The determination of a list of differentially expressed genes is a basic objective in many cDNA microarray experiments. We present a statistical approach that allows direct control over the percentage of false positives in such a list and, under certain reasonable assumptions, improves on existing methods with respect to the percentage of false negatives. The method accommodates a wide variety of experimental designs and can simultaneously assess significant differences between multiple types of biological samples. Two interconnected mixed linear models are central to the method and provide a flexible means to properly account for variability both across and within genes. The mixed model also provides a convenient framework for evaluating the statistical power of any particular experimental design and thus enables a researcher to a priori select an appropriate number of replicates. We also suggest some basic graphics for visualizing lists of significant genes. Analyses of published experiments studying human cancer and yeast cells illustrate the results.

The nucleotide sequence of a region of the Azotobacter vinelandii genome exhibiting sequence similarity to nifH has been determined. The order of open reading frames within this 6.1-kilobase-pair region was found to be anfH (alternative nitrogen fixation, nifH-like gene), anfD (nifD-like gene), anfG (potentially encoding a protein similar to the product of vnfG from Azotobacter chroococcum), anfK (nifK-like gene), followed by two additional open reading frames. The 5'-flanking region of anfH contains a nif promoter similar to that found in the A. vinelandii nifHDK gene cluster. The presumed products of anfH, anfD, and anfK are similar in predicted Mr and pI to the previously described subunits of nitrogenase 3. Deletion plus insertion mutations introduced into the anfHDGK region of wild-type strain A. vinelandii CA resulted in mutant strains that were unable to grow in Mo-deficient, N-free medium but grew in the presence of 1 microM Na2MoO4 or V2O5. Introduction of the same mutations into the nifHDK deletion strain CA11 resulted in strains that grew under diazotrophic conditions only in the presence of vanadium. The lack of nitrogenase 3 subunits in these mutant strains was demonstrated through two-dimensional gel analysis of protein extracts from cells derepressed for nitrogenase under Mo and V deficiency. These results indicate that anfH, anfD, and anfK encode structural proteins for nitrogenase 3.

Under diazotrophic conditions in the absence of molybdenum (Mo) and vanadium (V), Azotobacter vinelandii reduces N2 to NH4+ by using nitrogenase 3 (encoded by anfHDGK). However, dinitrogenase reductase 2 (encoded by vnfH) is also expressed under these conditions even though this protein is a component of the Vcontaining alternative nitrogenase. Mutant strains that lack dinitrogenase reductase 2 (VnfH-) grow slower than the wild-type strain in N-free, Mo-, and V-deficient medium. In this medium, these strains synthesize dinitrogenase reductase 1 (a component of the Mo-containing nitrogenase encoded by nifH), even though this component is not normally synthesized in the absence of Mo. Strains that lack both dinitrogenase reductases 1 and 2 (NifH-VnfH-) are unable to grow diazotrophically in Mo-and V-deficient medium. In this medium, NifH-VnfH-strains containing an anfH--lacZ transcriptional fusion exhibited less than 3% of the ,B-galactosidase activity observed in the wild type with the same fusion. P-Galactosidase activity expressed by VnfHmutants containing the anJH-lacZ fusion ranged between 57 and 78% of that expressed by the wild type containing the samne fusion. Thus, expression of dinitrogenase reductase 2 seems to be required for transcription of the anfHDGK operon, although, in Vnft-mutants, dinitrogenase reductase 1 appears to serve this function. Active dinitrogenase reductase 1 or 2 is probably required for this function since a nijM deletion mutant containing the anffl-lacZ fusion was unable to synthesize ,-galactosidase above background levels. An anfA deletion strain containing the anpfH-lacZ fusion exhibited ,B-galactosidase activity at 16% of that of the wild type containing the same fusion. However, in the presence of NH4', the ,B-galactosidase activity expressed by this strain more than doubled. This indicates that AnfA is required not only for normal levels of anfHDGK transcription but also for NH4'-and, to a lesser extent, Mo-mediated repression of this transcription.The regulation of the expression of the three nitrogenases in Azotobacter vinelandii is responsive to the presence or absence of ammonium (NH4'), molybdenum (Mo), and vanadium (V) in the culture medium. The synthesis of all three nitrogenases is repressed by NH4'. Nitrogenase 1 is found in cells grown in the presence of Mo and nitrogenase 2 is expressed in the presence of V but in the absence of Mo, whereas nitrogenase 3 is synthesized only in the absence of both Mo and V (4, 9, 13, and references therein). Our knowledge of the molecular basis for nitrogen and metal regulation in A. vinelandii is still rudimentary. The regulatory genes nifA, vnfA, and anfA have been identified, and some of their functions have been described on the basis of the phenotypes of NifA-, VnfA-, and AnfA-mutants (1, 14). The nifA gene product is required for transcription of the structural genes for nitrogenase 1 (1). NifA binds to an upstream activator sequence (6) or indirectly represses the synthesis of nitrogenase 1 in cells grown in Mo-deficient me...

The nucleotide sequence (3,600 bp) of a second copy of nifENX-like genes in Azotobacter vinelandii has been determined. These genes are located immediately downstream from vnfA and have been designated vnJENX. The vnJENX genes appear to be organized as a single transcriptional unit that is preceded by a potential RpoN-dependent promoter. While the nifEN genes are thought to be evolutionarily related to nifDK, the vaJEN genes appear to be more closely related to niJEN than to either nifDK, vnJDK, or anfDK. Mutant strains (CA47 and CA48) carrying insertions in vnfE and vnfN, respectively, are able to grow diazotrophically in molybdenum (Mo)-deficient medium containing vanadium (V) (Vnf+) and in medium lacking both Mo and V (Anf).However, a double mutant (strain DJ42.48) which contains a nijEN deletion and an insertion in vnJE is unable to grow diazotrophically in Mo-sufficient medium or in Mo-deficient medium with or without V. This suggests that NifE and NifN substitute for VnfE and VnfN when the vnjEN genes are mutationally inactivated. AnfA is not required for the expression of a vnJN-lacZ transcriptional fusion, even though this fusion is expressed under Mo-and V-deficient diazotrophic growth conditions. Azotobacter vinelandii is able to grow diazotrophically using any of three genetically distinct nitrogenases depending on the presence or absence of molybdenum (Mo) or vanadium (V) in the growth medium. The well-characterized Mo-containing nitrogenase (nitrogenase-1) is synthesized under conditions of Mo sufficiency. Under conditions where V replaces Mo, an alternative V-containing nitrogenase (nitrogenase-2) is expressed, and in the absence of both Mo and V, an alternative nitrogenase (nitrogenase-3) that does not appear to contain either Mo or V is made (18, 23). Each of these nitrogenase complexes is composed of two protein components, dinitrogenase reductase and dinitrogenase. Dinitrogenase reductase-1 is a dimer of two identical subunits with an Mr of approximately 60,000 (10, 11). Dinitrogenase reductase-2 and dinitrogenase reductase-3 are also dimers of two identical subunits (13,20). Dinitrogenase-1 is a tetramer (Mr, -240,000) consisting of two pairs of nonidentical subunits (a2P2). Dinitrogenase-2 and dinitrogenase-3, on the other hand, are probably hexamers, each containing three pairs of nonidentical subunits (a2P282) (25,39).The structural genes encoding nitrogenase-1 and nitrogenase-3 are organized as single operons (nifHDK and anf HDGKorflord2, respectively), while those encoding nitrogenase-2 form two independently regulated operons, vnfHorf Fd and vn.fDGK (8,25,26,39

No abstract