Control of the tissue inhibitor of metalloproteinases-1 promoter in culture-activated rat hepatic stellate cells: regulation by activator protein-1 DNA binding proteins.

In the injured liver hepatic stellate cells (HSCs) undergo a dramatic phenotypic transformation known as "activation" in which they become myofibroblast-like and express high levels of the tissue inhibitor of metalloproteinase 1 (TIMP-1). HSC activation is accompanied by transactivation of the TIMP-1 promoter. Truncation mutagenesis studies delineated a minimal active promoter consisting of nucleotides -102 to +60 relative to the major start site for transcription. Removal of an AP-1 site located at nucleotides -93 to -87 caused almost a complete loss of promoter activity. Analysis of AP-1 DNA binding activities during culture activation of HSCs initially indicated transient expression of proteins capable of forming a low mobility AP-1 DNA binding complex (LMAP-1). LMAP-1 was maximally induced at 24 hours of culture and then fell to undetectable levels at 120 hours. Western blot studies showed that both c-Fos and c-Jun underwent similar transient inductions. These temporal changes in c-Fos and c-Jun activities were unexpected because TIMP-1 mRNA expression is not detected in HSCs until culture day 3 to 5 and is thereafter sustained at a high level. Previous work in other cell lineages has established a key role for Pea3 binding proteins (Ets-1) in AP-1 mediated transactivation of the TIMP-1 promoter. We show that HSCs express relatively low levels Ets-1 and Ets-2 and show that mutagenesis of the Pea3 DNA binding site in the TIMP-1 promoter has less than a twofold effect on its activity in activated HSCs. Further analysis of AP-1 DNA binding activities in 7- to 14-day culture activated HSCs led to the discovery of high mobility AP-1 complexes (HMAP-1). HMAP-1 DNA binding activities were sequence specific with respect to AP-1 and absent from freshly isolated HSCs. Supershift EMSA and Western blot studies identified JunD, Fra2, and FosB as potential components of the HMAP-1. Mutations of the AP-1 site of the TIMP-1 promoter that prevented formation of HMAP-1 caused a 70% loss of activity in transfected activated HSCs. Taken together the data indicate that sustained upregulation of TIMP-1 gene expression may be at least partially controlled by a novel AP-1 dependent regulation of TIMP-1 promoter activity.

The cerebroside sulfate activator from pig kidney: purification and molecular structure.

The activator protein for hydrolysis of cerebroside sulfate by arylsulfatase A was purified from pig kidney in high yield. This protein, also known as sphingolipid activator protein-1 and saposin-B, was particularly rich in pig kidney. Purification was achieved by a simple procedure involving homogenation and heat treatment followed by affinity, ion exchange, and gel filtration chromatographies. The final product was better than 90% pure by gel electrophoresis and HPLC. It was possible to sequence more than 60 amino acids from the N-terminus with only a few uncertain residues. The sequence differed from that predicted for the human protein by about 10%, with most amino acid variations being conservative. There appeared to be a residual glycosyl substituent on asparagine 21, but the sugar content was low and the protein failed to bind to concanavalin A. The cerebroside sulfate activator proved to be exceptionally resistant to denaturation or protease digestion. The apparent molecular mass was approximately 20,000 Da on preparative gel-filtration columns, but was variable when estimated by HPLC gel filtration. Values ranging from 30,000 to over 100,000 Da were observed in neutral buffers, while values around 15,000-16,000 Da were seen in acidic buffers such as those used for assay of the biological activity. This was further decreased to a putative subunit of 7000-8000 Da under severe denaturing conditions. Pig kidney is a convenient source for the large-scale preparation of this interesting protein which has heretofore been obtained from human sources.

Sequence of the D-aspartyl/L-isoaspartyl protein methyltransferase from human erythrocytes. Common sequence motifs for protein, DNA, RNA, and small molecule S-adenosylmethionine-dependent methyltransferases.

