One-step immunopurification and lectinochemical characterization of the Duffy atypical chemokine receptor from human erythrocytes

Duffy antigen/receptor for chemokines (DARC) is a glycosylated seven-transmembrane protein acting as a blood group antigen, a chemokine binding protein and a receptor for Plasmodium vivax malaria parasite. It is present on erythrocytes and endothelial cells of postcapillary venules. The N-terminal extracellular domain of the Duffy glycoprotein carries Fy(a)/Fy(b) blood group antigens and Fy6 linear epitope recognized by monoclonal antibodies. Previously, we have shown that recombinant Duffy protein expressed in K562 cells has three N-linked oligosaccharide chains, which are mainly of complex-type. Here we report a one-step purification method of Duffy protein from human erythrocytes. DARC was extracted from erythrocyte membranes in the presence of 1% n-dodecyl-β-D-maltoside (DDM) and 0.05% cholesteryl hemisuccinate (CHS) and purified by affinity chromatography using immobilized anti-Fy6 2C3 mouse monoclonal antibody. Duffy glycoprotein was eluted from the column with synthetic DFEDVWN peptide containing epitope for 2C3 monoclonal antibody. In this single-step immunoaffinity purification method we obtained highly purified DARC, which migrates in SDS-polyacrylamide gel as a major diffuse band corresponding to a molecular mass of 40-47?kDa. In ELISA purified Duffy glycoprotein binds anti-Duffy antibodies recognizing epitopes located on distinct regions of the molecule. Results of circular dichroism measurement indicate that purified DARC has a high content of α-helical secondary structure typical for chemokine receptors. Analysis of DARC glycans performed by means of lectin blotting and glycosidase digestion suggests that native Duffy N-glycans are mostly triantennary complex-type, terminated with α2-3- and α2-6-linked sialic acid residues with bisecting GlcNAc and α1-6-linked fucose at the core.     

Expression and binding properties of a soluble chimeric protein containing the N-terminal domain of the Duffy antigen

The blood group Duffy antigen of human erythrocytes, which exists in two allelic forms, Fy(a) and Fy(b), is a promiscuous chemokine receptor. In this report we describe the expression and purification of a chimeric protein composed of the amino-terminal extracellular domain of the Duffy antigen (aa 3-60), C-terminal intracellular fragment of glycophorin A (GPA, aa 104-131), and the hexahistydyl tag. We obtained two forms of the recombinant protein containing the Fy(a) or Fy(b) epitope, denoted Fy(a)/GPA and Fy(b)/GPA, respectively. These constructs were expressed in Escherichia coli as periplasmic proteins and were purified by affinity chromatography on the Ni-NTA-agarose. Both proteins bound the monoclonal antibodies recognizing the common Fy6 epitope of the Duffy antigen and an epitope of the C-terminal fragment of GPA, and only the Fy(a)/GPA bound anti-Fy(a) antibody. However, binding of IL-8 to the recombinant proteins was not detected, which indicated that an N-terminal domain of the Duffy antigen is not sufficient for an effective chemokine binding. The lack of the chemokine binding was not likely to be due to the lack of glycosylation of the Fy/GPA, since IL-8 was effectively bound to de-N-glycosylated erythrocytes.     

The role of the Duffy antigen-related chemokine receptor in psoriasis vulgaris

Chemokines represent a family of potent biological mediators. Within the group of receptors mediating their effects, a promiscuous receptor has been found which is able to bind and inactivate diverse chemokines of both C-C and C-X-C families. It is co-localized with blood group antigens of the Duffy system on the same glycoprotein and expressed on red blood cells as well as post-capillary blood vessels. In the present study three aspects of Duffy pathophysiology were studied: firstly the amount of IL-8 and RANTES binding to red blood cells and its correlation to disease activity of psoriatic patients, secondly the distribution of Duffy phenotype among psoriatic patients and thirdly the expression of Duffy antigen in normal vs psoriatic skin. Red blood cells from psoriatic patients (n=50) were lysed by triton X (1%) and supernatants tested in IL-8- and RANTES sandwich-ELISA. Duffy phenotype of psoriatic patients (n=50) was assessed by typing red blood cells with specific antisera in indirect Coombs technique. For immunohistochemical detection in normal and psoriatic skin (n=10 respectively) a specific monoclonal antibody (Fy6) was used. Neither IL-8- nor RANTES-levels on red blood cells correlated to disease activity and distribution of Duffy phenotype in psoriatics was not significantly altered when compared to the normal population. Furthermore, Duffy antigen was expressed in a similar pattern in normal and psoriatic skin at all parts of vasculature, albeit much more abundantly in diseased skin. Altogether, chemokine binding to red blood cells seems of minor importance in psoriasis. However, Duffy antigen together with other binding mechanisms like proteoglycans may play a role at local level by binding locally produced chemokines. Thus biological effects of chemokines are both restricted and focussed to dermal tissue.     

