F1, and F2 mice from multiple lines carrying single mutations of the DE-1 or alpha A-CRYBP1 sites showed high levels of CAT activity in the lens, but not in any of the non-lens tissues. By contrast, despite activity of the wild-type promoter, none of the mutant promoter/CAT constructs were active in the transiently transfected and stably transformed lens cells. The mice carrying transgenes with either site-specific mutations in both the DE-1 and alpha A-CRYBP1 sites or a deletion of the entire DE-1 and part of the alpha A-CRYBP1 site (-60/+46) fused to the CAT gene did not exhibit CAT activity above background in any of the tissues examined, including the lens. Our results thus indicate that the DE-1 and alpha A-CRYBP1 sites are functionally redundant in transgenic mice. Moreover, the present data coupled with previous transfection and transgenic mouse experiments suggest that this functional redundancy is confined to lens expression within the mouse and is not evident in transiently transfected and stably transformed lens cells, making the cultured lens cells sensitive indicators of functional elements of crystallin genes.     

Structure-function analysis of the Abm12 beta mutation using site-directed mutagenesis and DNA-mediated gene transfer

The B6.C-H-2bm12 (bm12) mouse possesses a naturally occurring mutation in its class II MHC A beta gene. The three amino acid substitutions at positions 67, 70, and 71 that comprise this mutation lead to changes in both Ia expression and immune recognition of the resultant A beta A alpha molecule. The experiments reported here utilize a combination of oligonucleotide-mediated site-directed mutagenesis and DNA-mediated gene transfer to explore the roles played by each of the three mutant residues in these various phenotypic changes. A beta genes comprising all permutations of the residues distinguishing Ab beta from Abm12 beta were created and were individually co-transfected with Ab beta into mouse L cells. Sublines expressing high levels of membrane Ia were selected by preparative flow cytometry and were studied for reactivity with a panel of monoclonal anti-Ia antibodies, or for their ability to act as antigen-presenting cells (APC) for the stimulation of T cell hybridomas. During the generation of these transfectant lines, it was noted that expression of a high level of Abm12 beta Ab alpha was more difficult to achieve than a similar level of Ab beta Ab alpha. Northern blot analysis of specific A beta and A alpha mRNA levels in these various lines indicated that more class II mRNA, and presumably more A beta and A alpha chains, were required to achieve expression of Abm12 beta Ab alpha equal to that of Ab beta Ab alpha, suggesting that the previously noted reduction of Ia expression on cells from bm12 mice reflects a decreased ability of Abm12 beta Ab alpha chains to pair, or to reach the membrane. Staining of the panel of transfectants with monoclonal antibodies revealed that antibodies which did not distinguish Ab beta Ab alpha from Abm12 beta Ab alpha also reacted equally well with all molecules involving in vitro mutant A beta chains. Monoclonal antibodies reactive with Ab beta Ab alpha but not Abm12 beta Ab alpha were specific for an epitope primarily determined by the presence or absence of Arg 70 in Ab beta. In striking contrast, all three mutant positions were found to play crucial roles in T cell recognition, because all substitutions led to significant or complete loss of antigen-presenting function with all but one of the T hybridomas tested.(ABSTRACT TRUNCATED AT 400 WORDS)     

Antibodies to the estrogen receptor-alpha modulate rapid prolactin release from rat pituitary tumor cells through plasma membrane estrogen receptors.

Antibodies (Abs) raised against the estrogen receptor-alpha (ERalpha) were used to investigate the role of ERalpha proteins located at the plasma membrane in mediating the rapid, estrogen-stimulated secretion of prolactin (PRL) from rat pituitary GH(3)/B6/F10 cells. Exposure of the cells to 1 nM 17beta-estradiol (E(2)) significantly increased PRL release after 3 or 6 min. When ERalpha Abs that bind specifically to ERalpha but are too large to diffuse into cells were tested for activity at the cell membrane, Ab R4, targeted to an ERalpha hinge region sequence, increased PRL release in a time- and concentration-dependent fashion. Ab H151, directed against a different hinge region epitope, decreased PRL release and blocked the stimulatory action of E(2). Abs raised against the DNA binding domain (H226) or the carboxyl terminus (C542) were not biologically active. When each Ab was examined for recognition of ERalpha on the cell surface by immunocytochemistry, all except H151 generated immunostaining in aldehyde-fixed cells. In live cells, however, Ab H151 but not Ab R4 blocked the membrane binding of fluorescently tagged E(2)-BSA. Overall, the data indicate that plasma membrane ERalpha proteins mediate estrogen-stimulated PRL release from GH(3)/B6/F10 cells. These results may also convey information about conformationally sensitive areas of the membrane form of ERalpha involved in rapid, nongenomic responses to estrogens.     