A widely distributed protein methyltransferase catalyzes the transfer of a methyl group from S-adenosyl-methionine to the free carboxyl groups of D-aspartyl and/or L-isoaspartyl derivatives of L-aspartyl and L-asparaginyl residues. This enzyme has been postulated to function in the repair or the catabolism of age-damaged proteins. We present here the complete amino acid sequence of the more basic isozyme I of this enzyme from human erythrocytes. The sequence was determined by Edman degradation and mass spectral analysis of overlapping trypsin, Staphylococcus aureus V8 protease, Pseudomonas fragi endoproteinase Asp-N, cyanogen bromide, and hydroxylamine-generated fragments. The NH2-terminus is modified by acetylation and the protein contains 226 amino acids for a calculated molecular weight of 24,575. This value is in good agreement with the molecular weight determined for the purified protein by polyacrylamide gel electrophoresis in the presence of dodecyl sulfate and by gel filtration chromatography under nondenaturing conditions. The identification of 2 different amino acid residues at both positions 22 and 119 may indicate the presence of allelic variants or of two or more closely related structural genes. Finally, comparison of this sequence with those of methyltransferases for RNA, DNA, and small molecules, as well as other S-adenosylmethionine-utilizing enzymes, shows that many of these proteins share elements of three regions of sequence similarity and may be structurally or evolutionarily related.

Specificity of endoproteinase Asp-N (Pseudomonas fragi): cleavage at glutamyl residues in two proteins.

Endoproteinase Asp-N, a metalloprotease from a mutant strain of Pseudomonas fragi, has been reported to specifically cleave on the N-terminal side of aspartyl and cysteic acid residues. We utilized this enzyme to generate fragments for determining the amino acid sequence of the D-aspartyl/L-isoaspartyl methyltransferase isozyme I from human erythrocytes. Surprisingly, we identified cleavage sites for this enzyme at the N-terminal side of several glutamyl residues in addition to the expected cleavage sites at aspartyl residues. The ability of this enzyme to cleave polypeptides at both glutamyl and aspartyl residues was confirmed by mapping additional sites on erythrocyte carbonic anhydrase I. These results indicate that a more appropriate name for this enzyme may be Endoproteinase Asp/Glu-N.

Proteolytic processing of the beta-subunit of the lysosomal enzyme, beta-hexosaminidase, in normal human fibroblasts.

We have characterized the proteolytic processing of the beta-subunit of beta-hexosaminidase by identifying the amino termini of the various forms synthesized in cell-free translation and in cultured human fibroblasts. The procedures used had been developed for similar studies of the alpha-subunit (Little, L. E., Lau, M. M. H., Quon, D. V. K., Fowler, A. V., and Neufeld, E. F. (1988) J. Biol. Chem. 263, 4288-4292). Radioactive amino acids were incorporated biosynthetically into the different forms of the beta-subunit, which were isolated by immunoprecipitation, gel electrophoresis, and electroelution, and analyzed by automated Edman degradation. Translation by reticulocyte lysate in the presence of canine pancreas microsomes gave a product with alanine 43 at the amino terminus. The lysate could initiate translation at methionine 1 or methionine 13, depending on the SP6 mRNA provided. The product of signal peptidase action, the precursor form of the beta-subunit with amino-terminal alanine 43, was found in NH4+-induced secretions of cultured fibroblasts; intracellularly, this form was trimmed of two additional amino acids. The mature form was found to consist of three polypeptides joined by disulfide bonds; the amino termini were found to be valine 48, threonine 122, and lysine 315. Thus, in contrast to the alpha-subunit, the mature form of the beta-subunit of beta-hexosaminidase is derived from the precursor by internal proteolytic nicking rather than by removal of a large amino-terminal peptide segment.

The size, operation, and technical capabilities of protein and nucleic acid core facilities.

A survey of 40 protein and nucleic acid chemistry facilities has provided data about the capabilities of core facilities and the cost of the services they provide. Approximately 43% of the +158,000 average annual operating budget for a typical university facility is derived from service charges. After correcting for the various degrees of subsidization of the different facilities, it was found that it costs a typical university facility +65 to carry out an acid hydrolysis and amino acid analysis on a protein. A 25-residue peptide can be synthesized and cleaved for +2078, whereas sequencing the same peptide costs +874. A 25-residue oligonucleotide can be synthesized for +258. The total work output per month of an average facility corresponds to 65 amino acid analyses, 15 amino acid sequencing runs, three peptide syntheses, and 16 oligonucleotide syntheses. Depending on the approach used, from 85 to nearly 200 pmol of protein are required to obtain an accurate amino acid composition. To sequence the first 15 amino acids in a protein typically requires 150 pmol compared with 1.2 nmol of protein required to first carry out a tryptic digest and then isolate and sequence the first 15 residues in one of the resulting tryptic peptides.