Plasmodium vivax interaction with the human Duffy blood group glycoprotein: identification of a parasite receptor-like protein

The interaction between merozoites of the human pathogen, Plasmodium vivax, and the Duffy blood group glycoprotein on the surface of human erythrocytes is essential for the invasion of erythrocytes and the survival of the parasite. We have identified a P. vivax protein of 135 to 140 kDa which binds with receptor-like specificity to the human Duffy blood group glycoprotein. This interaction can be specifically inhibited by purified Duffy glycoprotein and by pretreating erythrocytes with a monoclonal antibody directed against a novel Duffy determinant. A protein with similar specificity for the Duffy glycoprotein from the phylogenetically related simian malaria, P. knowlesi, is shown to be immunologically related by the generation of cross-reactive antibodies. Despite their shared properties, these two Duffy associating proteins from P. vivax and P. knowlesi differ in some aspects of their interaction with the Duffy glycoprotein. The identification of these proteins will help elucidate the molecular mechanisms of erythrocyte invasion by Plasmodium.     

Heat-shock cognate protein (hsc71) and related proteins in mouse spermatogenic cells

A monoclonal antibody (13D3) has been developed that recognizes a 71 kilodalton (71 kDa) protein on two-dimensional immunoblots of proteins extracted from a mixture of mouse spermatogenic cells (mainly pachytene spermatocytes and spermatids). This protein was shown by immunoblotting and adenosine triphosphate (ATP)-binding characteristics to be identical to a 71 kDa mouse heat-shock cognate (hsc) protein, hsc71, present in 3T3 cells. Along with a 70 kDa heat-shock inducible protein (hsp70), and a 74 kDa heat-shock cognate protein (hsc74), hsc71 is a product of the mouse HSP70 multigene family. Although antibody 13D3 reacted strongly with hsc71, it reacted only faintly with hsp70 in 3T3 cells, and not at all with hsc74 or a germ cell-specific hsp70-like protein (P70) on immunoblots of mixed germ cells. Antibody 13D3 is unique among known antibodies in its pattern of reaction with these heat-shock proteins. In immunofluorescence studies on isolated germ cells, 13D3 reacted uniformly with the cytoplasm of pachytene spermatocytes, round spermatids, and residual bodies, but only with the midpiece of spermatozoa. Antibody 13D3 recognizes other proteins in addition to hsc71 on two-dimensional immunoblots of condensing spermatids and spermatozoa. Two of the proteins (70 kDa/pI 6.4 and 70 kDa/pI 6.5) were present in condensing spermatids and spermatozoa, and another protein (69 kDa/pI 7.0) was detected only in spermatozoa. The new proteins also were recognized by monoclonal antibody 7.10, which reacts specifically with hsp70, hsc71, hsc74, and P70. Although [35S]methionine was incorporated into the new proteins in condensing spermatids, hsc71, hsc74, and P70 were not labeled. These results suggest that unique heat-shock proteins are synthesized late in spermatogenesis.     

Detection and isolation of minisatellite Pc-1 binding proteins.

Minisatellites (MNs) are arrays of 5-100 nucleotide repeats that are dispersed throughout the genome of vertebrates. They demonstrate alteration in tumors and in cells exposed to various carcinogens, but the molecular mechanisms underlying the induction of mutations at MNs are largely unknown. Hypervariable MN Pc-1 isolated from the mouse genome consists of tandem repeats of d(GGCAG) flanked with locus-specific sequences at both ends. We have found that MN mutations are induced in NIH3T3 cells by treatment with okadaic acid using a Pc-1 MN fragment as a probe. In order to shed light on the molecular mechanisms, we isolated six MN Pc-1 binding proteins, pA, pB, pD, pE, pF and pG, from nuclear extracts of NIH3T3 cells treated with okadaic acid. While pA and pB bound to the G-rich strand of Pc-1, pD, pE, pF and pG bound to the complementary C-rich strand. Sequence specificities for DNA binding were revealed and one base substitution and insertion into the Pc-1 repeat unit dramatically changed the affinity of each protein, suggesting that they bind to Pc-1 and Pc-1-like MNs in vivo.     

