A novel HLA-B27 allele maps B27 allospecificity to the region around position 70 in the alpha 1 domain.

There are six known HLA-B alleles that share the HLA-B27 allospecificity, yet differ by one to six amino acid substitutions. Each of these B27 alleles can be readily assigned by one of the six representative IEF patterns. Two unrelated individuals, LH and HS, express B27 Ag that appear to be identical by IEF, but an HLA-B27 alloreactive CTL clone I-73 was found to react differently with these cells, suggesting these B27 molecules are not identical. We sequenced polymerase chain reaction-amplified B27 cDNA clones obtained from HS and compared its deduced amino acid sequence (B27-HS) with the B27 sequence of LH (B27-LH) which was previously designated the B*2701 allele. B27-HS and B27-LH differ by eight amino acids; three in alpha 1 domain and five in alpha 2 domain. These amino acid substitutions of B27-HS altered T cell recognition but not the B27 serologic epitope or IEF pattern. B27-HS differs from the six known B27 alleles by five to eight amino acid substitutions, and thus it represents the seventh allele of the HLA-B27 Ag family. This novel B27 allele might have been derived from a gene conversion event. Previously, two amino acid residues at positions 70 and 97 were suggested to be specific for B27 Ag family. B27-HS now reveals that Lys at position 70 is specific for B27 but Asn at position 97 is not. We propose that the region around position 70 might be crucial in determining the B27 serologic epitope and possibly in peptide Ag binding. This study also demonstrates that class I molecules of the same Ag specificity sharing an indistinguishable IEF pattern are not necessarily identical, and indicates that only the definitive determination of primary structure would identify all the class I alleles that are functionally relevant in regard to alloreactivity, T cell restriction, and disease association.     

Molecular characterization by high resolution isoelectric focusing of the products encoded by the class II region loci of the MHC in humans. II. DP alpha and DP beta gene variants

To characterize the molecular polymorphism of the DP alpha and DP beta gene products, the HLA-DP molecules expressed by more than 200 cell lines were individually immunoprecipitated by using the mAb B7/21 and their neuraminidase-treated DP alpha and DP beta chains analyzed in IEF gels. These cell lines, most of them from members of 32 families, allowed the definition, by segregation analysis, of the IEF patterns of the DP polypeptide chains encoded by 129 distinct haplotypes. Both DP alpha and DP beta chains display polymorphic IEF-banding patterns. Two DP alpha (A and B) and seven DP beta (A, B, C, D, E, F, and G) IEF variants were characterized. The DP alpha B variant was found in linkage disequilibrium with both DP beta B and DP beta D. Linkage disequilibrium was also encountered with alleles at the DR and DQ loci. Finally, the correlations between the IEF DP alpha and DP beta variants and the primed lymphocyte test-defined HLA-DP specificities were determined by using a panel of 24 primed lymphocyte test-typed cell lines.     

Use of isoelectric focusing to define major histocompatibility complex class I polymorphism in goats

We have used biochemical methods to extend and improve serological class I typing using a panel of 77 Swiss goats of the Saanen breed, comprising dam-offspring combinations from six half-sib sire families and several unrelated animals. Of these animals class I molecules were precipitated from cell lysates with the mAb B1.1G6 and HC10. Immunoprecipitates were analysed by SDS-PAGE and 1D-IEF. There was a good agreement between class I serological types and IEF banding patterns. We have identified three new class I specificities and subdivided the Bel 7 specificity. IEF has enabled us to make planned immunizations to produce antisera to the new specificities. New evidence for the expression of a second class I locus product in the Be7 haplotype has been found.     

Comparison of one-dimensional IEF patterns for serologically detectable HLA-A and B allotypes

