In this work, we identified an open reading frame 5' to the yeast HALI gene, that shares a 38% identity in the deduced amino acid sequence with gluconeogenic enzyme isocitrate lyase, encoded by ICL1. We therefore termed the new gene ICL2. The latter is not capable of complementing an icl1 deletion for growth on ethanol neither in its original context, nor when expressed under the control of the glycolytic PFK2 promoter. Nevertheless, fusions of the 5'-non-coding region of ICL2 to lacZ reporter gene revealed that the gene is transcribed and that the transcriptional regulation is similar to that of other gluconeogenic genes, i.e. high-level expression on ethanol that is drastically reduced on glucose media. Therefore, we attribute the lack of complementation to a lack of function of the encoded protein as an isocitrate lyase. The deduced amino acid sequences of Icl1 and Icl2 differ in a conserved motif used to identify isocitrate lyases, the hexapeptide KKCGHM, where the second lysine residue of Icl1 is replaced by an arginine in Icl2. However, we here demonstrated by in vitro mutagenesis of ICL1 that such an exchange, even though it affects Icl activity to some degree, does not lead to a complete lack of function. Thus, the results presented in this work argue for ICL2 encoding a non-functional isocitrate lyase and provide evidence that lysine 216 of Icl1 is not essential for catalysis.