Detection of a Plasmodium vivax erythrocyte binding protein by flow cytometry.

BACKGROUND: The malaria parasite Plasmodium vivax preferentially invades reticulocytes. It is therefore relevant for vaccine development purposes to identify and characterize P. vivax proteins that bind specifically to the surface of reticulocytes. We have developed a two-color flow cytometric erythrocyte binding assay (F-EBA) that has several advantages over traditional erythrocyte binding assays (T-EBAs) used in malaria research. We demonstrate the use of F-EBA using the P. vivax Duffy binding protein region II (PvDBP-RII) recombinant protein as a model. This protein binds to all erythrocytes that express the Duffy receptor (Fy) and discriminates binding between normocytes and reticulocytes. METHODS: F-EBAs were performed by incubating freshly isolated Aotus nancymai, Macaca mulatta, Saimiri boliviensis, and human erythrocytes with PvDBP-RII, a fluorescent anti-His tag detection antibody, and thiazole orange before flow cytometric analysis. T-EBAs employing immunoblot detection with an anti-His antibody were performed concomitantly. RESULTS: PvDBP-RII bound to A. nancymai, M. mulatta, and human Fy+ erythrocytes, but not human Fy- erythrocytes, by F-EBAs and T-EBAs. However, F-EBAs exhibited higher sensitivity and better concordance between experiments compared with T-EBAs. CONCLUSIONS: F-EBA is a rapid, simple, and reliable method for quantifying the ability of malaria proteins to bind to the surface of erythrocytes. F-EBA can discriminate binding between erythrocyte subpopulations without enrichment protocols and may be more reliable and sensitive than T-EBAs in identifying novel erythrocyte binding proteins.     

Studies of a Murine Monoclonal Antibody Directed against DARC: Reappraisal of Its Specificity

Duffy Antigen Receptor for Chemokines (DARC) plays multiple roles in human health as a blood group antigen, a receptor for chemokines and the only known receptor for Plasmodium vivax merozoites. It is the target of the murine anti-Fy6 monoclonal antibody 2C3 which binds to the first extracellular domain (ECD1), but exact nature of the recognized epitope was a subject of contradictory reports. Here, using a set of complex experiments which include expression of DARC with amino acid substitutions within the Fy6 epitope in E. coli and K562 cells, ELISA, surface plasmon resonance (SPR) and flow cytometry, we have resolved discrepancies between previously published reports and show that the basic epitope recognized by 2C3 antibody is ²²FEDVW²⁶, with ²²F and ²⁶W being the most important residues. In addition, we demonstrated that ³⁰Y plays an auxiliary role in binding, particularly when the residue is sulfated. The STD-NMR studies performed using 2C3-derived Fab and synthetic peptide corroborated most of these results, and together with the molecular modelling suggested that ²⁵V is not involved in direct interactions with the antibody, but determines folding of the epitope backbone.     

An antibody against the alzheimer's disease amyloid precursor protein recognizes distinct conformational isoforms

The Alzheimer's disease amyloid precursor protein (APP) consists of several isoforms, which are extensively post-translationally modified and processed. A monoclonal antibody, MAbE1, was raised against a synthetic peptide from an extracellular domain that is common to all isoforms of APP. Immunoblots and immunolocalization studies on cells of neuronal and other origins demonstrated that this antibody recognized a subclass of APP isoforms when compared to a monoclonal antibody raised against a bacterial fusion protein of APP, MAb22C11. Prominent protein bands of 71 kDa and 120 kDa were only detected on immunoblots of cell lysates and no immunoreactivity was observed in protein samples obtained from cell conditioned media. Immunofluorescence labelling with MAbE1 revealed predominantly perinuclear staining of cells of neuronal and glial origin. The data suggest that this monoclonal antibody detects distinct conformational isoforms of APP present in intracellular compartments.     

Immunochemical characterization of the human blood cell membrane glycoprotein recognized by the monoclonal antibody 12E7.