A new mouse monoclonal antibody (MoAb) 4E, which detects an epitope shared by HLA-B locus antigens, together with the MoAb W6/32, detecting a common HLA, B, C, determinant, and the MoAb 4B, detecting HLA-A2 and A28, were used to isolate HLA-A and -B antigens in sequential immunoprecipitation. The HLA antigens obtained from metabolically labeled cell extracts of B-lymphoblastoid cell lines or from phytohemagglutinin (PHA) activated peripheral blood lymphocytes were compared by one-dimensional isoelectric focusing (1D-IEF). The IEF banding patterns obtained with native HLA antigens segregated in a family with HLA. Neuraminidase treatment of isolated antigens reduced the number of bands to one or two, simplifying the analysis of characteristic patterns. Thus, we have cataloged IEF banding patterns for the majority of the serologically recognized HLA-A and -B allotypes obtained from 57 unrelated American Caucasians. While no HLA-A locus or HLA-B locus specific banding patterns were observed, the HLA-A antigens had, in general, slightly higher pl values than the HLA-B antigens. HLA-C antigens could not be detected in this assay system. The polymorphism detected by IEF banding patterns was as extensive as the serologically detected polymorphism identified by classical HLA serology. Moreover, variants for some HLA allotypes could be detected by this method. In addition to previously recognized A2 variants, new variants were identified for HLA-A1, A26, and Bw44. Each A1 and Bw44 variant was associated with particular haplotypes. The HLA-A2 antigens occurred on 43 HLA haplotypes in the unrelated Caucasian population. Only one of each A2 variants was identified in this population.     

Molecular basis of the isozymes of bovine glucose-6-phosphate isomerase

Glucose-6-phosphate isomerase occurs in different bovine tissues as multiple, catalytically active isozymes which can be resolved by polyacrylamide gel electrophoresis (PAGE) and isoelectric focusing (IEF). Most differentiated tissues have five distinct forms with apparent pI values of 7.2, 7.0, 6.8, 6.6, and 6.4. Young, mitotically active, cells of the intestinal mucosa and the epithelium of the eye lens show only the two more basic isozymes, while old cells in the cortex and nucleus of the eye lens accumulate the more acidic isozymes. All of the isozymes exhibit equal separation based on charge-to-mass ratio (PAGE) and charge (IEF), thus indicating only charge changes. The isozyme patterns are unchanged in the presence of reducing agents or protease inhibitors. Each isozyme was purified to homogeneity and shown to exhibit identical subunit molecular weights (59,000) on SDS-gel electrophoresis. Each of the isolated isozymes, when subjected to PAGE or IEF, exhibited a single band, indicating that the isozymes are not generated as a result of electrophoresis. When the most basic isozyme was incubated in vitro under mild alkaline conditions, there was a spontaneous generation of the more acidic isozymes with properties identical to those found in vivo. The isozymes, thus, appear to be the result of spontaneous, postsynthetic modifications involving the addition of equal numbers of negative charges and are consistent with the deamidation of specific asparagine and/or glutamine residues.     

Changes in expression and microheterogeneity of the genetic variants of human α1-acid glycoprotein in malignant mesothelioma

Human α₁-acid glycoprotein (AAG), an acute-phase plasma protein, is heterogeneous in the native state and polymorphic in the desialylated state. The AAG heterogeneity is mainly explained by a variable glycan chain composition in its five glycosylation sites. The AAG polymorphism is due to the presence of genetic variants. Three main variants are observed for AAG, ORM1 F1, ORM1 S and ORM2 A, which have a separate genetic origin. In this paper, we have used different isoelectric focusing (IEF) methods and chromatography on immobilized metal affinity adsorbent to study the relative occurrence of the genetic variants of AAG in relation to changes in microheterogeneity, in plasma and pleural effusions of patients with malignant mesothelioma (MM). The results were compared to those obtained with the variants in plasma of healthy individuals. Significant changes in variant distribution were observed in the MM samples, that corresponded to a rise in the proportion of the ORM1 variants and a fall in that of the ORM2 variant. However, the concentration in MM plasma increased for both variants. The AAG in MM plasma and effusion fluids was found to be more heterogeneous on IEF than AAG of healthy plasma. The evidence of stronger concentrations of both the high and low pI forms of AAG in the MM samples suggested two kinds of changes in charge heterogeneity. These two changes were shown to be attributed to different variants — i.e. the high pI forms to ORM1 F1 and S and the low pI forms to ORM2 A, after fractionation of AAG by chromatography on immobilized copper(II) ions. These results indicate specific changes in both the expression and glycosylation for each AAG variant, according to its separate genetic origin, in MM.     

Studies on the variants of the protein toxins ricin and abrin.