Proteolytic processing of the alpha-chain of the lysosomal enzyme, beta-hexosaminidase, in normal human fibroblasts.

The two subunits of beta-hexosaminidase undergo many post-translational modifications characteristic of lysosomal proteins, including limited proteolysis. To identify proteolytic cleavage sites in the alpha-chain, we have biosynthetically radiolabeled the transient forms, isolated these by immunoprecipitation, gel electrophoresis, and electroelution, and subjected them to automated Edman degradation. The position of the NH2-terminal amino acid was inferred from the elution cycle of the radioactive amino acid and the primary sequence encoded in the alpha-chain cDNA. The amino terminus of the precursor obtained by in vitro translation of SP6 alpha-chain mRNA in the presence of microsomes was leucine 23. The same amino terminus was found in precursor alpha-chain synthesized by normal human fibroblasts (IMR90) in a 1- or 3-h pulse or secreted by these cells in the presence of NH4Cl. The alpha-chain isolated after a 3-h pulse followed by a 5-h chase (intermediate form) included a mixture of molecular species of which the amino terminus was arginine 87 (most abundant), histidine 88, or leucine 90. After a 20-h chase (mature form) the latter species predominated. This mature form of the alpha-chain remained fully reactive with antibody raised against the carboxyl-terminal 15 amino acids, indicating little if any proteolysis at the carboxyl terminus. Thus synthesis and maturation of the alpha-chain of beta-hexosaminidase includes two major proteolytic cleavages: the first, between alanine 22 and leucine 23, removes the signal peptide to generate the precursor form, whereas the second occurs between the dibasic amino acids, lysine 86 and arginine 87. The second cleavage is followed by trimming of 3 additional amino acids to give the mature form of the alpha-chain.

Induction of helper and suppressor T cells by nonoverlapping determinants on the large protein antigen, beta-galactosidase.

The fine specificity of the T cell repertoire directed against T helper (Th)-inducing and T suppressor (Ts)-inducing determinants was examined with cyanogen bromide and tryptic peptides of Escherichia coli beta-galactosidase (GZ), a large tetrameric protein (monomer molecular weight = 116 kDa). Immunization with cyanogen bromide fragment 2 [CB-2, amino acids (a.a.) 3-92] induced both specific Th and Ts cells. Study of the induction of these functionally opposite T cell subpopulations with tryptic peptides of CB-2 indicated that Th and Ts were activated by separate, nonoverlapping determinants. Th-inducing activity resided in a nonapeptide, T6 (a.a. 44-52), whereas T4 (a.a. 27-37) induced Ts cells. The presence of distinct helper and suppressor determinants suggests that the specificity repertoire in these T cell subpopulations may differ, perhaps owing to the expression of antigen-recognizing receptors that are coded by unique gene families. Alternatively, antigen presentation structures may be physicochemically quite different, and bind to distinct parts of the peptide antigen.

The use of limited proteolysis to probe interdomain and active site regions of beta-galactosidase (Escherichia coli).

Limited proteolysis by pancreatic elastase (EC 3.4.21.36) and chymotrypsin (EC 3.4.21.1) was used to study the domain structure and active site of beta-galactosidase (EC 3.2.1.23) (Escherichia coli). Treatment with elastase resulted in a rapid cleavage between residues Ala-732 and Ala-733. No inactivation accompanied this cleavage suggesting that this bond is in a hinge region of the protein. Some slow cleavages beyond the initial one were observed to occur and were accompanied by inactivation. Treatment of beta-galactosidase with chymotrypsin resulted in cleavages first between Trp-585 and Ser-586 and then between Phe-601 and Cys-602. The first of these cleavages resulted in total inactivation of beta-galactosidase. The presence of monovalent ions or isopropyl-beta-D-thiogalactopyranoside protected against the cleavages but when Mg2+ or Mn2+ was present in the reaction mixture, the bond between Trp-585 and Ser-586 was more susceptible to the action of chymotrypsin. These data demonstrate that the conformation of beta-galactosidase around Trp-585 and Ser-586 is dramatically affected by the binding of ions and isopropyl-beta-D-thiogalactopyranoside. The mutant M15 beta-galactosidase, which is missing residues 11 through 41 and is an inactive dimer rather than an active tetramer, was found to be much more labile to proteases than native beta-galactosidase, but the same initial cleavages were found to occur. In addition, trypsin cleaved the M15 protein between Arg-431 and Trp-432 while native beta-galactosidase was stable to trypsin.

cDNA cloning and amino acid sequence of bovine brain 2',3'-cyclic-nucleotide 3'-phosphodiesterase.