Inhibition of in vitro tumor cell growth by RP215 monoclonal antibody and antibodies raised against its anti-idiotype antibodies

RP215 monoclonal antibody (Mab) was shown to recognize carbohydrate-associated epitope(s) in the heavy chains of cancer cell-expressed immunoglobulins, designated in general as CA215 pan cancer marker. Growth inhibitions of tumor cells in vitro by RP215 Mab and antibodies against its anti-idiotype (anti-id) antibodies were investigated. Polyclonal rabbit anti-id antibodies and the corresponding rat anti-id Mabs were generated and characterized. Following immunizations in mice, antisera raised against anti-id antibodies were analyzed by typical immunoassays. It was observed that mouse anti-anti-id sera (Ab3) revealed binding affinity and specificity to CA215 that are comparable to those of RP215. Both RP215 and Ab3 were shown to induce apoptosis of cultured cancer cells in vitro by TUNEL and MTT assays. These experimental observations were consistent with that of in vivo tumor growth inhibition by RP215 in previous nude mouse experiments. Therefore, heterologous or homologous anti-id antibodies of RP215 that contain the internal image of its specific epitope in CA215 may serve as effective anti-cancer vaccines for therapeutic treatments of various cancers in humans. The relative stability of RP215-specific carbohydrate-associated epitope was compared to that of human IgG at extreme pH’s (≤2 or ≥12) or following NaIO₄ treatments. The major molecular forms of CA215 were further documented with various enzyme immunoassays and found to have similar secondary structures to those of normal human immunoglobulin G.     

An IL-17F/A heterodimer protein is produced by mouse Th17 cells and induces airway neutrophil recruitment

IL-17A and IL-17F are related homodimeric proteins of the IL-17 family produced by Th17 cells. In this study, we show that mouse Th17 cells also produce an IL-17F/A heterodimeric protein. Whereas naive CD4(+) T cells differentiating toward the Th17 cell lineage expressed IL-17F/A in higher amounts than IL-17A/A homodimer and in lower amounts than IL-17F/F homodimer, differentiated Th17 cells expressed IL-17F/A in higher amounts than either homodimer. In vitro, IL-17F/A was more potent than IL-17F/F and less potent than IL-17A/A in regulating CXCL1 expression. Neutralization of IL-17F/A with an IL-17A-specific Ab, and not with an IL-17F-specific Ab, reduced the majority of IL-17F/A-induced CXCL1 expression. To study these cytokines in vivo, we established a Th17 cell adoptive transfer model characterized by increased neutrophilia in the airways. An IL-17A-specific Ab completely prevented Th17 cell-induced neutrophilia and CXCL5 expression, whereas Abs specific for IL-17F or IL-22, a cytokine also produced by Th17 cells, had no effects. Direct administration of mouse IL-17A/A or IL-17F/A, and not IL-17F/F or IL-22, into the airways significantly increased neutrophil and chemokine expression. Taken together, our data elucidate the regulation of IL-17F/A heterodimer expression by Th17 cells and demonstrate an in vivo function for this cytokine in airway neutrophilia.     

Tumor-specific idiotype vaccines. II. Analysis of the tumor-related network response induced by the tumor and by internal image antigens (Ab2 beta)

In this study the tumor-specific immuneresponse induced by irradiated tumor cells (L1210/GZL) and by anti-idiotype antibodies was analyzed. The anti-idiotype antibodies (Ab2) were made against the paratope of a monoclonal antitumor antibody (11C1) that recognizes a tumor-associated antigen which cross-reacts with the mouse mammary tumor virus-encoded envelope glycoprotein 52. Two Ab2, 2F10 and 3A4, induced idiotypes expressed by the monoclonal antitumor antibodies 11C1 and 2B2. Cytotoxic T cells, generated by immunization with irradiated tumor cells, lyse 2F10 and 3A4 hybridoma cells. Furthermore, immunization with Ab2 induces tumor-specific cytotoxic T lymphocytes. The frequency of tumor-reactive cytotoxic T lymphocyte was found to be similar in mice immunized with Ab2 or irradiated tumor cells when examined at the precursor level. However, only 2F10 induces protective immunity against the growth of L1210/GZL tumor cells. The depletion of a L3T4+ T cell population from 2F10 immune mice was found to increase the effectiveness of transferred T cells to induce inhibition of tumor growth. The inability of 3A4 to induce antitumor immunity could be correlated with the presence of a population of Lyt2+ regulatory T cells. Collectively, these results demonstrate the existence of a regulatory network controlling the expression of effective tumor immunity. Our results demonstrate that selection of binding site-related Ab2 may not be a sufficient criteria for the development of an idiotype vaccine. A better understanding of the regulatory interactions induced by anti-idiotypes is needed for the design of effective antitumor immunotherapy.     