This study elucidates some structural and biological features of galactose-binding variants of the cytotoxic proteins ricin and abrin. An isolation procedure is reported for ricin variants from Ricinus communis seeds by using lactamyl-Sepharose affinity matrix, similar to that reported previously for variants of abrin from Abrus precatorius seeds [Hegde, R., Maiti, T. K. & Podder, S. K. (1991) Anal. Biochem. 194, 101-109]. Ricin variants, subfractionated on carboxymethyl-Sepharose CL-6B ion-exchange chromatography, were characterized further by SDS/PAGE, IEF and a binding assay. Based on the immunological cross-reactivity of antibody raised against a single variant of each of ricin and abrin, it was established that all the variants of the corresponding type are immunologically indistinguishable. Analysis of protein titration curves on an immobilized pH gradient indicated that variants of abrin I differ from other abrin variants, mainly in their acidic groups and that variance in ricin is a cause of charge substitution. Detection of subunit variants of proteins by two-dimensional gel electrophoresis showed that there are twice as many subunit variants as there are variants of holoproteins, suggesting that each variant has a set of subunit variants, which, although homologous, are not identical to the subunits of any other variant with respect to pI. Seeds obtained from polymorphic species of R. communis showed no difference in the profile of toxin variants, as analyzed by isoelectric focussing. Toxin variants obtained from red and white varieties of A. precatorius, however, showed some difference in the number of variants as well as in their relative intensities. Furthermore, variants analyzed from several single seeds of A. precatorius red type revealed a controlled distribution of lectin variants in three specific groups, indicating an involvement of at least three genes in the production of Abrus lectins. The complete absence or presence of variants in each group suggested a post-translational differential proteolytic processing, a secondary event in the production of abrin variants.     

Charge heterogeneity of cholera toxin and its subunits

Abstract Analytical isoelectric focusing (IEF) in thin layers of polyacrylamide gels resolved cholera toxin into 3 isomeric forms differing in charge (isoelectric points 6.80, 6.65 and 6.55). All these forms had identical molecular weights, and were also antigenically similar, as demonstrated by their reactivity to antisera to cholera toxin. Both the B and A subunits possessed charge heterogeneity. The B subunit was detected in a free form when a solution of cholera toxin was aged for a few days. Antisera to cholera toxin, irrespective of mode of immunisation, contained antibodies to both the intact cholera toxin and the free B subunit as demonstrated by the immunoblotting technique based on IEF.     

Distribution of thermophilic aerobic sporeforming bacteria in food ingredients

Samples of sugar, starch, spices, and miscellaneous products were tested for thermophilic sporeformers of Bacillus to determine the dominant species present. Surface colonies selected at random were identified. Six species of Bacillus were isolated: B. stearothermophilus, B. coagulans, B. licheniformis, B. subtilis, B. circulans, and B. pumilus. Samples of starch and pepper were tested for thermophilic sporeformers of Bacillus to determine the distribution of rough and smooth variants. Colonies were classified as rough or smooth variants by colonial characteristics. The distribution of variant forms in these two products was significantly different. Starch samples showed predominantly rough variants; pepper samples showed predominantly smooth variants.     

BoLA class I charge heterogeneity reflects the expression of more than two loci

Internationally recognized allo-antisera in lymphocyte microcytotoxicity assays are thought to detect allelic products of a single highly polymorphic class I locus. A recent report suggested that two bovine lymphocyte antigen (BoLA) class I loci are expressed at the protein level. However, 1D-IEF analysis of BoLA class I molecules reveals multi-band patterns which cannot be reconciled with the reported number of loci. The aim of this study was to investigate the origins of the charge diversity of BoLA class I molecules observed using 1D-IEF. BoLA class I molecules appear to be glycosy-lated at a single N-linked position with a complex type carbohydrate moiety which has up to three terminal sialic acid residues. Class I molecules immunoprecipitated from resting bovine PBL are not phosphorylated. Neither modification is responsible for the observed charge heterogeneity. Peptide mapping reveals that different BoLA charge variants have distinct digestion patterns. Furthermore, a number of different polypeptides are associated with each serological specificity. These polypeptides appear to be encoded by different loci which exist in linkage disequilibrium. The number of charge variants with different peptide maps indicates that the BoLA system has a minimum of three class I loci expressed at the protein level.     

