Crystal structure of PhnH: an essential component of carbon-phosphorus lyase in Escherichia coli

Organophosphonates are reduced forms of phosphorous that are characterized by the presence of a stable carbon-phosphorus (C-P) bond, which resists chemical hydrolysis, thermal decomposition, and photolysis. The chemically inert nature of the C-P bond has raised environmental concerns as toxic phosphonates accumulate in a number of ecosystems. Carbon-phosphorous lyase (CP lyase) is a multienzyme pathway encoded by the phn operon in gram-negative bacteria. In Escherichia coli 14 cistrons comprise the operon (phnCDEFGHIJKLMNOP) and collectively allow the internalization and degradation of phosphonates. Here we report the X-ray crystal structure of the PhnH component at 1.77 Å resolution. The protein exhibits a novel fold, although local similarities with the pyridoxal 5′-phosphate-dependent transferase family of proteins are apparent. PhnH forms a dimer in solution and in the crystal structure, the interface of which is implicated in creating a potential ligand binding pocket. Our studies further suggest that PhnH may be capable of binding negatively charged cyclic compounds through interaction with strictly conserved residues. Finally, we show that PhnH is essential for C-P bond cleavage in the CP lyase pathway.     

Bioconversion of sunflower oil,rapeseed oil and ricinoleic acid by Candida tropicalis M25

On the basis of mutational analysis, the genes for phosphonate uptake and degradation in Escherichia coli were shown to be organized in a 10.9-kb operon of 14 genes (named phnC to phnP) and induced by phosphate (P_i) starvation [Metcalf and Wanner (1993) J Bacteriol 175: 3430–3442]. The repression of phosphonate utilization by P_i has hindered both the biochemical characterization of the carbon-phosphorus (C-P) lyase activity and the development of improved methods for phosphonate biodegradation in biotechnology. We have cloned the genes phnG to phnP (associated with C-P lyase activity) with the lac promoter to provide expression of C-P lyase in the presence of P_i. A number of strains lacking portions of the phn operon have been constructed. In vivo complementation of the strains, in which phnC to phnP (including both Pn transport and catalysis genes) or phnH to phnP (including only catalysis genes) was deleted, with plasmids carrying various fragments of the phn operon revealed that the expression of phnC-phnP gene products is essential to restore growth on minimal medium with phosphonate as the sole phosphorus source, while phnG-phnM gene products are required for C-P lyase activity as assessed by in vivo methane production from methylphosphonic acid. The minimum size of the DNA required for the whole-cell C-P lyase activity has been determined to be a 5.8-kb fragment, encompassing the phnG to phnM genes. Therefore, there is no requirement for the phnCDE-encoded phosphonate transport system, suggesting that cleavage of the C-P bond may occur on the outer surface of the inner membrane of E. coli cells, releasing the carbon moiety into the periplasm. These data are in agreement with the observation that phosphonates cannot serve as the carbon source for E.␣coli growth. Received: 23 September 1997 / Received revision: 5 January 1998 / Accepted: 24 January 1998     

Molecular genetic studies of a 10.9-kb operon in Escherichia coli for phosphonate uptake and biodegradation

Bacteria that use phosphonates as a phosphorus source must be able to break the stable carbon-phosphorus bond. In Escherichia coli phosphonates are broken down by a C-P lyase that has a broad substrate specificity. Evidence for a lyase is based on in vivo studies of product formation because it has been proven difficult to detect the activity in vitro. By using molecular genetic techniques, we have studied the genes for phosphonate uptake and degradation in E. coli, which are organized in an operon of 14 genes, named phnC to phnP. As expected for genes involved in P acquisition, the phnC-phnP operon is a member of the PHO regulon and is induced many hundred-fold during phosphate limitation. Three gene products (PhnC, PhnD and PhnE) comprise a binding protein-dependent phosphonate transporter, which also transports phosphate, phosphite, and certain phosphate esters such as phosphoserine; two gene products (PhnF and PhnO) may have a role in gene regulation; and nine gene products (PhnG, PhnH, PhnI, PhnJ, PhnK, PhnL, PhnM, PhnN, and PhnP) probably comprise a membrane-associated C-P lyase enzyme complex. Although E. coli can degrade many different phosphonates, the ability to use certain phosphonates appears to be limited by the specificity of the PhnCDE transporter and not by the specificity of the C-P lyase.     

Isolation of a glyphosate-metabolising Pseudomonas: detection,partial purification and localisation of carbon-phosphorus lyase

A Pseudomonas isolate (GLC11) capable of growth in the presence of up to 125 mM glyphosate [N-phosphonomethyl glycine (PMG)] has been isolated. Unlike the previously isolated Pseudomonas PG2982 and other bacterial strains, isolate GLC11 grows equally well in commercial formulation and analytical grade PMG. Utilisation of PMG as a phosphorus source is repressed by inorganic phosphate (Pi) in both isolates. Enzymatic activity responsible for carbon-phosphorus bond cleavage (C-P lyase) was detected in cell-free extracts of both isolates and was partially purified. Resolution on DE-52 anion exchange chromatography yielded a single peak of C-P lyase activity. The molecular mass of C-P lyase as analysed by gel permeation chromatography is approximately 200 kDa. The enzyme activity was localised in the periplasmic space of bacteria. The specific activity of C-P lyase was different for different phosphonates when used as substrates. Correspondence to: R. K. Bhatnagar     

Evidence for two phosphonate degradative pathways in Enterobacter aerogenes.