Identification of a novel retinoic acid-responsive element within the lamin A/C promoter

A-type lamins are not present in either early embryos or the embryonal carcinoma (EC) cell line. P19 cells, which are EC cell line, are able to express A-type lamins upon retinoic acid (RA) treatment. Here we report that a novel RA-responsive element, termed lamin A/C-RA-responsive element (L-RARE), is located within the lamin A/C promoter. RA activated the luciferase activity of the reporter which had four tandem repeats of the wild-type L-RARE, while a loss of function mutant, which altered CACCCCC to CACtatC within L-RARE, did not respond. Four specific binding complexes of L-RARE, Complexes-A, -B, -C, and -D, were detected in protein extracts obtained from P19 cells treated with and without RA. Specific antibodies revealed that Sp1 and Sp3 were included in Complex-A and Complexes-B and -C, respectively. Thus, L-RARE was important in the RA-mediated activation of the lamin A/C promoter and was recognized by DNA binding proteins.     

Tissue distribution and immunoreactivity of heme-regulated eIF-2 alpha kinase determined by monoclonal antibodies.

A highly purified preparation of heme-regulated inhibitor (HRI), an eIF-2 alpha kinase, from rabbit reticulocyte lysates has been used for generating monoclonal antibodies (mAB). Two hybridoma clones secreting HRI-specific antibodies (mAB A and mAB F) were obtained. Both antibodies immunoprecipitated biosynthetically labeled as well as phosphorylated HRI in reticulocyte lysates and also recognized denatured HRI in a Western blot. In in vitro protein kinase assays, preincubation of HRI with the antibodies significantly diminished both autokinase and eIF-2 alpha kinase activities. HRI from reticulocyte lysates could be quantitatively removed by immunoprecipitation with mAB F, and such HRI-depleted lysates were able to maintain protein synthesis under conditions of heme deficiency. With these monoclonal antibodies, HRI was detected only in the reticulocytes and bone marrow of anemic rabbits, among several rabbit tissues tested. The antibodies did not detect cross-reacting HRI in rat or human reticulocytes or in mouse erythroleukemic cells or human K562 cells even after induction of differentiation, although eIF-2 alpha kinase activity was detected in them. Polyclonal anti-rabbit HRI antibody detected HRI in rat reticulocytes. However, no cross-reacting HRI was detected by polyclonal antibody in human reticulocytes or other cell types tested. These findings suggest that HRI is not ubiquitous, and may be erythroid-specific, and that it is antigenically different in different species.     

A protein whose binding to Na,K-ATPase is regulated by ouabain

Immunoprecipitation of Na,K-ATPase from kidney homogenate by antibodies against alpha1-subunit results in the precipitation of several proteins together with the Na,K-ATPase. A protein with molecular mass of about 67 kD interacting with antibodies against melittin (melittin-like protein, MLP) was found in the precipitate when immunoprecipitation was done in the presence of ouabain. If immunoprecipitation was done using antibodies against melittin, MLP and Na,K-ATPase alpha1-subunit were detected in the precipitate, and the amount of alpha1-subunit in the precipitate was increased after the addition of ouabain to the immunoprecipitation medium. MLP was purified from mouse kidney homogenate using immunoaffinity chromatography with antibodies against melittin. The addition of MLP to purified FITC-labeled Na,K-ATPase decreases fluorescence in medium with K+ and increases it in medium with Na+. The enhancement of fluorescence depends upon the MLP concentration. The N-terminal sequence of MLP determined by the Edman method is the following: HPPKRVRSRLNG. No proteins with such N-terminal sequence were found in the protein sequence databases. However, we revealed five amino acid sequences that contain this peptide in the middle part of the chain at distance 553 amino acids from the C-terminus (that corresponds to protein with molecular mass of about 67 kD). Analysis of amino acid sequence located between C-terminus and HPPKRVRSRLNG in all found sequences has shown that they were highly conservative and include WD40 repeats. It is suggested that the 67-kD MLP either belongs to the found protein family or was a product of proteolysis of one of them.     

