Communicated by P.BorstTo determine how microbody enzymes enter microbodies, we are studying the genes for cytosolic and glycosomal (microbody) isoenzymes in Trypanosoma brucei. We have found three genes (A, B and C) coding for phosphoglycerate kinase (PGK) in a tandem array in T. brucei. Gene B codes for the cytosolic and gene C for the glycosomal isoenzyme. Genes B and C are 95% homologous, and the predicted protein sequences share -45% amino acid homology with other eukaryote PGKs. The microbody isoenzyme differs from the cytosolic form and other PGKs in two respects: a high positive charge and a carboxy-terminal extension of 20 amino acids. Our results show that few alterations are required to redirect a protein from cytosol to microbody. From a comparison of our results with the unpublished data for three other glycosomal glycolytic enzymes we infer that the high positive charge represents the major topogenic signal for uptake of proteins into glycosomes.

The yeast mitochondrial genes coding for cytochrome c oxidase subunit I (COX1) and the ATPase subunits 8 and 6 are organized in one transcription unit. Precise mapping of RNA termini with S1 nuclease and primer extension analysis shows that the 3′ end of the COX1 mRNA and the 5′ end of the ATPase precursor RNA are juxtaposed within a conserved dodecamer sequence (5′‐ AAUAAUAUUCUU ‐3′). Sequence comparison reveals that this motif is present downstream of nearly all protein‐encoding genes, including extragenic unassigned reading frames (URFs) and two URFs located within introns. Also the 3′ terminus of an RNA species derived from the URF ‐containing intron of the large rRNA gene maps within such a dodecamer sequence. It is likely, therefore, that this motif serves as a processing point in the generation of mature mRNA. From a comparison of the various transcription units, we infer that RNAs that originate from an endonucleolytic cleavage at this sequence have stable 3′ termini, while further processing of the 5′ ends occurs. The efficiency of the initial cleavage varies between the different positions at which the motif is present.

Trypanosoma brucei contains two isoenzymes for glyceraldehyde‐phosphate dehydrogenase (GAPDH); one enzyme resides in a microbody‐like organelle, the glycosome, the other one is found in the cytosol. We show here that the glycosomal enzyme is encoded by two tandemly linked genes of identical sequence. These genes code for a protein of 358 amino acids, with a mol. wt of 38.9 kd. This is considerably larger than all other GAPDH proteins studied so far, including the enzyme that is located in the cytosol of the trypanosome. The glycosomal enzyme shows 52‐57% homology with known sequences of GAPDH proteins from 10 other organisms, both prokaryotes and eukaryotes. The residues that are involved in NAD+ binding, catalysis and subunit contacts are well conserved between all these GAPDH molecules, including the trypanosomal one. However, the glycosomal protein of T. brucei has some distinct features. Firstly, it contains a number of insertions, 1‐8 amino acids long, which are responsible for the high mol. wt of the protein. Secondly, an unusually high number of positively charged amino acids confer a high isoelectric point (pI 9.3) to the protein. Part of the additional basic residues are present in the insertions. We discuss the genomic organization of the genes for the glycosomal GAPDH and the possibility that the particular features of the protein are involved in its transfer from the cytoplasm, where it is synthesized, into the glycosome.

To determine how microbody enzymes enter microbodies, we are studying the genes for glycosomal (microbody) enzymes in Trypanosoma brucei. Here we present our results for triosephosphate isomerase (TIM), which is found exclusively in the glycosome. We found a single TIM gene without introns, having one major polyadenylated transcript of 1500 nucleotides with a long untranslated tail of approximately 600 nucleotides. By a novel method, suitable for low abundance transcripts, we demonstrate that TIM mRNA contains the 35‐nucleotide leader sequence (mini‐exon) also found on several other trypanosome mRNAs. The TIM gene and a DNA segment of at least 6 kbp upstream of the gene are transcribed at an equal rate in isolated nuclei, suggesting that the gene is part of a much larger transcription unit. The predicted protein is of the same size as TIMs from other organisms and shares approximately 50% amino acid homology with other eukaryote TIMs, somewhat less with prokaryote TIMs. Trypanosome TIM is the most basic of all TIMs sequenced thus far. This is, in part, due to the presence of two clusters of positively charged residues in the molecule which may act as a signal for entry into glycosomes.

We have examined the initiation of transcription of the mitochondrial genes for ribosomal RNA (rRNA) in the yeast Kluyveromyces lactis and show that these are transcribed independently from individual promoters. The mature large rRNA contains a 5' di- or triphosphate end which can be labelled in vitro with [alpha-32P]GTP using guanylyltransferase and this enabled us to determine the nucleotide sequence of its 5' terminus. For the small rRNA, a minor in vitro capped RNA species hybridizes in the region where--as judged from S1 nuclease protection experiments--the precursor of this RNA starts. We have determined the DNA sequence around the beginning of both rRNA genes and this reveals the existence of an identical nonanucleotide sequence (5' -ATATAAGTA- 3') just preceding the positions where the rRNAs start. This sequence is identical to the one preceding the rRNA genes in the mtDNA of the distantly related yeast Saccharomyces cerevisiae (Osinga, K.A. and Tabak, H.F. (1982) Nucl.Acids Res. 10, 3617-3626) and supports our proposal that this sequence motif is part of a yeast mitochondrial promoter. We have noticed that the same sequence is located in the putative origin of replication present in hypersuppressive petite mutants of S. cerevisiae and consider the possibility that this sequence is involved in RNA priming of DNA replication.