Human Liver Cell Trafficking Mutants: Characterization and Whole Exome Sequencing

The HuH7 liver cell mutant Trf1 is defective in membrane trafficking and is complemented by the casein kinase 2α subunit CK2α’’. Here we identify characteristic morphologies, trafficking and mutational changes in six additional HuH7 mutants Trf2-Trf7. Trf1 cells were previously shown to be severely defective in gap junction functions. Using a Lucifer yellow transfer assay, remarkable attenuation of gap junction communication was revealed in each of the mutants Trf2-Trf7. Electron microscopy and light microscopy of thiamine pyrophosphatase showed that several mutants exhibited fragmented Golgi apparatus cisternae compared to parental HuH7 cells. Intracellular trafficking was investigated using assays of transferrin endocytosis and recycling and VSV G secretion. Surface binding of transferrin was reduced in all six Trf2-Trf7 mutants, which generally correlated with the degree of reduced expression of the transferrin receptor at the cell surface. The mutants displayed the same transferrin influx rates as HuH7, and for efflux rate, only Trf6 differed, having a slower transferrin efflux rate than HuH7. The kinetics of VSV G transport along the exocytic pathway were altered in Trf2 and Trf5 mutants. Genetic changes unique to particular Trf mutants were identified by exome sequencing, and one was investigated in depth. The novel mutation Ile34Phe in the GTPase RAB22A was identified in Trf4. RNA interference knockdown of RAB22A or overexpression of RAB22AI34F in HuH7 cells caused phenotypic changes characteristic of the Trf4 mutant. In addition, the Ile34Phe mutation reduced both guanine nucleotide binding and hydrolysis activities of RAB22A. Thus, the RAB22A Ile34Phe mutation appears to contribute to the Trf4 mutant phenotype.     

The Selection of S. CEREVISIAE Mutants Defective in the Start Event of Cell Division

Thirty-three temperature-sensitive mutations defective in the start event of the cell division cycle of Saccharomyces cerevisiae were isolated and subjected to preliminary characterization. Complementation studies assigned these mutations to four complementation groups, one of which, cdc28, has been described previously. Genetic analysis revealed that these complementation groups define single nuclear genes, unlinked to one another. One of the three newly identified genes, cdc37, has been located in the yeast linkage map on chromosome IV, two meiotic map units distal to hom2.—Each mutation produces stage-specific arrest of cell division at start, the same point where mating pheromone interrupts division. After synchronization at start by incubation at the restrictive temperature, the mutants retain the capacity to enlarge and to conjugate.     

Radiation-Sensitive Mutants of CAENORHABDITIS ELEGANS

Nine rad (for abnormal radiation sensitivity) mutants hypersensitive to ultraviolet light were isolated in the small nematode Caenorhabditis elegans. The mutations are recessive to their wild-type alleles, map to four of the six linkage groups in C. elegans and define nine new games named rad-1 through rad-9. Two of the mutants—rad-1 and rad-2—are very hypersensitive to X rays, and three—rad-2, rad-3 and rad-4—are hypersensitive to methyl methanesulfonate under particular conditions of exposure. The hypersensitivity of these mutants to more than one DNA-damaging agent suggests that they may be abnormal in DNA repair. One mutant—rad-5, a temperature-sensitive sterile mutant—shows an elevated frequency of spontaneous mutation at more than one locus; rad-4, which shows a cold-sensitive embryogenesis, reduces meiotic X-chromosome nondisjunction tenfold and partially suppresses some but not all mutations that increase meiotic X-chromosome nondisjunction; the viability of rad-6 hermaphrodites is half that of rad-6 males at 25°; and newly mature (but not older) rad-8 hermaphrodites produce many inviable embryo progeny. Meiotic recombination frequencies were measured for seven rad mutants and found to be close to normal.     