2',3'-Cyclic-nucleotide 3'-phosphodiesterase (EC 3.1.4.37) has been widely used as a marker for myelin-oligodendrocytes in the central nervous system. Evidence has been provided that the enzyme is identical with one of the Wolfgram proteins of central nervous system myelin. The amino acid sequence of bovine 2',3'-cyclic-nucleotide 3'-phosphodiesterase was determined by both protein and cDNA sequence analyses. Protein sequence analysis was done on bovine elastase 2',3'-cyclic-nucleotide 3'-phosphodiesterase, a low molecular weight enzyme obtained by solubilization with pancreatic elastase (EC 3.4.21.36) (Nishizawa, Y., Kurihara, T., and Takahashi, Y. (1980) Biochem. J. 191, 71-82; Kurihara, T., Nishizawa, Y., Takahashi, Y., and Odani, S. (1981) Biochem. J. 195, 153-157). Based on the carboxyl-terminal sequence of bovine elastase 2',3'-cyclic-nucleotide 3'-phosphodiesterase, synthetic oligodeoxyribonucleotides were prepared and used as probes for screening a cDNA library of bovine brain. A cDNA of 2305 base pairs was obtained and sequenced, and the complete amino acid sequence of bovine 2',3'-cyclic-nucleotide 3'-phosphodiesterase was deduced. Bovine 2',3'-cyclic-nucleotide 3'-phosphodiesterase deduced contains 400 amino acids including initiation methionine and has a molecular weight of 44,850. Bovine elastase 2',3'-cyclic-nucleotide 3'-phosphodiesterase corresponds to the 236 amino acids of bovine 2',3'-cyclic-nucleotide 3'-phosphodiesterase. RNA blot analysis revealed a single-species mRNA of about 2600 bases.

Repertoires of T cells directed against a large protein antigen, beta-galactosidase. II. Only certain T helper or T suppressor cells are relevant in particular regulatory interactions.

11 cyanogen bromide (CB) peptides, comprising 70% of the large protein, Escherichia coli beta-galactosidase (GZ), were studied for their ability to induce T suppressor (Ts) cells capable of strongly suppressing the in vitro anti-fluorescein (FITC) response to GZ-FITC. Only CB-2 (amino acid residues 3-92) and CB-3 (residues 93-187) were found to bear such Ts-inducing epitopes. In examining the specificity of T helper cell (Th) targets susceptible to CB-2 and CB-3-specific Ts, it appeared that only nearly Th targets could be suppressed. Thus, CB-10-primed Th were not suppressed by either Ts; even CB-3-primed Ts did not suppress CB-2-specific Th, although CB-2-specific Ts were effective. Furthermore, analysis of the suppression pattern revealed a hierarchical use of potential epitopes on native GZ in triggering functional regulatory T cells. A dominant Th epitope near the amino terminus of GZ tops a hierarchy of potential Th, most of which are never engaged. The dominant determinant seems to exist on the peptide CB-2-3 (residues 3-187), and presumably is destroyed by its cleavage at Met 92; the Th cells that it induces are suppressible by each of the Ts-inducing peptides. In the GZ system, where the native antigen is quite large, the interactions between Th and Ts are highly circumscribed. This may be attributable to the topology of antigen fragments produced during processing; any relevant fragment must bear at least a Ts- and Th-reactive determinant to permit intercellular regulation. A final implication of these results is that, not only does the existence of a Th-inducing determinant depend on its being an appropriate distance from a B cell epitope, but the existence of Ts-inducing determinants likewise depends on the existence of a neighboring Th-B cell association.

The amino acid sequence of thiogalactoside transacetylase of Escherichia coli.

