In contrast to unamplified DNA fragments, amplified fragments contain exons of the human DHFR gene. DHFR mRNAs were mapped as the major 5'-end at position -71 of the human DHFR gene. However, determination of the amplification rate of the DHFR gene using a fluorescence-activated cell sorter for selection (Johnston et al., 1983) gave a rate of 1x10-3 amplification events/cell division.
Cell cycle experiments using a physical procedure to remove mitotic cells from culture dishes for synchronization showed that the amount of DHFR enzyme peaked in the early S phase of mouse cells (2-3 fold; Mariani et al., 1981). A brief summary of the results of the analysis of the human DHFR gene system is described below. When grown in the absence of M TX, all cell lines lost the enhanced DHFR enzyme activity.
However, the number of double minutes did not correlate with either the amount of DHFR enzyme or the instability of the resistant phenotype in the absence of selection. The second screening procedure involved phenotypic expression of the human DHFR eDNA to identify E.
DNA Sequence of the DHFR Coding Region
In vitro translation experiments showed that this RNA encodes the DHFR enzyme. These results showed that the increased DHFR enzyme content in MTX-resistant cell lines is due to an increase in the amount of DHFR-specific mRNA. Polysomal poly(A)-containing RNA from 6A3 cells was used to create an eDNA library, and two methods were used to select for DHFR-specific recombinant clones.
The first method was differential hybridization (St. John and Davis, 1979), using 32 P-labeled eDNA from 6A3 cells and VA2 B cells to screen replica filters of recombinant clones. Positive clones containing DHFR inserts were identified by a darker hybridization signal using the 6A3 eDNA probe compared to the VA2-B probe. Comparison of the DNA sequence of the 5' end of a phenotypically expressed eDNA clone (pHD84) with the sequence of a mouse DHFR eDNA clone confirmed the insert encoding DHFR.
Different DHFR cDNAs were used to probe RNA transfer hybridization experiments, using 6A3 polysomal RNA. These results indicated that at least three RNAs (3.8, 1.0, and 0.8 kb) exist for human DHFR and that these RNAs differed in the length of their 3' untranslated regions.
Chapter 5: Identification of a Human DHFR Pseudogene
Since this homologous DHFR region cannot encode the DHFR enzyme, the cosmid clone contains a DHFR intronless pseudogene (1jJHD 1). 1982) have identified two additional DHFR pseudogenes that were distinct from l(JHDl. Analysis of other gene systems has led to the hypothesis that an RNA intermediate is involved in the formation of intronless pseudogenes (Nishioka et al., 1980; Vanin et al. 1980; Karin and Richards, 1982. The DNA sequence homology between the DHFR cDNAs and l(JHDl) extends beyond the polyadenylation site of the 1.0 kb DHFR mRNA.
Therefore, if an RNA intermediate is involved in the generation of l(JHDl), then an RNA greater than 1.0 kb of DHFR mRNA must be involved. Subsequent analysis has located the human DHFR gene on human chromosome 5, using hybridization experiments of DNA transfer with DNA from mouse-human and Chinese hamster-human cell hybrids and 32P-labeled pHD84 probe (Maurer et al., 1984) This analysis also showed that the unamplified DHFR-specific EcoRI fragments described above, were not shared with the functional DHFR gene.
Isolation of the Human DHFR Gene
However, the corresponding introns of the human, mouse and Chinese hamster DHFR genes show wide variation in their size, except for the first intron. In contrast to the variation in intron size, the total lengths of the DHFR genes are equivalent among the three mammalian species. There is a relatively high G+C content in the 5' ends of the human gene (average 70%), while there is a relatively low G+C content in the rest of the gene (average 40%).
There is no detectable sequence homology in the intron sequences of the human and mouse DHFR genes other than the 15-35 nucleotides immediately adjacent to the intron/exon splice junctions. The 5' noncoding regions of the mouse and human genes also share significant nucleotide sequence identity (65%) up to at least 327 bp upstream of the translation initiation codon. This sequence homology suggests that this region plays an important regulatory role in DHFR expression.
In contrast to the 5'-end of DHFR genes from mouse and human cells, the 3'-untranslated regions show no obvious sequence homology. Since the 3'-untranslated sequence of the mouse gene is implicated in the control of growth-dependent expression of DHFR (discussed above), it would be interesting to determine whether a similar mechanism exists to control the expression of the corresponding human gene.
Chapter 7: Mapping of the Major and Minor 5'-ends of DHFR-Specific RNA
Almost all cells treated with the above concentrations of MTX died in the first 2 weeks. 34; The subscripts in the symbols of the variants refer to the time when the analysis of DHFR activity was performed. An analysis of the chromosome composition of the MTX-resistant variants again revealed a marked variability.
The most striking feature of the chromosome composition of the MTX-resistant variants was the presence of double minute chromosomes. These cells were then tested for the behavior of the DHFR activity and the number of double minute chromosomes after further growth in the absence of MTX. Quantitative behavior of the DHFR activity and the number of double-minute chromosomes per cell in variant 6A3 after removing the selective pressure.
The content of the latter in the methotrexate-resistant cell lines mentioned above is increased up to several hundred times. Fractionation of the poly(A)-containing RXA by electrophoresis through a 1.2% (w/v) agarose/CH3HgOH plate gel was performed as described by Bailey & Davidson (1976), except that the concentration of CH3HgOH in the gel was 20 mM and the electrophoresis buffer was 50 mM sodium borate (pH 8.2).
CHAPTER 4
Nonhomologous amino acids in the enzyme sequence from rat L120 cells (Stone et al., 1979). As noted in the cloning of the 6 kb EcoRI fragment of the DHFR gene in A. This scheme best illustrates the extensive human and mouse 5' non-coding nucleotide homology.
However, the 45-48 bp quadruple repeat of the mouse gene (Crouse et al., 1982) is present in only one copy in the human sequence. The only exception is the -365 bp nucleotide sequence homology of the 5' non-coding region in the human and rat insulin genes (Bell et al., 1982). The portions of the mouse DHFR gene (M) shown (coding regions and adjacent intron segments) are numbered according to Crouse et al., (1982) and Setzer et al., (1982).
Best fit plot of 5' non-coding regions of human and mouse DHFR genes. The positions of the first coding regions of the human and mouse genes are indicated. The sizes of the coding sequences in each exon (in bp) and the sizes of the human gene introns (in kb) are indicated.
The sizes of the Chinese hamster genintrons are those reported by Carothers et al., (1983). Alignment of the major and minor 5' ends of DHFR-specific RNAs with the sequence upstream of the human DHFR reading frame.