The Utilization during Mitotic Cell Division of Loci Controlling Meiotic Recombination and Disjunction in DROSOPHILA MELANOGASTER

To inquire whether the loci identified by recombination-defective and disjunction-defective meiotic mutants in Drosophila are also utilized during mitotic cell division, the effects of 18 meiotic mutants (representing 13 loci) on mitotic chromosome stability have been examined genetically. To do this, meiotic-mutant-bearing flies heterozygous for recessive somatic cell markers were examined for the frequencies and types of spontaneous clones expressing the cell markers. In such flies, marked clones can arise via mitotic recombination, mutation, chromosome breakage, nondisjunction or chromosome loss, and clones from these different origins can be distinguished. In addition, meiotic mutants at nine loci have been examined for their effects on sensitivity to killing by UV and X rays.—Mutants at six of the seven recombination-defective loci examined (mei-9, mei-41, c(3)G, mei-W68, mei-S282, mei-352, mei-218) cause mitotic chromosome instability in both sexes, whereas mutants at one locus (mei-218) do not affect mitotic chromosome stability. Thus many of the loci utilized during meiotic recombination also function in the chromosomal economy of mitotic cells.—The chromosome instability produced by mei-41 alleles is the consequence of chromosome breakage, that of mei-9 alleles is primarily due to chromosome breakage and, to a lesser extent, to an elevated frequency of mitotic recombination, whereas no predominant mechanism responsible for the instability caused by c(3)G alleles is discernible. Since these three loci are defective in their responses to mutagen damage, their effects on chromosome stability in nonmutagenized cells are interpreted as resulting from an inability to repair spontaneous lesions. Both mei-W68 and mei-S282 increase mitotic recombination (and in mei-W68, to a lesser extent, chromosome loss) in the abdomen but not the wing. In the abdomen, the primary effect on chromosome stability occurs during the larval period when the abdominal histoblasts are in a nondividing (G2) state.—Mitotic recombination is at or above control levels in the presence of each of the recombination-defective meiotic mutants examined, suggesting that meiotic and mitotic recombination are under separate genetic control in Drosophila.—Of the six mutants examined that are defective in processes required for regular meiotic chromosome segregation, four (l(1)TW-6^cs, ca^nd, mei-S332, ord) affect mitotic chromosome behavior. At semi-restrictive temperatures, the cold sensitive lethal l(1)TW-6^cs causes very frequent somatic spots, a substantial proportion of which are attributable to nondisjunction or loss. Thus, this locus specifies a function essential for chromosome segregation at mitosis as well as at the first meiotic division in females. The patterns of mitotic effects caused by ca^nd, mei-S332, and ord suggest that they may be leaky alleles at essential loci that specify functions common to meiosis and mitosis. Mutants at the two remaining loci (nod, pal) do not affect mitotic chromosome stability.     

Mutants of the Killer Plasmid of SACCHAROMYCES CEREVISIAE Dependent on Chromosomal Diploidy for Expression and Maintenance

Mutants of the killer plasmid of Saccharomyecs cerevisiae have been isolated that depend upon chromosomal diploidy for the expression of plasmid functions and for replication or maintenance of the plasmid itself. These mutants are not defective in any chromosomal gene needed for expression or replication of the killer plasmid.—Haploids carrying these mutant plasmids (called d for diploid-dependent) are either unable to kill or unable to resist being killed or both and show frequent loss of the plasmid. The wild-type phenotype (K⁺R⁺) is restored by mating the d plasmid-carrying strain with either (a) a wild-type sensitive strain which apparently has no killer plasmid; (b) a strain which has been cured of the killer plasmid by growth at elevated temperature; (c) a strain which has been cured of the plasmid by growth in the presence of cycloheximide; (d) a strain which has lost the plasmid because it carries a mutation in a chromosomal mak gene; or (e) a strain of the opposite mating type which carries the same d plasmid and has the same defective phenotype, indicating that the restoration of the normal phenotype is not due to recombination between plasmid genomes or complementation of plasmid or chromosomal genes.—Sporulation of the phenotypically K⁺R⁺ diploids formed in matings between d and wild-type nonkiller strains yields tetrads, all four of whose haploid spores are defective for killing or resistance or maintenance of the plasmid or a combination of these. Every defective phenotype may be found among the segregants of a single diploid clone carrying a d plasmid. These defective segregants resume the normal killer phenotype in the diploids formed when a second round of mating is performed, and the segregants from a second round of meiosis and sporulation are again defective.     