Detection of 10 variants of biliverdin reductase in rat liver by two-dimensional gel electrophoresis

We have identified and characterized multiple forms of biliverdin reductase (BVR) in control rat liver cytosol. Two-dimensional electrophoresis of the purified BVR resolved a minimum of 10 discrete protein zones. All 10 proteins were BVR as judged by immunological cross-reactivity toward rabbit anti-rat BVR. Based on the isoelectric focusing pattern of separation, the BVR variants could be organized into five net-charge groups designated as BVR-IEF1 to BVR-IEF5 and three molecular mass groups designated as BVR-MW1-BVR-MW3, respectively. The pI values of the net-charge groups were: BVR-IEF1, 6.23; IEF2, 5.91; IEF3, 5.76; IEF4, 5.61; IEF5, 5.48. The Mr values of the molecular mass groups were: BVR-MW1, 30,400; MW2, 30,700; MW3, 31,400. Single dimension slab gel isoelectric focusing offered greater resolution of the net charge variants, and BVR-IEF3 was further resolved into two variants, IEF3a and IEF3b, with pIs of 5.77 and 5.75, respectively. The six net-charge variants also resolved on a preparative chromatofocusing column and were designated as BVR-CF1-BVR-CF6. The pH values of the peak fractions were: BVR-CF1, 6.91; CF2, 6.33; CF3, 6.03; CF4, 5.82; CF5, 5.45; CF6, 5.27. Correspondence between the isoelectric focusing net-charge variants and the chromatofocusing net-charge variants was established. The Mr and net-charge variants did not represent partially degraded forms of biliverdin reductase produced during purification since the pattern of resolution of variants on slab gel isoelectric focusing or two-dimensional electrophoresis did not change by purifying the proteins in the presence of protease inhibitors and 5 mM EDTA. BVR-CF2 and BVR-CF4 were purified and examined for pH-dependent cofactor requirements for activity. Both net-charge variants and two pH optima that were cofactor-dependent; maximum activity with NADPH, however, was at pH 8.5 and with NADH at pH 6.7. With both variants, however, a higher catalytic rate was observed with NADH than with NADPH at their respective pH optima. Furthermore, BVR-CF2 exhibited a higher catalytic rate than did BVR-CF4 with either cofactor throughout the pH range of 5-9.     

Molecular and evolutionary analyses of a variable series of genes in Borrelia burgdorferi that are related to ospE and ospF, constitute a gene family, and share a common upstream homology box.

In this study we report on the molecular characterization of a series of genes that constitute a gene family related to ospE and ospF. Some members of this family appear to represent recombined or variant forms of ospE and ospF. Variant ospE and ospF genes were found in several Borrelia burgdorferi isolates, demonstrating that their occurrence is not a phenomenon relevant to only a single isolate. Hybridization analyses revealed that the upstream sequence originally identified 5' of the full-length ospEF operon exists in multiple copies ranging in number from two to six depending on the isolate. This repeated sequence, which we refer to as the upstream homology box (UHB), carries a putative promoter element. In some isolates, UHB elements were found to flank copies of ospE and ospF that exist independently of each other. We refer to this group of UHB-flanked genes collectively as the UHB gene family. The evolutionary relationships among UHB gene family members were assessed through DNA sequence analysis and gene tree construction. These analyses suggest that some UHB-flanked genes might actually represent divergent forms of other previously described genes. Analysis of the restriction fragment length polymorphism patterns of the UHB-flanked genes among B. burgdorferi isolates demonstrated that these patterns are highly variable among isolates, suggesting that these genes are not phylogenetically conserved. The variable restriction fragment length polymorphism patterns could indicate recombinational activity in these sequences. The presence of numerous copies of the UHB elements and the high degree of homology among UHB-flanked genes could provide the necessary elements to allow for homologous recombination, leading to the generation of recombination variants of UHB gene family members.     

Clonal diversity in human B cell lymphoma. I. Idiotypic and genetic analysis of lymphoma heterohybrids

Secretory heterohybrid clones from seven pristine human B cell lymphomas of diverse histologic types were established to investigate the question of tumor clonal diversity. We found that in six tumors, heterohybrid-derived Ig showed similar band patterns in IEF; families of anti-Id prepared from tumor Ig reacted uniformly with individual heterohybrids and original tumor; and the V gene loci displayed little variation on Southern analysis. In one patient who was followed with serial multiple site biopsies over a 14-mo period, clonal Id was preserved until the final stage of his disease, in spite of cytotoxic treatment. In a single follicular tumor (J.M.), each of the anti-Id reacted uniformly with the parent tumor and the individual heterohybrids, except that three of six clones failed to react with a single anti-Id family member. A Southern analysis of the VH gene locus revealed an identical gene rearrangement that was shared by the parent tumor and each heterohybrid. However, there was considerable heterogeneity of J.M. heterohybrid Ig in IEF gels, and we demonstrated the production of variant lambda L chains by the heterohybrid clones. One type of lambda L chain had a normal mobility in SDS-PAGE gels but larger lambda variants were produced by four of six heterohybrids. A Southern analysis of the VL gene displayed considerable variation in the type of lambda rearrangement present in the various heterohybrids, suggesting extensive diversity at the VL gene locus. In a second tumor (S.C.) that exhibited uniform anti-Id tumor reactivity we were also able to demonstrate the presence of a second minor tumor cell population (a biclonal tumor). Our data suggest that intraclonal VH variation may vary considerably with lymphoma subtype and mutagenic exposure and that an additional mechanism for generating spontaneous intraclonal heterogeneity is genetic variation at the VL locus.     