We screened mini-Mu plasmid libraries from Enterobacter aerogenes IFO 12010 for plasmids that complement Escherichia coli phn mutants that cannot use phosphonates (Pn) as the sole source of phosphorus (P). We isolated two kinds of plasmids that, unexpectedly, encode genes for different metabolic pathways. One kind complements E. coli mutants with both Pn transport and Pn catalysis genes deleted; these plasmids allow degradation of the 2-carbon-substituted Pn alpha-aminoethylphosphonate but not of unsubstituted alkyl Pn. This substrate specificity is characteristic of a phosphonatase pathway, which is absent in E. coli. The other kind complements E. coli mutants with Pn catalysis genes deleted but not those with both transport and catalysis genes deleted; these plasmids allow degradation of both substituted and unsubstituted Pn. Such a broad substrate specificity is characteristic of a carbon-phosphorus (C-P) lyase pathway, which is common in gram-negative bacteria, including E. coli. Further proof that the two kinds of plasmids encode genes for different pathways was demonstrated by the lack of DNA homology between the plasmids. In particular, the phosphonatase clone from E. aerogenes failed to hybridize to the E. coli phnCDEFGHIJKLMNOP gene cluster for Pn uptake and degradation, while the E. aerogenes C-P lyase clone hybridized strongly to the E. coli phnGHIJKLM genes encoding C-P lyase but not to the E. coli phnCDE genes encoding Pn transport. Specific hybridization by the E. aerogenes C-P lyase plasmid to the E. coli phnF, phnN, phnO, and phnP genes was not determined. Furthermore, we showed that one or more genes encoding the apparent E. aerogenes phosphonatase pathway, like the E. coli phnC-to-phnP gene cluster, is under phosphate regulon control in E. coli. This highlights the importance of Pn in bacterial P assimilation in nature.     

Accumulation of Intermediates of the Carbon-Phosphorus Lyase Pathway for Phosphonate Degradation in phn Mutants of Escherichia coli

The catabolism of phosphonic acids occurs in Escherichia coli by the carbon-phosphorus lyase pathway, which is governed by the 14-cistron phn operon. Here, several compounds are shown to accumulate in strains of E. coli with genetic blocks in various phn cistrons when the strains are fed with phosphonate.Phosphonates (Pn), which contain the carbon-phosphorus bond, are quite abundant in nature, primarily as components of phosphonolipids, where 2-aminoethyl phosphonate (AEPn) is analogous to ethanolamine phosphate as the constituent of phospholipids. In addition, Pn are constituents of polysaccharides, glycoproteins, glycolipids, and several antibiotics (25). Furthermore, large amounts of manmade Pn enter the environment (24). Pn utilization polypeptides in Escherichia coli are specified by the phnCDEFGHIJKLMNOP operon (15). Of the 14 cistrons, one (phnF) encodes a repressor protein, as inferred from sequence alignment (6); three (phnCDE) encode an ABC transport system for Pn (22); seven (phnGHIJKLM) have been postulated to encode the carbon-phosphorus (CP) lyase activity (26); and three (phnNOP) have been postulated to encode “auxiliary enzymes.” Among the latter genes, the phnO gene has been shown to specify an enzyme with aminoalkylphosphonate N-acetyltransferase activity (5). We previously identified the product of the phnN gene as an enzyme capable of catalyzing the phosphorylation of ribose 1,5-bisphosphate to 5-phosphoribosyl α-d-1-diphosphate (ribose 1,5-bisphosphate phosphokinase; EC2.7.4.23) (11). The relation of these two compounds to Pn catabolism remains to be established. Since both the substrate and the product of ribose 1,5-bisphosphate phosphokinase are phosphorylated compounds, we inferred that at least some of the other intermediates of the pathway might be also phosphate esters. It might therefore be possible to identify these intermediates by labeling cells with radioactive phosphate ion in the presence of Pn followed by visualization by appropriate thin-layer chromatography (TLC) procedures. In addition, we employed mutants defective in individual genes of the pathway, and we looked for the accumulation of radiolabeled compounds as candidates for members of the phn pathway. Finally, the mutant strains used harbored the ΔpstS605 allele to render the expression of the phn operon constitutive and, thus, independent of the phosphate supply.     