Expression of cagA,virB/D Complex and/or vacA Genes in Helicobacter pylori Strains Originating from Patients with Gastric Diseases

In order to better understand pathogenicity of Helicobacter pylori, particularly in the context of its carcinogenic activity, we analysed expression of virulence genes: cagA, virB/D complex (virB4, virB7, virB8, virB9, virB10, virB11, virD4) and vacA in strains of the pathogen originating from persons with gastric diseases. The studies were conducted on 42 strains of H. pylori isolated from patients with histological diagnosis of non-atrophic gastritis—NAG (group 1, including subgroup 1 containing cagA⁺ isolates and subgroup 2 containing cagA^- strains), multifocal atrophic gastritis—MAG (group 2) and gastric adenocarcinoma—GC (group 3). Expression of H. pylori genes was studied using microarray technology. In group 1, in all strains of H. pylori cagA⁺ (subgroup 1) high expression of the gene as well as of virB/D was disclosed, accompanied by moderate expression of vacA. In strains of subgroup 2 a moderate expression of vacA was detected. All strains in groups 2 and 3 carried cagA gene but they differed in its expression: a high expression was detected in isolates of group 2 and its hyperexpression in strains of group 3 (hypervirulent strains). In both groups high expression of virB/D and vacA was disclosed. Our results indicate that chronic active gastritis may be induced by both cagA⁺ strains of H. pylori, manifesting high expression of virB/D complex but moderate activity of vacA, and cagA^- strains with moderate expression of vacA gene. On the other hand, in progression of gastric pathology and carcinogenesis linked to H. pylori a significant role was played by hypervirulent strains, manifesting a very high expression of cagA and high activity of virB/D and vacA genes.     

Parameters of Mitotic Recombination in Minute Mutants of DROSOPHILA MELANOGASTER

A sample of 16 Minutes, representing 12 loci distributed over all the chromosome arms and including 3 pairs of alleles and 4 deficiencies, has been studied with respect to several developmental and recombinational parameters. Cell marker mutants located in most of the chromosome arms were used to assess (1) spontaneous and X-ray-induced mitotic recombination frequencies of each Minute, and (2) clone sizes of the different cell marker clones. These parameters were analyzed both in the wing disc and in the abdominal histoblasts.—Whereas spontaneous frequencies are not affected by the presence of the Minutes studied, the different Minutes characteristically increase the frequency of recombination clones arising after X-irradiation. The recombinant clones which are M⁺/M⁺ are significantly larger than clones in the same fly which retain the M⁺/M condition. This is particularly striking in clones in the wing disc, slightly so in clones in the tergites. The occurrence of mitotic recombination in the fourth chromosome is reported for the first time.—Chaeta length and developmental delay correlates with the recombinational parameters in different ways. Possible causal interrelationships of the different traits of the Minute syndrome are discussed.     

Abnormal micronuclear telomeres lead to an unusual cell cycle checkpoint and defects in Tetrahymena oral morphogenesis

Telomere mutants have been well studied with respect to telomerase and the role of telomere binding proteins, but they have not been used to explore how a downstream morphogenic event is related to the mutated telomeric DNA. We report that alterations at the telomeres can have profound consequences on organellar morphogenesis. Specifically, a telomerase RNA mutation termed ter1-43AA results in the loss of germ line micronuclear telomeres in the binucleate protozoan Tetrahymena thermophila. These cells also display a micronuclear mitotic arrest, characterized by an extreme delay in anaphase with an elongated, condensed chromatin and a mitotic spindle apparatus. This anaphase defect suggests telomere fusions and consequently a spindle rather than a DNA damage checkpoint. Most surprisingly, these mutants exhibit unique, dramatic defects in the formation of the cell's oral apparatus. We suggest that micronuclear telomere loss leads to a “dynamic pause” in the program of cortical development, which may reveal an unusual cell cycle checkpoint.     

The proportion of terramycin-resistant mutants in B. megatherium cultures

The number of terramycin-resistant mutants in Bacillus megatherium cultures, their mutation rate, and the growth rate of the wild and mutant cells have been determined under various conditions. These values are in agreement with the following equations (Northrop and Kunitz, 1957):— See PDF for Equation λ = mutation rate, A = growth rate constant of wild cells, B = growth rate constant of mutants, See PDF for Equation equilibrium. The value of the mutation rate as determined from equation (6) agrees with that found by the null fraction method.     

G Protein β Subunit–null Mutants Are Impaired in Phagocytosis and Chemotaxis Due to Inappropriate Regulation of the Actin Cytoskeleton

Chemotaxis and phagocytosis are basically similar in cells of the immune system and in Dictyostelium amebae. Deletion of the unique G protein β subunit in D. discoideum impaired phagocytosis but had little effect on fluid-phase endocytosis, cytokinesis, or random motility. Constitutive expression of wild-type β subunit restored phagocytosis and normal development. Chemoattractants released by cells or bacteria trigger typical transient actin polymerization responses in wild-type cells. In β subunit–null cells, and in a series of β subunit point mutants, these responses were impaired to a degree that correlated with the defect in phagocytosis. Image analysis of green fluorescent protein–actin transfected cells showed that β subunit– null cells were defective in reshaping the actin network into a phagocytic cup, and eventually a phagosome, in response to particle attachment. Our results indicate that signaling through heterotrimeric G proteins is required for regulating the actin cytoskeleton during phagocytic uptake, as previously shown for chemotaxis. Inhibitors of phospholipase C and intracellular Ca²⁺ mobilization inhibited phagocytosis, suggesting the possible involvement of these effectors in the process.     

