Cryptic polyadenylation sites within the coding sequence of three yeast genes expressed in tobacco

Effects of codon optimization on the mRNA levels of heterologous genes in filamentous fungi

Filamentous fungi, particularly Aspergillus species, have recently attracted attention as host organisms for recombinant protein production. Because the secretory yields of heterologous proteins are generally low compared with those of homologous proteins or proteins from closely related fungal species, several strategies to produce substantial amounts of recombinant proteins have been conducted. Codon optimization is a powerful tool for improving the production levels of heterologous proteins. Although codon optimization is generally believed to improve the translation efficiency of heterologous genes without affecting their mRNA levels, several studies have indicated that codon optimization causes an increase in the steady-state mRNA levels of heterologous genes in filamentous fungi. However, the mechanism that determines the low mRNA levels when native heterologous genes are expressed was poorly understood. We recently showed that the transcripts of heterologous genes are polyadenylated prematurely within the coding region and that the heterologous gene transcripts can be stabilized significantly by codon optimization, which is probably attributable to the prevention of premature polyadenylation in Aspergillus oryzae. In this review, we describe the detailed mechanism of premature polyadenylation and the rapid degradation of mRNA transcripts derived from heterologous genes in filamentous fungi.

Spurious polyadenylation of Norovirus Narita 104 capsid protein mRNA in transgenic plants

Noroviruses are members of the family Caliciviridae, and cause a highly communicable gastroenteritis in humans. We explored the potential to develop a plant-based vaccine against Narita 104 virus, a Genogroup II Norovirus. In stably transgenic potato, we obtained very poor expression of Narita 104 virus capsid protein (NaVCP) despite the use of a strong constitutive promoter (dual enhancer 35S) driving the native coding sequence. We identified potentially detrimental sequence motifs that could mediate aberrant mRNA processing via spurious polyadenylation signals. Northern blots and RT-PCR analysis of total RNA revealed truncated transcripts that suggested premature polyadenylation. Site-directed mutagenesis to remove one potential polyadenylation near-upstream element resulted in an increased expression of NaVCP when transiently expressed in leaves of Nicotiana benthamiana. Further, cloning of the truncated cDNAs from transgenic NaVCP potato plants and transiently transfected N. benthamiana allowed us to identify at least ten different truncated transcripts resulting from premature polyadenylation of full length NaVCP transcripts. Comparative studies using real time PCR analysis from cDNA samples revealed lower accumulation of full length transcripts of NaVCP as compared to those from a gene encoding Norwalk Virus capsid protein (a related Genogroup I Norovirus) in transiently transfected plants. These findings provide evidence for impaired expression of NaVCP in transgenic plants mediated by spurious polyadenylation signals, and demonstrate the need to scrupulously search for potential polyadenylation signals in order to improve transgene expression in plants.

Yeast RNase III triggers polyadenylation-independent transcription termination

Transcription termination of messenger RNA (mRNA) is normally achieved by polyadenylation followed by Rat1p-dependent 5'-3' exoribonuleolytic degradation of the downstream transcript. Here we show that the yeast ortholog of the dsRNA-specific ribonuclease III (Rnt1p) may trigger Rat1p-dependent termination of RNA transcripts that fail to terminate near polyadenylation signals. Rnt1p cleavage sites were found downstream of several genes, and the deletion of RNT1 resulted in transcription readthrough. Inactivation of Rat1p impaired Rnt1p-dependent termination and resulted in the accumulation of 3' end cleavage products. These results support a model for transcription termination in which cotranscriptional cleavage by Rnt1p provides access for exoribonucleases in the absence of polyadenylation signals.

Cryptic polyadenylation of transcripts of an RNA virus gene introduced into tobacco plants

We constructed an expression vector for the coat protein (CP) gene and the 3' untranslated region (3' UTR) of RNA virus (sweet potato feathery mottle virus severe strain (SPFMV-S)) lacking a foreign terminator. Out of seven transgenic tobacco plants, expression of the transgene was observed in six plants. RT-PCR analysis revealed that the transcripts had a poly(A) tail, and in most of them, polyadenylation occurred on the 5' side of the 3' UTR. These results suggest that the viral sequence contains a cryptic polyadenylation signal that permits 3'-end processing of the transcripts.