Characterization of class I bovine lymphocyte antigens (BoLA) by one-dimensional isoelectricfocusing

BoLA class I antigens were characterized in a group of British and Dutch Friesian cattle by one-dimensional isoelectric focusing (1D-IEF) and the results compared with serology using alloantisera and microcytotoxicity. For IEF analysis, non-stimulated peripheral blood mononuclear cells (PBM) were metabolically labelled with 35S methionine, detergent lysates were prepared and MHC molecules precipitated with the monoclonal antibodies (mAbs) W6/32 or B1.1G6. Staphylococcus protein A precipitated antigens were separated on a vertical slab gel under denaturing conditions. The banding patterns seen for the W6/32 precipitated molecules obtained by 1D-IEF were compared with the serological specificities. Characteristic banding patterns were observed for most serological specificities as well as workshop undefined haplotypes. These patterns were seen both in families and the outbred population. In families IEF haplotypes segregated with serotypes. Additional MHC class I products were suggested by variable banding patterns for different w10 haplotypes and when using the different mAbs. A pulse chase experiment with a w12 animal also suggested more than one expressed product. The w2 and w5 specificities were not precipitated by either W6/32 or B1.1G6 and w6.2 and w6.4 were precipitated by W6/32 but not by B1.1G6. These results show that 1D-IEF is useful for BoLA typing. For the characterization of class I antigens, however, much depends on the mAbs used.     

Structural optimization of an aptamer generated from Ligand-Guided Selection (LIGS) resulted in high affinity variant toward mIgM expressed on Burkitt's lymphoma cell lines

Aptamers are synthetic, short nucleic acid molecules capable of specific target recognition. Aptamers are selected using a screening method termed Systematic Evolution of Ligands by Exponential enrichment (SELEX). We recently have introduced a variant of SELEX called “Ligand-Guided-Selection” (LIGS) that allows the identification of specific aptamers against known cell-surface proteins. Utilizing LIGS, we introduced three specific aptamers against membrane-bound IgM (mIgM), which is the hallmark of B cells. Out of the three aptamers selected against mIgM, an aptamer termed R1, in particular, was found to be interesting due to its ability to recognize mIgM on target cells and then block anti-IgM antibodies binding their antigen. We systematically truncated parent aptamer R1 to design shorter variants with enhanced affinity. Importantly, herein we show that the specificity of the most optimized variant of R1 aptamer is similar to that of anti-IgM antibody, indicating that the specificity of the ligand utilized in selective elution of the aptamer determines the specificity of the LIGS-generated aptamer. Furthermore, we report that truncated variants of R1 are able to recognize mIgM-positive human B lymphoma BJAB cells at physiological temperature, demonstrating that LIGS-generated aptamers could be re-optimized into higher affinity variants. Collectively, these findings show the significance of LIGS in generating highly specific aptamers with potential applications in biomedicine.     

Comparative studies on glycerol 3-phosphate dehydrogenase in bees and wasps

Electrophoresis of tissue extracts has shown the presence of multiple electrophoretic forms of glycerol 3-phosphate dehydrogenase (EC 1.1.1.8) in many Hymenoptera. The patterns are most complex in the two bumblebee genera, Bombus and Psithyrus, where from five to six variants are observed.Homogeneous preparations of the major flight muscle variant of glycerol 3-phosphate dehydrogenase have been isolated from thoraces of three bumblebee species and of yellow jacket. The amino acid compositions of these four enzymes plus that from the honeybee have been determined and compared.The Michaelis constant for dihydroxyacetone phosphate was measured for the honeybee, bumblebee, and yellow jacket enzymes. Differences were observed between species but not between bumblebee isozymes.B. nevadensis glycerol 3-phosphate dehydrogenase binds reversibly to both B. nevadensis and honeybee actin. The alar-muscular variants bind more strongly than the omniregional variants.     