The 12E7 murine monoclonal antibody recognizes a protease-sensitive component of human red cells, platelets and lymphocytes which could not be detected on granulocytes. Scatchard analyses indicated that the 125I-labelled antibody binds to 1000, 4000 and 27,000 antigen sites on each red cell, platelet and lymphocyte respectively, with a binding constant ranging from 4 x 10(7) to 9 x 10(7) M-1. The membrane components recognized by the monoclonal antibody were characterized by immunostaining on nitrocellulose sheets. A 28 kDa sialoglycoprotein was visualized following electrophoretic transfer of the red cell and lymphocyte membrane proteins separated by SDS/polyacrylamide-gel electrophoresis. Another component of 25 kDa was also clearly identified in the lymphocyte and platelet lysates, but was barely detectable in the red cell membrane preparations. Enzyme treatment of intact platelets, as well as analysis of the membrane and cytosolic preparations from these cells, have shown that the 25 kDa component was of cytoplasmic origin. The mobility of the 28 kDa membrane component is decreased following neuraminidase treatment of intact blood cells, but these cells still react normally with the monoclonal antibody, indicating that sialic acids are not required for binding. The 28 kDa component is present on red cell membranes prepared from S-s-U-, En(a-) and Gerbich(-) individuals, demonstrating that it is a new sialoglycoprotein not derived from glycophorins A, B, C or D. The 28 kDa component was totally solubilized with 0.1% Triton X-100 from red cell membranes and behaves like the other red cell membrane sialoglycoproteins since it was extracted in the aqueous phase following chloroform/methanol/water or butanol/water partitionings. The 28 kDa component could be partially purified by h.p.l.c. gel permeation chromatography and preparative SDS/polyacrylamide-gel electrophoresis. The material finally obtained strongly inhibits the 12E7 monoclonal as well as human anti-Xga antibodies, suggesting either that the 28 kDa glycoprotein carries both antigens or that the 12E7 and Xga-active molecules copurified.     

Characterization and localization of tropomyosin proteins in Xenopus embryos with specific antibodies

In the process of monoclonal antibody (mAb) production against the 38kDa protein which is lacking in the gastrula arrested mutant embryos in Xenopus we incidentally obtained two kinds of mAb (designated as B11 and 2D10 antibodies, respectively) recognizing tropomyosin (TM) proteins in Xenopus embryos. The characterization of the corresponding antigens to those mAb was performed by immunoblotting and silver staining for two-dimensional (2-D) gels in the present study. The localization of the antigens in Xenopus embryos was also investigated by fluorescent microscopy.
By 2-D immunoblots with those mAb, three distinct protein spots or TM isoforms were recognized in Xenopus embryos; a 38 kDa spot with a pl of approximately 4.8 reacted with both antibodies in embryos at stages later than the mid-tailbud (stage 28) and two 30 kDa spots, which are probably isomers, with a pl of approximately 4.8 were detected with 2D10 antibody in embryos at stages extending from the fertilized to the mid-neurula (stage 20). By immunofluorescent microscopy, B11 antibody was shown to react mainly with muscle cells and their precursor cells. In contrast, 2D10 antibody stained the cytoplasm of almost all cells in embryos at stages from the fertilized to the tadpole.
Judging from the results obtained with immunoblotting and fluorescent microscopy, it is likely that the 38 kDa spot is a skeletal muscle TM isoform and the two 30 kDa spots are non-muscle TM isoforms.     

Plasmodium vivax: malarial proteins associated with the membrane-bound caveola-vesicle complexes and cytoplasmic cleft structures of infected erythrocytes

The identification of antigens of parasite origin associated with the altered membrane of Plasmodium vivax-infected erythrocytes was undertaken in this study. The 125I-lactoperoxidase catalyzed surface radiolabeling of trophozoite-infected erythrocytes revealed new bands of 95 and 70 kDa not labeled in normal erythrocytes. Erythrocyte membrane-enriched preparations from [35S]methionine biosynthetically labeled-infected erythrocytes also indicated that in addition to bands at 95 and 70 kDa, several other parasite proteins were possibly membrane associated. Five monoclonal antibodies (Mabs) reactive with P. vivax produced an immunofluorescent pattern of numerous small dots scattered over the entire infected erythrocyte. This pattern mimics that of Schuffner's stippling; small red dots seen in Giemsa-stained P. vivax-infected erythrocytes, which represent accumulations of dye in caveola-vesicle complexes (CVC). Four of the monoclonal antibodies immunoprecipitated a Triton X-100 detergent-insoluble 95-kDa parasite protein which was localized by immunofluorescent assay and immunoelectron microscopy exclusively to the CVC. Two of these Mabs were immunofluorescence reactive with the surface of intact infected erythrocytes in suspension. The fifth Mab, which also localized exclusively to the CVC structures, immunoprecipitated a Triton X-100 extractable protein of 70 kDa. Two other monoclonal antibodies reacted exclusively with the numerous membranous cleft structures found in the cytoplasm of infected erythrocytes. This cleft-associated parasite antigen was 28 kDa in size. Some of these Mabs recognize epitopes and produce similar IFA patterns on erythrocytes infected with P. cynomolgi, P. knowlesi, and P. ovale parasites, but not with P. falciparum- or P. brasilianum-infected erythrocytes.     

