Identification of critical amino acid residues on human dihydrofolate reductase protein that mediate RNA recognition

Abstract

Previous studies have shown that human dihydrofolate reductase (DHFR) acts as an RNA-binding protein, in which it binds to its own mRNA and, in so doing, results in translational repression. In this study, we used RNA gel mobility shift and nitrocellulose filter-binding assays to further investigate the specificity of the interaction between human DHFR protein and human DHFR mRNA. Site-directed mutagenesis was used to identify the critical amino acid residues on DHFR protein required for RNA recognition. Human His-Tag DHFR protein specifically binds to human DHFR mRNA, while unrelated proteins including thymidylate synthase, p53 and glutathione-S-transferase were unable to form a ribonucleoprotein complex with DHFR mRNA. The Cys6 residue is essential for RNA recognition, as mutation at this amino acid with either an alanine (C6A) or serine (C6S) residue almost completely abrogated RNA-binding activity. Neither one of the cysteine mutant proteins was able to repress the in vitro translation of human DHFR mRNA. Mutations at amino acids Ile7, Arg28 and Phe34, significantly reduced RNA-binding activity. An RNA footprinting analysis identified three different RNA sequences, bound to DHFR protein, ranging in size from 16 to 45 nt, while a UV cross-linking analysis isolated an approximately 16 nt RNA sequence bound to DHFR. These studies begin to identify the critical amino acid residues on human DHFR that mediate RNA binding either through forming direct contact points with RNA or through maintaining the protein in an optimal structure that allows for the critical RNA-binding domain to be accessible.

Figure 1

Map of the human DHFR cDNA. The 700-nt DHFR cDNA includes 59 nt of the 5′-UTR, the coding region (564 nt) and 77 nt of the 3′-UTR.

Figure 2

(A) Specificity of human DHFR protein binding to human DHFR mRNA. ³²P-radiolabeled human DHFR RNA (100 000 c.p.m.; 3.8 fmol) was incubated in the absence (lane 1) or presence of 21.3 pmol (lane 2), 42.6 pmol (lane 3), 85.2 pmol (lane 4) of His-Tag recombinant human DHFR protein as described in the Materials and Methods. Radiolabeled DHFR mRNA was also incubated in the presence of 238 pmol of His-Tag E.coli DHFR (lane 5), 139 pmol of His-Tag human TS (lane 6), 192 pmol of GST (lane 7) and 63.3 pmol of GST-p53 (lane 8). (B) RNA gel-shift competition experiment to show specificity of binding of DHFR protein to DHFR mRNA. A radiolabeled DHFR mRNA probe was incubated in the absence (lane 1) or presence (lanes 2–10) of His-Tag DHFR protein (42.6 pmol) as described in the Materials and Methods. Competition studies were performed with a 25- (lane 3), 125- (lane 4) and 250-fold (lane 5) molar excess of unlabeled DHFR mRNA and a 250-fold molar excess of yeast tRNA (lane 6), GAPDH mRNA (lane 7), 18S rRNA (lane 8), TS30 (lane 9) or β-actin antisense RNA (lane 10).

Figure 2

(A) Specificity of human DHFR protein binding to human DHFR mRNA. ³²P-radiolabeled human DHFR RNA (100 000 c.p.m.; 3.8 fmol) was incubated in the absence (lane 1) or presence of 21.3 pmol (lane 2), 42.6 pmol (lane 3), 85.2 pmol (lane 4) of His-Tag recombinant human DHFR protein as described in the Materials and Methods. Radiolabeled DHFR mRNA was also incubated in the presence of 238 pmol of His-Tag E.coli DHFR (lane 5), 139 pmol of His-Tag human TS (lane 6), 192 pmol of GST (lane 7) and 63.3 pmol of GST-p53 (lane 8). (B) RNA gel-shift competition experiment to show specificity of binding of DHFR protein to DHFR mRNA. A radiolabeled DHFR mRNA probe was incubated in the absence (lane 1) or presence (lanes 2–10) of His-Tag DHFR protein (42.6 pmol) as described in the Materials and Methods. Competition studies were performed with a 25- (lane 3), 125- (lane 4) and 250-fold (lane 5) molar excess of unlabeled DHFR mRNA and a 250-fold molar excess of yeast tRNA (lane 6), GAPDH mRNA (lane 7), 18S rRNA (lane 8), TS30 (lane 9) or β-actin antisense RNA (lane 10).

Figure 3

(A) RNA-binding activity of wild-type and mutant DHFR proteins using the nitrocellulose filter-binding assay. ³²P-radiolabeled DHFR mRNA (100 000 c.p.m.; 3.8 fmol) was incubated with 42.6 pmol of wild-type or mutant His-Tag recombinant DHFR proteins using the nitrocellulose filter-binding assay as described in the Materials and Methods. Each value represents the mean ± SD of at least three experiments. (B) Enzyme activity of wild-type and mutant human DHFR proteins. The enzyme activity of wild-type and mutant DHFR proteins was determined using the spectrophotometric assay, as outlined in the Materials and Methods. Each value represents the mean ± SD of at least three experiments.

Figure 3

(A) RNA-binding activity of wild-type and mutant DHFR proteins using the nitrocellulose filter-binding assay. ³²P-radiolabeled DHFR mRNA (100 000 c.p.m.; 3.8 fmol) was incubated with 42.6 pmol of wild-type or mutant His-Tag recombinant DHFR proteins using the nitrocellulose filter-binding assay as described in the Materials and Methods. Each value represents the mean ± SD of at least three experiments. (B) Enzyme activity of wild-type and mutant human DHFR proteins. The enzyme activity of wild-type and mutant DHFR proteins was determined using the spectrophotometric assay, as outlined in the Materials and Methods. Each value represents the mean ± SD of at least three experiments.

Figure 4

Effect of wild-type and mutant human DHFR proteins on translation of human DHFR mRNA. Human DHFR mRNA (0.24 nmol) was incubated in the absence (lane 1) or presence of wild-type or mutant DHFR proteins (85.2 pmol) as described in the Materials and Methods. Lane 1 contains only DHFR mRNA; lane 2, wild-type human DHFR protein; lane 3, C6A; lane 4, C6S; lane 5, I7A; lane 6, R28A; lane 7, F34S; lane 8, E30Q; lane 9, F31A; lane 10, I7F; lane 11, His-Tag E.coli DHFR protein.

Figure 5

UV cross-linking analysis. ³²P-radiolabeled human DHFR mRNA (100 000 c.p.m.; 3.8 fmol) was incubated with 42.6 pmol of wild-type or mutant human His-Tag DHFR protein as described in the Materials and Methods. After UV cross-linking, the reaction mixture was incubated with RNase A to digest the unprotected RNAs. The UV cross-linked complexes were then resolved on SDS–12.5% PAGE. Lane 1, probe only; lane 2, His-Tag DHFR protein without UV cross-linking; lane 3, His-Tag DHFR protein; lane 4, C6A; lane 5, E30A; lane 6, GST.

Figure 6

Isolation of bound RNAs from DHFR RNP complexes. ³²P-radiolabeled human DHFR mRNA (3 000 000 c.p.m.; 114 fmol) was incubated with wild-type, His-Tag human DHFR protein (127.8 pmol), followed by digestion with RNase T1 and RNase A, and the addition of heparin. RNAs bound to the DHFR protein were eluted and resolved on a 15% polyacrylamide–8 M urea gel as described in the Materials and Methods. Lane 1, RNA markers; lane 2, RNAs isolated from the DHFR RNP complex.