The amino acid sequence of thiogalactoside transacetylase, a dimer, has been determined. The monomer contains 202 amino acid residues in a single polypeptide chain and has a molecular weight of 22,671. The analysis was carried out by treatment of the carboxymethylated protein with cyanogen bromide and with trypsin. All seven cyanogen bromide peptides were isolated in pure form and were ordered by peptides isolated from tryptic digests. The sequence analysis was aided by determination of the DNA sequence of the lacA gene. The amino terminus of the protein is heterogenous because the initiator methionine is only partially cleaved. Another rather unusual feature of this cytoplasmic protein is a very hydrophobic segment in the center portion of the chain. Comparison of the amino acid sequence of thiogalactoside transacetylase to those of the lac repressor, beta-galactosidase, and lactose permease did not reveal any marked similarities. Therefore, there is no obvious evolutionary relatedness among proteins of the Lactose Operon.

Purification of thiogalactoside transacetylase by affinity chromatography.

Thiogalactoside transacetylase, the product of the lacA gene of the lactose operon of Escherichia coli, has been purified by an improved procedure. The enzyme binds tightly to immobilized Cibacron Blue F3GA columns and can be eluted by potassium chloride in high concentrations. Final purification was obtained by affinity chromatography on an agarose-coenzyme A column followed by gel filtration.

Purification, structure, and properties of hybrid beta-galactosidase proteins.

Structural studies are reported on seven hybrid proteins produced by gene fusions that contain a "foreign" amino acid sequence substituting for part of the NH2-terminal region of the beta-galactosidase polypeptide. All of these hybrid proteins retain beta-galactosidase enzyme activity. A simple and rapid purification scheme for the hybrid beta-galactosidase is described, involving ammonium sulfate fractionation, DEAE-Bio-Gel, and Bio-Gel A-1.5 chromatography. The proteins are tetramers and have activity equivalent to that of wild type enzyme. Their amino acid sequences were determined by isolation and sequence determination of the cyanogen bromide peptide containing the joining site. The subunit sizes vary from 1009 to 1355 residues compared to 1023 for wild type. Up to 26 amino acid residues at the NH2 terminus of beta-galactosidase can be substituted by the new sequence. The nature of the new sequence apparently has no influence on stability or activity of the hybrid, but those hybrids with more of the beta-galactosidase sequence deleted are less stable to heat or urea treatment and tend to dissociate to dimeric form. All hybrids are less stable to heat and urea than wild type. Antipeptide antibodies raised against peptides derived from the NH2-terminal region of wild type beta-galactosidase were found to bind to the hybrid proteins, although they do not bind to the normal enzyme. These results indicate that the quaternary structure is disturbed but not disrupted by substitution of the different sequence, and these results help to localize one of the intersubunit contact regions in beta-galactosidase.

The active site regions of lacZ and ebg beta-galactosidases are homologous.

The active site-directed inhibitor 4-nitrophenyl-beta-D-galactopyranosylmethyltriazene, previously shown (Fowler, A. V., Zabin, I., Sinnott, M. L., and Smith, P. J. (1978) J. Biol. Chem. 253, 5283-5285) to alkylate methionine 502 in lacZ beta-galactosidase, was used to label the second naturally occurring beta-galactosidase of Escherichia coli (ebgo). The reagent was also used to label two mutant forms of the enzyme (ebga and ebgb) selected for enhanced lactase activity. In the case of ebgo and ebga, 75 and 85% of the label, respectively, was incorporated into a tryptic peptide which is homologous (38% identity) to residues 483-503 of the lacZ beta-galactosidase sequence. In the ebgo and ebga enzymes, a serine probably is alkylated. In the case of the ebgb enzyme, 61% of the label is found on a tryptic peptide homologous (69% identity) with residues 457-468 of the lacZ beta-galactosidase. In this peptide, a glutamic acid and a tyrosine residue are both alkylated.

Inactivation of beta-Galactosidase by iodination of tyrosine-253.

Beta-Galactosidase is rapidly inactivated by iodination catalyzed by lactoperoxidase but is not inactivated in the presence of the substrate analogue, isopropyl beta-D-thiogalactoside (IPTG). Enzyme activity is lost upon the incorporation of 1 mol of iodine per mol of monomer, without dissociation of the tetrameric structure. Tryptic digests of beta-galactosidase iodinated with 125I in the presence and absence of IPTG were separated by high-performance liquid chromatography and were compared. One fraction was found to be more highly labeled in the digest from the inactivated protein. After isolation of the peptide, amino acid analysis indicated it to be Asp-Tyr-Leu-Arg, residues 252-255. Thus, Tyr-253 is the most reactive tyrosine in beta-galactosidase. This suggests that the conformation of this region of the protein may be altered by binding of IPTG to make Tyr-253 less accessible to iodination. Alternatively, Tyr-253 could be an active-site residue.

Repertoires of T cells directed against a large protein antigen, beta-galactosidase. I. Helper cells have a more restricted specificity repertoire than proliferative cells.

The proliferative and helper T cell repertoires were compared in the CBA/J mouse for the response to the large protein antigen, tetrameric beta-galactosidase (GZ = 1021 a.a/monomer). The systems assessed the ability of cyanogen bromide (CB) peptides of GZ to: 1) prime for a T cell proliferative response to GZ; or 2) generate T cell help, measured by the production of anti-FITC PFC in the in vitro response to GZ-FITC. Priming for in vitro proliferation was attempted with 11 CB peptides comprising 70% of the GZ molecule. Strong priming was found with five peptides and intermediate priming was found with four other peptides; two peptides were without effect (CB-20 = a.a 767-862, and CB-4 = a.a. 188-202). Despite this indication of generally dispersed recognition of GZ epitopes, only two CB peptides, CB-2 (a.a. 3-92) and CB-10 (a.a. 378-418) were able to induce a T helper cell response. The surprising dearth of helper T cell-inducing epitopes may be peculiar to the limited fluorescein (FITC) substitution on GZ-FITC (17-25 FITC residues per tetramer) or it may reflect the constraints involved in T cell recognition required for T-B collaboration. Also considered was the possibility that the helper T cell repertoire might be distinct from the proliferative repertoire, the latter reflecting DNA synthesis and recruitment by other functional T cell subpopulations.

Sequence studies on the maltose-binding protein of Escherichia coli.

The amino terminal sequence of the maltose-binding protein as well as the leader region were reported earlier [2]. The amino acid composition of the protein indicated that the protein has no cysteine and very few methionine, arginine and histidine residues. In contrast, the residues of lysine, alanine and aspartic acid (or asparagine) account for about a third of the approximately 360 amino acid residues in the protein. Five cyanogen bromide peptides ranging in size from 10 to 154 residues have now been isolated by Sephadex G50 and G100 gel filtration. These five peptides account for the whole protein. Sequence determination of these fragments as well as of others has now been initiated.

Position of the lacZX90 mutation and hybridization between complete and incomplete beta-galactosidase.

The position of the termination codon in lacZX90 was determined by isolation of a lac+ revertant. Lysine was found to replace tyrosine at position 1,012 of beta-galactosidase, indicating that X90 protein lacked the carboxyl-terminal 10 residues. A heat- and urea-sensitive hybrid enzyme was formed in vivo when supC, which supplies tyrosine to the position in the polypeptide corresponding to the nonsense codon, was used to suppress lacZX90. This result shows that suppression that adds back the original amino acid may not lead to the production of the wild-type enzyme if the latter is multimeric, because incomplete chains can be incorporated into the oligomer.

beta-Galactosidase alpha-complementation. Overlapping sequences.

Enzyme activity is restored to two defective beta-galactosidase molecules (M15 protein lacking amino acid residues 11-41 and M112 protein lacking residues 23-31) by incubation with CNBr2 (residues 3-92 of beta-galactosidase). M15 and M112 proteins (alpha-acceptors) are dimers. Complemented enzyme, like wild type, has a tetrameric structure. Cleavage of CNBr2 with glutamic acid-specific protease yielded a much smaller alp ha-donor (3-41 peptide) which was also effective in complementation, indicating that the M15 protein can supply all of the residues from 42-92 for the structure of complemented enzyme. Treatment of M112 protein/3-41 peptide complemented enzyme with trypsin under very mild conditions followed by examination of the products demonstrated that the alpha-donor pep]tide supplies the NH2-terminal segment of complemented enzyme. Similar trypsin treatment of M15 protein/CNBr2 indicated that in this complemented enzyme the polypeptide region beyond those residues missing in the alpha-acceptor can be provided by either the alpha-donor or the alpha-acceptor. Both M15 protein and M112 protein are more susceptible to mild tryptic proteolysis than complemented enzyme, indicating a more open structure. Several antipeptide antibodies that react with these two proteins do not react with beta-galactosidase. M112 protein, like M15 protein, can be activated by anti-beta-galactosidase but to a much higher level.