The leukotoxin of Pasteurella haemolytica binds to β2 integrins on bovine leukocytes

The putative receptor proteins of Pasteurella haemolytica leukotoxin were isolated from bovine polymorphonuclear neutrophil lysate by affinity chromatography on a leukotoxin-specific monoclonal antibody column to which the leukotoxin was pre-bound. SDS-PAGE of the purified proteins showed four bands at 180 kDa, 170 kDa, 150 kDa and 95 kDa, in addition to the expected 102-kDa leukotoxin band and a series of bands with molecular masses lower than 102 kDa representing the disintegrated leukotoxin. N-terminal amino acid sequencing of the 170-kDa band showed homology with human and murine CD11b. The purified proteins reacted specifically with monoclonal antibodies specific for CD11a, CD11b, CD11c (the alpha chains of beta(2) integrins), and CD18 (the beta chain of beta(2) integrins). Pre-incubation of polymorphonuclear neutrophils with a monoclonal antibody specific for CD18 reduced the cytotoxicity of the leukotoxin to the cells. These results indicate that the leukotoxin binds to the beta(2) integrins on bovine leukocytes, very likely via CD18.     

Mesogleal cells of the jellyfish Aurelia aurita are involved in the formation of mesogleal fibres

The extracellular matrix of the jellyfish Aurelia aurita (Scyphozoa, Cnidaria), known as the mesoglea, is populated by numerous mesogleal cells (Mc). We determined the pattern of the Mc and the mesoglea, raised polyclonal antibodies (RA47) against the major mesogleal protein pA47 (47 kDa) and checked their specificity. In the mesoglea, RA47 stains pA47 itself. In immunoblots of Mc, RA47 stains bands of 120 kDa and 80 kDa; weaker staining is observed at pA47. The same staining pattern is seen on blots of jellyfish epidermal cells and of whole Hydra (Hydrozoa) or isolated mesoglea of Hydra. Our data indicate that pA47 is synthesized by Mc and epidermal cells as high molecular precursors. Using immunostaining techniques, we showed Mc to be involved in the formation of mesogleal non-collagenous (called "elastic" in classic morphological studies) fibres. The biochemical and morphological data suggest that Mc originate from the epidermis.     

Isolation and characterization of deletion mutants affected in early genes of bacteriophage ?80

Summary When Escherichia coli cells that had been irradiated with ultraviolet light were infected with bacteriophage 80, five major (pE, pB, pA, pC and pD) and two minor (pU and pV) proteins were found to be synthesized during early stages of infection. The genss coding for the five major proteins were mapped on the 80 chromosome using various deletion mutants which lacked the capacity to synthesize some or all the major proteins. The size and positions of all the deletions were determined by gel electrophoresis of EcoRI digests of phage DNA and by electron microscopy of heteroduplexes between DNAs of the deletion and wild-type phage. The five major proteins designated pE(25K), pB(40K), pA(45K), pC(34K) and pD(31K) were shown to be encoded in this order presumably by a single operon that was located at 60.2–67.4% on the 80 genome. These proteins were found to be involved in phage recombination. The absence of pE or pB resulted in a Red^– phenotype and the absence of three proteins (pE, pB and pA) resulted in a Fec^– phenotype. The exact positions of the genes for the minor proteins pU(29K) and pV(26K) have not been determined.     

