Gratuitous overexpression of genes in Escherichia coli leads to growth inhibition and ribosome destruction

Abstract

We attempted to test the idea that the relative abundance of each individual tRNA isoacceptor in Escherichia coli can be altered by varying its cognate codon concentration. In order to change the overall codon composition of the messenger pool, we have expressed in E. coli lacZ with the aid of T7 RNA polymerase and, separately, we have expressed a truncated tufB gene (designated ⌬tufB) with the aid of E. coli RNA polymerase so that their respective gene products individually accounted for 30% of the total bacterial protein. Unexpectedly, the maximum expression of either test gene has no specific effect on the relative rates of synthesis of the tRNA species that we studied. Instead, we find that there is a cumulative breakdown of rRNAs, which results in a loss of ribosomes and protein synthetic capacity. After either of the test genes is maximally induced, there is a growing fraction of protein synthesis invested in ␤-galactosidase or ⌬tufB that is matched by a comparable decrease of the fraction of normal protein synthesis. We have also observed enhanced accumulation of two heat shock proteins during overexpression. Finally, after several hours of overexpression of either test protein, the bacteria are no longer viable. These results are relevant to the practical problems of obtaining high expression levels for cloned proteins.There is a good correlation between the frequencies of codon usage in the mRNA pools and the abundance of the corresponding tRNA species for bacteria growing in rich laboratory media (10). Furthermore, when the quality of the culture media is systematically varied, the tRNA abundance changes in expected ways. In particular, the relative fraction of major tRNA species that translate major codons increases as the growth rates of the bacteria increase (4, 6). This variation is thought to be parallel to an increased usage of major codons in the mRNA pool at the higher growth rates (3,5,14). The present experiments were initiated to study the influence of the codon frequencies in the mRNA pool on the synthesis of individual tRNA species by bacteria.We have followed the consequences for Escherichia coli of overexpressing to very high levels genes whose protein products have no function for the cells, so-called gratuitous proteins. In this way, we have manipulated the codon composition of the mRNA pool, and we have measured the influence of this manipulation on the abundance of a number of tRNA species. The two gratuitous proteins that we have used were chosen because they are encoded by mRNA species with very different codon frequencies. One such protein is ␤-galactosidase, which is metabolically useless under the growth conditions we studied and which has an mRNA with a predominantly minor codon composition (21). The other is a truncated version of elongation factor Tu, called ⌬EF-Tu, which is functionally inactive and is encoded primarily by major codons (21). The results of overexpressing these two proteins were similar and not at all the expected ones.Expression of either ␤-galactosidase or ...

Modification of the apoptotic-like effects of MBP protein overexpression in E. coli by fusion with 14-3-3 derived polypeptides

“…The mechanism of this effect is unclear. One possible explanation is that an unspecific`protein burden' causes starvation of the bacteria: it has been shown that overexpression of various proteins is accompanied by an uniform growth-rate inhibition (Dong et al, 1995). On the other hand, however, the growth inhibition appears to be stronger than expected (Kurland and Dong, 1996).…”

Synthetic gene design—The rationale for codon optimization and implications for molecular pharming in plants

“…Thus, they hypothesized that as amino acid charging of tRNA becomes limiting, only the tRNAs that are robust to starvation can support high translation levels of heterologous protein expression. Therefore successful codon bias of a gene depends on maintaining high levels of charged tRNAs and minimizing levels of uncharged tRNAs which can inhibit translation (Bonomo & Gill 2005;Dong et al 1995;Harcum 2002;Welch, Govindarajan, et al 2009). …”

Bacterial growth inhibition by overproduction of protein

“…After the initial attempts to make vast quantities of the targeted protein have failed, the experimenters may, in desperation, study their cultures. They then find that the fully induced bacteria are behaving wierdly: they grow slowly (Springer et al, 1985;Franklyn and Schimmel, 1990;Menguito et al, 1993;Tubulekas and Hughes, 1993;Vind et al, 1993;Bowrin et al, 1994), they produce enhanced quantities of heat-shock proteins (Bahl et al, 1987;Parsell and Sauer, 1989;Vind et al, 1993;Dong et al, 1995), and they may eventually stop growing altogether (Dong et al, 1995). Under the light microscope the induced bacteria may look like 'snakes' and they may accumulate inclusion bodies (Marston, 1986).…”