Genetic polymorphism of human C4-binding protein

Two different forms of human C4-bp, C4-bp A and C4-bp B, have been identified by isoelectric focusing (IEF) of neuraminidase-treated EDTA-plasma samples. Family studies demonstrate Mendelian segregation of these forms, indicating that they are under gentic control. This conclusion is supported by IEF analysis of the two variants purified by affinity chromatography. Under completely denaturing conditions, C4-bp B was found to be composed of two subunits that focused at different pH, whereas C4-bp A contains only the more basic one. These results suggest that a single autosomal locus with at least two codominant alleles coding for the subunits controls the IEF variation of C4-bp in humans. The allele designated C4BP*1 codes for a subunit that, after neuraminidase treatment, focuses at pH = 6.65. The allele C4BP*2 codes for a different subunit that focuses at pH = 6.60. The C4-bp A phenotype corresponds to the genotype C4BP*1,C4BP*1 and the phenotype C4-bp B to the genotype C4BP*1,C4BP*2. The phenotype corresponding to the C4BP*2,C4BP*2 homozygous genotype has not been encountered thus far. Initial linkage data indicate that the C4BP locus is not closely linked to either the HLA or to the C3 loci.     

Characterization of the molecular basis of the alpha 1-antitrypsin F allele.

alpha 1-Antitrypsin (alpha 1AT), the major serum inhibitor of neutrophil elastase, is a highly polymorphic serum protein associated with characteristic isoelectric-focusing (IEF) patterns for most variants. To characterize the molecular basis of the anodal F variant, the DNA sequence of the coding exons of an FZ individual was determined. The F allele differed from the normal M1(Val213) alpha 1AT allele by a single nucleotide transversion of cytosine to thymidine, which results in the amino acid substitution Arg223 CGT----Cys TGT. Inheritance of the F mutation was confirmed by family analysis using allele-specific amplification. In the context that the normal alpha 1AT molecule has only one cysteine residue, a mutation resulting in the addition of a second cysteine may influence the three-dimensional form of the protein and/or permit interaction with other plasma proteins with free-SH groups and may be responsible for the observation that the major F alpha 1AT bands often migrate as doublets in IEF gels.     

Variants of peroxiredoxins expression in response to hydroperoxide stress

We examined patterns of the proteins that were expressed in human umbilical vein endothelial cells (HUVEC) in response to oxidative stress by two-dimensional polyacrylamide gel electrophoresis (2D-PAGE). When HUVEC were exposed to H2O2 at 100 microM for 60 min, the intensities of eight spots increased and those of eight spots decreased on 2D gels, as compared with control gels, after staining with silver. These changes were also observed after exposure of cells to hydroperoxides such as cumene hydroperoxide and tert-butyl hydroperoxide, but not after exposure to other reagents that induce oxidative stress such as S-alkylating compounds, nitric oxide, and salts of heavy metals. Therefore, these proteins were designated hydroperoxide responsive proteins (HPRPs). Microsequencing analysis revealed that these HPRPs corresponded to at least six pairs of proteins. Of these, four pairs of HPRPs were thioredoxin peroxidase I (TPx I), TPx II, TPx III, and the product of human ORF06, all of which belong to the peroxiredoxin (Prx) family and all of which are involved in the elimination of hydroperoxides. The other two pairs corresponded to heat shock protein 27 (HSP27) and glyceraldehyde-3-phosphate dehydrogenase (G3PDH), respectively. The variants that appeared in response to hydroperoxides had molecular masses similar to the respective native forms, but their pI values were lower by 0.2-0.3 pH units than those of the corresponding native proteins. These variants were detected on 2D gels after cells had been exposed to hydroperoxides in the presence of an inhibitor of protein synthesis. All variants were generated within 30 min of exposure to 100 microM H2O2. The variants of TPx I and TPx II appeared within 2 min of the addition of H2O2 to the culture medium. The HPRPs returned to their respective native forms after the removal of stress. Our results indicated that at least six proteins were structurally modified in response to hydroperoxides. Analysis by 2D-PAGE of 32P-labeled proteins revealed that the variant of HSP27 was its phosphorylated form while the other HPRPs were not modified by phosphorylation. Taken together, the results suggest that 2D-PAGE can reveal initial responses to hydroperoxide stress at the level of protein modification. Moreover, it is possible that the variants of four types of Prx might reflect intermediate states in the process of hydroperoxide elimination.