Frameshift Suppression in SACCHAROMYCES CEREVISIAE. II. Genetic Properties of Group II Suppressors

Suppressors of ICR-induced mutations that exhibit behavior similar to bacterial frameshift suppressors have been identified in the yeast Saccharomyces cerevisiae. The yeast suppressors have been divided into two groups. One of these groups (Group II: SUF1, SUF3, SUF4, SUF5 and SUF6) appears to include a set of informational suppressors in which the vehicle of suppression is glycyl-tRNA. Some of the genetic properties of Group II suppressors are described in this communication.——Corevertants of the Group II frameshift mutations his4–519 and leu2–3 have been characterized to determine the spectrum of reversion events induced by the frameshift mutagen ICR-170. Seventythree ICR-induced corevertants were analyzed. With the exception of one corevertant, which carried an allele of SUF1, all carried alleles of SUF3 or SUF5. SUF1, SUF3, SUF4 and SUF6 were represented among spontaneous and UV-induced corevertants. In the course of these experiments one of the suppressors was mapped. SUF5, the probable structural gene for tRNA^GLY1, is located between ade2 and ade9 on chromosome XV.——SUF1, SUF4 and SUF6 have novel properties and comprise a distinct subset of suppressors. Although these suppressors show no genetic linkage to each other, they share several common features including lethality in haploid pairwise combinations, reduced tRNA^GLY3 isoacceptor activity and increased efficiency of suppression in strains carrying the cytoplasmically inherited [PSI] element. In addition, strains carrying SUF1, SUF4 or SUF6 are phenotypically unstable and give rise to mitotic Suf⁺ segregants at high frequency. These segregants invariably contain a linked, second-site mutation that maps in or adjacent to the suppressor gene itself. Strains carrying any of these suppressors also give rise to mitotic segregants that exhibit enhanced efficiency of suppression; mutations responsible for this phenotype map at two loci, upf1 and upf2. These genes show no genetic linkage to any of the Group II suppressors.——Methods that permit positive selection for mutants with decreased or enhanced efficiency of suppression have been devised in order to examine large numbers of variants. The importance of these interacting mutants is underscored by their potential utility in studying suppressor function at the molecular level.     

Yeast Cycloheximide-resistant crl Mutants Are Proteasome Mutants Defective in Protein Degradation

In 1988 McCusker and Haber generated a series of mutants which are resistant to the minimum inhibitory concentration of the protein synthesis inhibitor cycloheximide. These cycloheximide-resistant, temperature-sensitive (crl) mutants, in addition, exhibited other pleiotropic phenotypes, e.g., incorrect response to starvation, hypersensitivity against amino acid analogues, and other protein synthesis inhibitors. Temperature sensitivity of one of these mutants, crl3–2, had been found to be suppressed by a mutation, SCL1–1, which resided in an α-type subunit of the 20S proteasome. We cloned the CRL3 gene by complementation and found CRL3 to be identical to the SUG1/CIM3 gene coding for a subunit of the 19S cap complex of the 26S proteasome. Another mutation, crl21, revealed to be allelic with the 20S proteasomal gene PRE3. crl3–2 and crl21 mutant cells show significant defects in proteasome-dependent proteolysis, whereas the SCL1–1 suppressor mutation causes partial restoration of crl3–2-induced proteolytic defects. Notably, cycloheximide resistance was also detected for other proteolytically deficient proteasome mutants (pre1–1, pre2–1, pre3–1, pre4–1). Moreover, proteasomal genes were found within genomic sequences of 9 of 13 chromosomal loci to which crl mutations had been mapped. We therefore assume that most if not all crl mutations reside in the proteasome and that phenotypes found are a result of defective protein degradation.     

Isolation and Characterization of pso Mutants Sensitive to Photo-Addition of Psoralen Derivatives in SACCHAROMYCES CEREVISIAE

We have isolated mutants sensitive to photo-addition of bi-functional and mono-functional derivatives of psoralen in Saccharomyces cerevisiae. Three of these pso mutants were analyzed in detail. They segregate in meiosis like Mendelian genes and complement each other, as well as existing radiation-sensitive (rad and rev) mutants. The study of heterozygous diploid strains (PSO⁺/pso) indicates that the three pso genes are recessive. The mutant pso1–1 demonstrates a cross-sensitivity to UV and γ-rays, whereas mutants pso2–1 and pso3–1 are specifically sensitive to photo-addition of psoralen derivatives. The comparison of exponentially growing cells to stationary-phase cells demonstrates that for the three mutants the defect in repair capacity of DNA cross-links and monoadducts concerns G1 and early S-phase cells. The pso2–1 mutant is, however, also defective in G2 repair and loses diploid resistance when it is in the homozygous state.—The block in repair capacity in these novel mutants is discussed in relation to the three other repair pathways known to be involved in the repair of furocoumarins photo-induced lesions in yeast DNA.     

Two Chromosomal Genes Required for Killing Expression in Killer Strains of SACCHAROMYCES CEREVISIAE

The killer character of yeast is determined by a 1.4 x 10⁶ molecular weight double-stranded RNA plasmid and at least 12 chromosomal genes. Wild-type strains of yeast that carry this plasmid (killers) secrete a toxin which is lethal only to strains not carrying this plasmid (sensitives). ——— We have isolated 28 independent recessive chromosomal mutants of a killer strain that have lost the ability to secrete an active toxin but remain resistant to the effects of the toxin and continue to carry the complete cytoplasmic killer genome. These mutants define two complementation groups, kex1 and kex2. Kex1 is located on chromosome VII between ade5 and lys5. Kex2 is located on chromosome XIV, but it does not show meiotic linkage to any gene previously located on this chromosome. ——— When the killer plasmid of kex1 or kex2 strains is eliminated by curing with heat or cycloheximide, the strains become sensitive to killing. The mutant phenotype reappears among the meiotic segregants in a cross with a normal killer. Thus, the kex phenotype does not require an alteration of the killer plasmid. ——— Kex1 and kex2 strains each contain near-normal levels of the 1.4 x 10⁶ molecular weight double-stranded RNA, whose presence is correlated with the presence of the killer genome.     

Scw1p antagonizes the septation initiation network to regulate septum formation and cell separation in the fission yeast Schizosaccharomyces pombe

Cytokinesis in the fission yeast Schizosaccharomyces pombe is regulated by a signaling pathway termed the septation initiation network (SIN). The SIN is essential for initiation of actomyosin ring constriction and septum formation. In a screen to search for mutations that can rescue the sid2-250 SIN mutant, we obtained scw1-18. Both the scw1-18 mutant and the scw1 deletion mutant (scw1Δ mutant), have defects in cell separation. Both the scw1-18 and scw1Δ mutations rescue the growth defects of not just the sid2-250 mutant but also the other temperature-sensitive SIN mutants. Other cytokinesis mutants, such as those defective for actomyosin ring formation, are not rescued by scw1Δ. scw1Δ does not seem to rescue the SIN by restoring SIN signaling defects. However, scw1Δ may function downstream of the SIN to promote septum formation, since scw1Δ can rescue the septum formation defects of the cps1-191β-1,3-glucan synthase mutant, which is required for synthesis of the primary septum.     

std1, a Gene Involved in Glucose Transport in Schizosaccharomyces pombe

A wild-type strain, Sp972 h⁻, of Schizosaccharomyces pombe was mutagenized with ethylmethanesulfonate (EMS), and 2-deoxyglucose (2-DOG)-resistant mutants were isolated. Out of 300 independent 2-DOG-resistant mutants, 2 failed to grow on glucose and fructose (mutants 3/8 and 3/23); however, their hexokinase activity was normal. They have been characterized as defective in their sugar transport properties, and the mutations have been designated as std1-8 and std1-23 (sugar transport defective). The mutations are allelic and segregate as part of a single gene when the mutants carrying them are crossed to a wild-type strain. We confirmed the transport deficiency of these mutants by [¹⁴C]glucose uptake. They also fail to grow on other monosaccharides, such as fructose, mannose, and xylulose, as well as disaccharides, such as sucrose and maltose, unlike the wild-type strain. Lack of growth of the glucose transport-deficient mutants on maltose revealed the extracellular breakdown of maltose in S. pombe, unlike in Saccharomyces cerevisiae. Both of the mutants are unable to grow on low concentrations of glucose (10 to 20 mM), while one of them, 3/23, grows on high concentrations (50 to 100 mM) as if altered in its affinity for glucose. This mutant (3/23) shows a lag period of 12 to 18 h when grown on high concentrations of glucose. The lag disappears when the culture is transferred from the log phase of its growth on high concentrations. These mutants complement phenotypically similar sugar transport mutants (YGS4 and YGS5) reported earlier by Milbradt and Hoefer (Microbiology 140:2617–2623, 1994), and the clone complementing YGS4 and YGS5 was identified as the only glucose transporter in fission yeast having 12 transmembrane domains. These mutants also demonstrate two other defects: lack of induction and repression of shunt pathway enzymes and defective mating.     