Catabolism and Detoxification of 1-Aminoalkylphosphonic Acids: N-Acetylation by the phnO Gene Product

In Escherichia coli uptake and catabolism of organophosphonates are governed by the phnCDEFGHIJKLMNOP operon. The phnO cistron is shown to encode aminoalkylphosphonate N-acetyltransferase, which utilizes acetylcoenzyme A as acetyl donor and aminomethylphosphonate, (S)- and (R)-1-aminoethylphosphonate, 2-aminoethyl- and 3-aminopropylphosphonate as acetyl acceptors. Aminomethylphosphonate, (S)-1-aminoethylphosphonate, 2-aminoethyl- and 3-aminopropylphosphonate are used as phosphate source by E. coli phn⁺ strains. 2-Aminoethyl- or 3-aminopropylphosphonate but not aminomethylphosphonate or (S)-1-aminoethylphosphonate is used as phosphate source by phnO strains. Neither phn⁺ nor phnO strains can use (R)-1-aminoethylphosphonate as phosphate source. Utilization of aminomethylphosphonate or (S)-1-aminoethylphosphonate requires the expression of phnO. In the absence of phnO-expression (S)-1-aminoethylphosphonate is bacteriocidal and rescue of phnO strains requires the simultaneous addition of d-alanine and phosphate. An intermediate of the carbon-phosphorus lyase pathway, 5′-phospho-α-d-ribosyl 1′-(2-N-acetamidoethylphosphonate), a substrate for carbon-phosphorus lyase, was found to accumulate in cultures of a phnP mutant strain. The data show that the physiological role of N-acetylation by phnO-specified aminoalkylphosphonate N-acetyltransferase is to detoxify (S)-1-aminoethylphosphonate, an analog of d-alanine, and to prepare (S)-1-aminoethylphosphonate and aminomethylphosphonate for utilization of the phosphorus-containing moiety.     

Molecular cloning, mapping, and regulation of Pho regulon genes for phosphonate breakdown by the phosphonatase pathway of Salmonella typhimurium LT2.

Two pathways exist for cleavage of the carbon-phosphorus (C-P) bond of phosphonates, the C-P lyase and the phosphonatase pathways. It was previously demonstrated that Escherichia coli carries genes (named phn) only for the C-P lyase pathway and that Enterobacter aerogenes carries genes for both pathways (K.-S. Lee, W. W. Metcalf, and B. L. Wanner, J. Bacteriol. 174:2501-2510, 1992). In contrast, here it is shown that Salmonella typhimurium LT2 carries genes only for the phosphonatase pathway. Genes for the S. typhimurium phosphonatase pathway were cloned by complementation of E. coli delta phn mutants. Genes for these pathways were proven not to be homologous and to lie in different chromosomal regions. The S. typhimurium phn locus lies near 10 min; the E. coli phn locus lies near 93 min. The S. typhimurium phn gene cluster is about 7.2 kb in length and, on the basis of gene fusion analysis, appears to consist of two (or more) genes or operons that are divergently transcribed. Like that of the E. coli phn locus, the expression of the S. typhimurium phn locus is activated under conditions of Pi limitation and is subject to Pho regulon control. This was shown both by complementation of the appropriate E. coli mutants and by the construction of S. typhimurium mutants with lesions in the phoB and pst loci, which are required for activation and inhibition of Pho regulon gene expression, respectively. Complementation studies indicate that the S. typhimurium phn locus probably includes genes both for phosphonate transport and for catalysis of C-P bond cleavage.     

苜蓿中华根瘤菌desA基因功能的鉴定

为研究苜蓿中华根瘤菌脂肪酸脱饱和酶desA基因在不饱和脂肪酸合成、共生结瘤固氮以及应对逆境胁迫中的功能,为高效利用苜蓿中华根瘤菌提供理论依据,本文通过异体遗传互补和脂肪酸组成薄层层析,分析SmdesA编码蛋白是否具有脱饱和酶的活性并参与不饱和脂肪酸的合成,构建SmdesA的缺失突变株和互补菌株,比较各菌株在不同逆境胁迫条件下的生长速率以及回接宿主植物后与紫花苜蓿共生结瘤的能力.结果表明SmdesA不能互补大肠杆菌CY57中EcfabA的突变,但具有将饱和脂肪酸脱饱和形成不饱和的棕榈油酸和十八碳烯酸的能力.另外,SmdesA缺失突变对苜蓿中华根瘤菌的脂肪酸组成影响不大,但会显著影响低温和高盐条件下菌株的生长速率以及与紫花苜蓿共生结瘤的能力.我们推测,SmdesA参与的脱饱和途径可能是苜蓿中华根瘤菌不饱和脂肪酸合成的补偿途径,其编码的蛋白DesA不是不饱和脂肪酸合成的关键酶,但在应对逆境胁迫和共生结瘤中具有重要的生物学功能.     

Interspecies regulation of the recA gene of gram-negative bacteria lacking an E. coli-like SOS operator

The recA genes of Agrobacterium tumefaciens, Rhizobium meliloti, