Recombinant sickle hemoglobin containing a lysine substitution at Asp-85(alpha): expression in yeast, functional properties, and participation in gel formation.

Clinical modalities based on inhibition of gelation of HbS are hindered by the lack of quantitative information on the extent of participation of different amino acid residues in the aggregation process. One such site is Asp-85(alpha), which is involved in a parallel interdouble strand ionic interaction with Lys-144(beta) according to the crystal structure of HbS, but electron microscopy does not specifically show Asp-85(alpha) as a contact site for fiber formation. Using a yeast recombinant system, we have substituted this site by Lys to abolish ion pairing and to make a quantitative determination of its participation in aggregation. The purified double mutant was shown to have the expected pI, the calculated molecular weight, correct amino acid composition, and peptide map. The recombinant double mutant has an oxygen affinity of 10 mm Hg, which is identical to that for HbA and HbS under the same conditions; it also has high cooperativity with an average n value of 2.7. The change in P50 in response to chloride ions was about 25% less than that for HbA or HbS and is ascribed to the introduction of a new positive charge near one of the major oxygen-linked chloride binding sites of hemoglobin. The gelation concentration of the double mutant was measured by a new procedure (Bookchin et al, 1994); the maximal amount of soluble hemoglobin (Csat) in the presence of dextran indicated a decreased tendency for gelation with a Csat of 53 mg/mL compared with 34 mg/mL for HbS. This inhibitory effect is smaller than that of the E6V(beta)/L88A(beta) (Csat, 67 mg/mL) and the E6V(beta)/K95I(beta) (Csat, 90 mg/mL) recombinant hemoglobins. Thus, we would classify Asp-85(alpha) as a moderate contributor to the strength of the HbS aggregate. This wide range of gelation values demonstrates that some sites are more important than others in promoting HbS aggregation.

Inhibition of the G1-S transition of the cell cycle by inhibitors of deoxyhypusine hydroxylation.

The formation of the unusual amino-acid hypusine in eIF-5A (eukaryotic initiation factor 5A) is associated with cellular proliferation. We used a panel of compounds, including mimosine, to probe the relationship between the exit from the G1 phase of the cell cycle, i.e., the onset of DNA replication, and the formation of hypusine by the enzyme deoxyhypusyl hydroxylase (DOHH). These two parameters displayed the same dose dependency and structure-activity relationship. Only compounds that inhibited DOHH also suppressed proliferation. This effect was observed: (i) in spontaneously proliferating, virally transformed, and mitogen-stimulated cells; (ii) for both anchorage-dependent and anchorage-independent proliferation; and (iii) with normal and malignant cell lines. DOHH reactivation occurred rapidly after inhibitor withdrawal and correlated with synchronized entry into S. The changes in the expression of specific genes during the G1-to-S transition mimicked the physiological pattern. These findings suggest that hypusine formation in eIF-5A which occurs in a specific, invariant sequence motif acquired early in evolution, may be involved in the G1-to-S transition in the eukaryotic cells tested.

Random chemical modification of the oxygen-linked chloride-binding sites of hemoglobin: those in the central dyad axis may influence the transition between deoxy- and oxy-hemoglobin.

The features of random chemical modification are defined with reference to acetylation of bovine hemoglobin, which has been performed in a random manner so that all of the amino groups that participate in functional chloride binding (i.e., those that are oxygen-linked) could be identified. Random chemical modification, which has objectives different from those of both specific (selective) and extensive chemical modification, has been achieved for bovine hemoglobin with the mild reagent, 14C-methyl acetate phosphate; retention of function was demonstrated by a Hill coefficient of n = 2.2 for the modified hemoglobin. After removal of unmodified Hb chains, the mixture of randomly modified acetylated alpha or beta chains was subjected to tandem treatment with trypsin and chymotrypsin. Peptides were purified by HPLC and identified by amino acid analysis. The amount of radioactivity in the acetylated amino group of a purified peptide was taken as an estimate of the degree of chloride binding. For bovine Hb, two amino groups of the alpha-chain (Val-1 and Lys-99) and three amino groups of the beta-chain (Met-1, Lys-81, and Lys-103) were shown to be oxygen-linked (i.e., to have incorporated significantly more radioactivity in the deoxy conformation compared to the same site in the oxy conformation). Three of these sites were already known chloride-binding sites [i.e., Val-1(alpha), the N-terminus of the alpha-chain, and two sites between the 2 beta-chains of bovine hemoglobin, Met-1(beta) and Lys-81(beta); these findings support the conclusions of the random modification approach. Two other chloride-binding sites, Lys-99(alpha) and Lys-103(beta), align the sides of the central dyad axis connecting the two well-known major chloride-binding sites of bovine Hb. The interrelationship of these five chloride-binding sites was assessed by improved molecular graphics. When viewed through the central dyad axis, the functional chloride-binding sites in the central cavity appear to be symmetrically related and to connect the two major chloride-binding sites. Modifiers or mutants that are directed at these regions in the central dyad axis may favor the deoxy conformation to provide a lower oxygen affinity by preventing the constriction of the central cavity that normally occurs upon oxygenation.