Improved transcriptional activators and their use in mis-expression traps in Arabidopsis

The synthetic transcription factor LhG4 has been used in numerous mis-expression studies in plants. We show that the sequence encoding the LhG4 activation domain, derived from Saccharomyces cerevisiae GAL4, contains several cryptic polyadenylation signals in Arabidopsis. The GAL4-derived sequence was modified according to preferred Arabidopsis codon usage, generating LhG4AtO which was faithfully transcribed in Arabidopsis plants. In protoplasts, LhG4AtO achieved maximum transactivation of the pOp promoter with 10-fold less input DNA than LhG4. The same methods were used to compare 10 other LhG4 derivatives that carried alternative natural or synthetic activation domains. Lh214 and Lh314, which contain synthetic activation domains comprising trimers of a core acidic activation domain, directed threefold more GUS expression from the pOp promoter with 20-fold less input DNA than LhG4. In contrast, when expressed from the CaMV 35S promoter in transgenic plants carrying a pOp-GUS reporter, Lh214 and Lh314 yielded transformants with substantially lower GUS activities than other constructs including LhG4AtO and LhG4 which performed similarly. When incorporated into an enhancer-trapping vector, however, LhG4AtO and Lh314 yielded enhancer traps with approximately twice the frequency of LhG4, suggesting that the modified activation domains offer improved performance when expressed from weaker transcription signals. To increase the number of LhG4 patterns available for mis-expression studies, we describe a population of enhancer-trap lines obtained with LhG4AtO in a pOp-GUS background. We show that enhancer-trap lines can transactivate an unlinked pOp-green fluorescent protein (pOp-GFP) reporter in the pattern predicted by staining for GUS activity.

The new RENT family of repetitive elements in Nicotiana species harbors gene regulatory elements related to the tCUP cryptic promoter

The tCUP cryptic constitutive promoter was discovered in the tobacco genome by T-DNA (transfer DNA) tagging with a promoterless GUS-nos gene. Here, we show that the portion of the tCUP sequence containing a variety of cryptic gene regulatory elements is related to a new family of moderately repetitive sequences (10(2) copies), the RENT (repetitive element from Nicotiana tabacum) family. The RENT family is found only in certain Nicotiana species. Five RENT elements were cloned and sequenced. The RENT elements are a minimum of 5 kb in length and share 80-90% sequence similarity throughout their length. The 5' termini are the same in the isolated RENT family members and are characterized by a conserved border sequence (TGTTGA(T or C)ACCCAATTTT(T or C)). The 3' ends of RENT sequence similarity vary in location and sequence. The tCUP cryptic promoter originated from a unique truncated RENT element that interrupts a phytochelatin synthase-like gene that may have undergone rearrangements prior to or resulting from T-DNA insertion. No evidence was found for expressed coding regions within the RENT elements; however, like the cryptic gene regulatory elements within the tCUP sequence, the isolated RENT elements possess promoter activity and translational enhancer activity.

The bifunctional LKR/SDH locus of plants also encodes a highly active monofunctional lysine-ketoglutarate reductase using a polyadenylation signal located within an intron

Both plants and animals catabolize lysine (Lys) via two consecutive enzymes, Lys-ketoglutarate reductase (LKR) and saccharopine dehydrogenase (SDH), which are linked on a single polypeptide encoded by a single LKR/SDH gene. We have previously shown that the Arabidopsis LKR/SDH gene also encodes a monofunctional SDH that is transcribed from an internal promoter. In the present report, we have identified two cDNAs derived from cotton (Gossypium hirsutum) boll abscission zone that encode a novel enzymatic form of Lys catabolism, i.e. a catabolic monofunctional LKR. The monofunctional LKR mRNA is also encoded by the LKR/SDH gene, using two weak polyadenylation sites located within an intron. In situ mRNA hybridization and quantitative reverse transcriptase-polymerase chain reaction analyses also suggest that the cotton monofunctional LKR is relatively abundantly expressed in parenchyma cells of the abscission zone. DNA sequence analysis of the LKR/SDH genes of Arabidopsis, maize (Zea mays), and tomato (Lycopersicon esculentum) suggests that these genes can also encode a monofunctional LKR mRNA by a similar mechanism. To test whether the LKR/SDH and monofunctional LKR enzymes possess different biochemical properties, we used recombinant Arabidopsis LKR/SDH and monofunctional LKR enzymes expressed in yeast (Saccharomyces cerevisiae) cells. The K(m) of the monofunctional LKR to Lys was nearly 10-fold lower than its counterpart that is linked to SDH. Taken together, our results suggest that the LKR/SDH locus of plants is a super-composite locus that can encode three related but distinct enzymes of Lys catabolism. These three enzymes apparently operate in concert to finely regulate Lys catabolism during plant development.