Plasmodium berghei XAT: protective 155/160 kDa antigens are located in parasitophorous vacuoles of schizont-stage parasite

Effective blood-stage malaria vaccine candidates have been mainly developed from the proteins in exposed locations on the parasite such as the surface of free merozoites or infected red blood cells. In the present study, we identified and localized novel protective antigens derived from the blood-stage of Plasmodium berghei XAT after establishment of hybridomas producing protective monoclonal antibodies (mAbs) against the parasites. The protective antigens were expressed in schizonts but not in trophozoites, and located in the parasitophorous vacuoles in the infected erythrocyte cytoplasm. The antigens, with molecular weight of 155/160 kDa, were not identical to any merozoite/schizont antigens that have been reported as target molecules recognized by mAbs developed to rodent malaria parasites. The characterization of new malarial antigenic targets of potentially protective antibody responses following infection would give us new insights for the selection of candidate antigens for malaria vaccine.     

The human Duffy blood group in rhesus monkeys1

There is good evidence that susceptibility to Plasmodium vivax infection and to P. knowlesi erythrocyte invasion is influenced by certain human Duffy (Fy) blood group antigens. Since P. knowlesi readily infects rhesus monkeys (Macaca mulatto), it was not surprising to find an Fy-like antigen on rhesus erythrocytes. Using human Fy antisera in elution and absorption experiments, we found that all 40 rhesus monkeys tested displayed the Fy(a-b +) phenotype. Furthermore, the rhesus Fy^bantigen was inactivated by chymotrypsin but not by trypsin, suggesting that it is homologous to the human Fy^bantigen. Preliminary serological analyses and enzyme hydrolysis experiments suggest that none of the 13 blood group systems that we have described in rhesus are analogous to the human Fy system. Thus, it appears that there is no Duffy-like polymorphism in rhesus monkeys.     

The mutation G298A-->Ala100Thr on the coding sequence of the Duffy antigen/chemokine receptor gene in non-caucasian Brazilians

Ala100Thr has been suggested to be a Caucasian genetic marker on the FY*B allele. As the Brazilian population has arisen from miscegenation among Portuguese, Africans, and Indians, this mutation could possibly be found in Euro- and Afro-Brazilians, or in Brazilian Indians. Fifty-three related individuals and a random sample of 100 subjects from the Brazilian population were investigated using the polymerase chain reaction and four restriction fragment length polymorphisms. Confirming the working hypothesis, among the related individuals three Afro-Brazilians (two of them a mother and daughter) and a woman of Amerindian descent had the Ala100Thr mutation on the FY*B allele. Five non-related Euro-Brazilians also carried the mutation. All nine individuals presented the Fy(a-b+) phenotype. We conclude that the Ala100Thr mutation can occur in populations other than Caucasians and that this mutation does not affect Duffy expression on red blood cells. Gene frequencies for this allele in the non-related individuals were in agreement with those of other populations. The Duffy frequencies of two Amerindian tribes were also investigated.     

Quantitation and intracellular localization of the 85K heat shock protein by using monoclonal and polyclonal antibodies.

Two monoclonal antibodies have been produced against the human 85,000-molecular-weight heat shock protein (hsp85). One of these, 16F1, cross-reacts with the murine homolog and is shown by peptide map immunoblots to be directed against an epitope different from that recognized by the other monoclonal antibody, 9D2. Both monoclonal antibodies recognize only a single Mr-85,000 species in two-dimensional immunoblots. Immunoprecipitation did not reveal an association of this heat shock protein with any other protein in HeLa cells. Immunoperoxidase staining showed a purely cytosolic distribution at both light and electron microscopic levels and no association with membranes, mitochondria, or other organelles. The 9D2 monoclonal and a polyclonal antimurine hsp85 antibody were used to identify the antigens and to quantitate their levels in a variety of normal tissues by immunoautoradiography. Relative abundance in the various tissues as determined by Coomassie blue staining correlates reasonably well with the immunoreactivity. Testis and brain, for example, have high hsp85 levels, whereas heart and skeletal muscle have little or none. The Mr-85,000 sodium dodecyl sulfate-polyacrylamide gel band in testis and brain lysates was further confirmed to be hsp85 by one-dimensional partial proteolytic peptide mapping. Based on these data and our previous observations showing that synthesis and levels of the protein are altered by depriving cultured cells of glucose, we speculate that intracellular hsp85 levels depend on differences in the intermediary metabolism of glucose in the various tissues. Furthermore, it appears that high basal levels of this heat shock protein may not necessarily protect cells against heat shock, since testis is one of the most heat-sensitive tissues and has the highest hsp85 level.