Identity Coefficients in Finite Populations. I. Evolution of Identity Coefficients in a Random Mating Diploid Dioecious Population

Properties of identity relation between genes are discussed, and a derivation of recurrent equations of identity coefficients in a random mating, diploid dioecious population is presented. Computations are run by repeated matrix multiplication. Results show that for effective population size (N_e) larger than 16 and no mutation, a given identity coefficient at any time t can be expressed approximately as a function of (1—f), (1—f)³ and (1— f)⁶, where f is the mean inbreeding coefficient at time t. Tables are presented, for small N_e values and extreme sex ratios, showing the pattern of change in the identity coefficients over time. The pattern of evolution of identity coefficients is also presented and discussed with respect to N _eu, where u is the mutation rate. Applications of these results to the evolution of genetic variability within and between inbred lines are discussed.     

Proteasome Control of [URE3] Prion Propagation by Degradation of Anti-Prion Proteins Cur1 and Btn2 in Saccharomyces cerevisiae

[URE3] is a prion of the nitrogen catabolism controller, Ure2p, and [PSI+] is a prion of the translation termination factor Sup35p in S. cerevisiae. Btn2p cures [URE3] by sequestration of Ure2p amyloid filaments. Cur1p, paralogous to Btn2p, also cures [URE3], but by a different (unknown) mechanism. We find that an array of mutations impairing proteasome assembly or MG132 inhibition of proteasome activity result in loss of [URE3]. In proportion to their prion—curing effects, each mutation affecting proteasomes elevates the cellular concentration of the anti-prion proteins Btn2 and Cur1. Of >4,600 proteins detected by SILAC, Btn2p was easily the most overexpressed in a pre9Δ (α3 core subunit) strain. Indeed, deletion of BTN2 and CUR1 prevents the prion—curing effects of proteasome impairment. Surprisingly, the 15 most unstable yeast proteins are not increased in pre9Δ cells suggesting altered proteasome specificity rather than simple inactivation. Hsp42, a chaperone that cooperates with Btn2 and Cur1 in curing [URE3], is also necessary for the curing produced by proteasome defects, although Hsp42p levels are not substantially altered by a proteasome defect. We find that pre9Δ and proteasome chaperone mutants that most efficiently lose [URE3], do not destabilize [PSI+] or alter cellular levels of Sup35p. A tof2 mutation or deletion likewise destabilizes [URE3], and elevates Btn2p, suggesting that Tof2p deficiency inactivates proteasomes. We suggest that when proteasomes are saturated with denatured/misfolded proteins, their reduced degradation of Btn2p and Cur1p automatically upregulates these aggregate-handling systems to assist in the clean-up.     

Mutants Affected in Alkaline Phosphatase Expression: Evidence for Multiple Positive Regulators of the Phosphate Regulon in ESCHERICHIA COLI

The expression of alkaline phosphatase (the product of the phoA gene) in Escherichia coli is believed to be subject to both positive control by the phoB gene product and negative control by the phoR gene product. We have isolated a large number of PhoA^- mutants in the phoR^- genetic background. Among mutants altered in the positive control of alkaline phosphatase, some were phoB mutants; others had a mutation in a new gene, designated phoM. We believe that the phoM gene codes for a positive regulator that acts together with the phoB gene product in phoA gene expressions.—The PhoM phenotype was found to be masked in phoR⁺ strains. This and other evidence support a positive regulatory role for the phoR gene product as well.—Our experiments demonstrate that phoA is under positive control by three different positive regulators: the products of the phoB, phoM and phoR genes. The phoB gene product is always needed together with either the phoR or phoM gene product. In addition, the phoR gene product acts as a negative regulator.—We describe a model for phoA gene expression consistent with this new evidence.