Role of exogenous hemin in the synthesis of hemoproteins and nonheme proteins during glucose repression in Saccharomyces cerevisiae

Exogenous addition of hemin to glucose-repressed cells of Saccharomyces cerevisiae stimulates the incorporation of amino acid into cytoplasmic proteins twofold. There was no significant change in the synthesis of total cytoplasmic RNA whereas a 40% increase in the synthesis of poly(A)-containing RNA was observed upon hemin treatment. Cell-free translation of cytoplasmic mRNAs and immunoprecipitation analysis of the translated products with antibodies against subunit V of cytochrome oxidase and the alpha and beta subunits of F1-ATPase reveals that there is an eightfold enrichment of the mRNA for subunit V of cytochrome oxidase upon hemin treatment. The effect of hemin on the alpha and beta subunits of F1-ATPase is only marginal, suggesting a differential role for heme in the synthesis of hemoproteins and nonheme proteins during glucose repression.     

Developmental expression of G protein alpha subunits in mouse spermatogenic cells: evidence that G alpha i is associated with the developing acrosome.

Guanine nucleotide-binding proteins (G proteins) are important signal transducing molecules found in all cells. G proteins are associated with the plasma membrane/outer acrosomal membrane region of acrosome-intact sperm and at least one G protein is involved in the zona pellucida-induced acrosome reaction. With the goal of elucidating the functions of these proteins during spermatogenesis, we investigated the types of G proteins present in spermatogenic cells and when they first become associated with the developing acrosome. Using bacterial toxin-catalyzed [32P]ADP-ribosylation in conjunction with immunoprecipitation and immunofluorescence utilizing antibodies directed against specific regions of various G protein isotypes, the alpha subunits of Gi1, Gi2, Gi3, and G(o) were detected in mouse spermatocytes and spermatids. An antiserum recognizing a conserved sequence of G alpha i subtypes localized to the proacrosomal granules of spermatocytes and the developing acrosome of spermatids. Levels of G alpha o diminished as spermatocytes developed into spermatids such that G alpha o was not detected in cauda epididymal sperm. Immunoreactivity using G alpha o-specific antisera did not display a distinct regionalization within any of the spermatogenic cell types. G alpha s was not detected in the developing spermatogenic cells or sperm. The association of G alpha i with the developing acrosome suggests a role for G proteins may have a role in acrosome biogenesis as well as being part of a complex required later for signal transduction leading to acrosomal exocytosis.     

The processing alpha1,2-mannosidase of Saccharomyces cerevisiae depends on Rer1p for its localization in the endoplasmic reticulum.

The yeast alpha1,2-mannosidase Mns1p is involved in N-linked oligosaccharide processing in Saccharomyces cerevisiae by converting Man9GlcNAc2 to a single isomer of Man8GlcNAc2. alpha1,2-Mannosidase is a 63 kDa type II resident membrane protein of the endoplasmic reticulum that has none of the known endoplasmic reticulum localization signals (HDEL/KDEL, KKXX, or RRXX). Using antibodies against recombinant alpha1,2-mannosidase, indirect immunofluorescence showed that alpha1,2-mannosidase localization is abnormal in rer1 cells and that the alpha1,2-mannosidase localizes in the vacuoles of rer1/deltapep4 cells whereas in wild-type and deltapep4 cells it is found in the endoplasmic reticulum. 35S-labeled cell extracts were subjected to double immunoprecipitation, first with antibodies to alpha1,2-mannosidase, then with either alpha1,2-mannosidase antibodies or antibodies to alpha1,6-mannose residues added in the Golgi. The labeled proteins were examined by autoradiography after sodium dodecyl sulfate polyacrylamide gel electrophoresis. A significant proportion of the labeled alpha1,2-mannosidase was immunoprecipitated by alpha1,6-mannose antibodies in wild-type, deltapep4 and rer1/deltapep4 cells with endogenous levels of alpha1,2-mannosidase, and in wild-type, deltapep4, rer1 and rer1/deltapep4 cells overexpressing alpha1,2-mannosidase. The alpha1,2-mannosidase of rer1/deltapep4 cells had a slower mobility on the gels than alpha1,2-mannosidase precipitated from wild-type or deltapep4 cells, indicating increased glycosylation due to transport through the Golgi to the vacuoles. It is concluded that the endoplasmic reticulum localization of alpha1,2-mannosidase in wild-type cells depends on Rer1p for retrieval from an early Golgi compartment.