Deficiency of acylpeptide hydrolase in small-cell lung carcinoma cell lines.

During protein biosynthesis, processing of the N terminus of many proteins may occur through acetylation and deacetylation. The enzyme acylpeptide hydrolase is likely involved in deacetylation of nascent peptide chains or of bioactive peptides. The related enzyme, acylase, hydrolyzes the acetyl amino acid product of the acylpeptide hydrolase reaction to acetate and a free amino acid. There is a reciprocal relationship between the substrates for these enzymes (i.e., substrates for one enzyme are competitive inhibitors for the other). In several cultured cell lines, including normal and malignant cells, the ratio of acylpeptide hydrolase to acylase enzyme activities appears to be coordinated and characteristic for a given cell type. Thus, in normal cultured lung cells, hamster ovary cells, hepatoma cells, and lymphocyte cells, nearly equal amounts of these enzymes are expressed, conducive to optimal processing of acetylated N-terminal residues. Four lines of erythroleukemic cell lines were found to express nearly twice as much acylase as acylpeptide hydrolase activity. In the Ehrlich ascites tumor cell line, where 80% of the proteins have been reported to remain acetylated at their N terminus, acylpeptide hydrolase is hardly expressed but acylase activity is not reduced. The 3p21 region of human chromosome 3, which contains the DNF15S2 locus that encodes acylpeptide hydrolase (Jones et al., Proc Natl Acad Sci USA 1991;88:2194), undergoes deletion in some carcinoma cells; the gene that encodes for the acylase is also present on region 3p of the same chromosome. We found that both acylpeptide hydrolase and acylase activities are practically absent in six small-cell lung carcinoma cell lines tested.(ABSTRACT TRUNCATED AT 250 WORDS)

Genetic relationship between acylpeptide hydrolase and acylase, two hydrolytic enzymes with similar binding but different catalytic specificities.

An 87% identity has been found between the reported cDNA sequence that encodes acylpeptide hydrolase (EC 3.4.19.1) [Mitta, M., Asada, K., Uchimura, Y., Kimizuka, F., Kato, I., Sakiyama, F. & Tsunasawa, S. (1989) J. Biochem. 106, 548-551] and a cDNA transcribed from a locus (DNF15S2) on the short arm of human chromosome 3, reported by Naylor et al. [Naylor, S.L., Marshall, A., Hensel, C., Martinez, P.F., Holley, B. & Sakaguchi, A.Y. (1989) Genomics 4, 355-361]; the DNF15S2 locus suffers deletions in small cell lung carcinoma associated with a reduction or loss of acylase activity (EC 3.5.1.14). Acylpeptide hydrolase catalyzes the hydrolysis of the terminal acetylated amino acid preferentially from small acetylated peptides. The acetylamino acid formed by acylpeptide hydrolase is further processed to acetate and a free amino acid by an acylase. The substrates for the acylpeptide hydrolase and the acylase behave in a reciprocal manner since acylpeptide hydrolase binds but does not process acetylamino acids and the acylase binds acetylpeptides but does not hydrolyze them; however, the two enzymes share the same specificity for the acyl group. These findings indicate some common functional features in the protein structures of these two enzymes. Since the gene coding for acylpeptide hydrolase is within the same region of human chromosome 3 (3p21) that codes for the acylase and deletions at this locus are also associated with a decrease in acylase activity, there is a close genetic relationship between the two enzymes. There could also be a relationship between the expression of these two enzymes and acetylated peptide growth factors in some carcinomas.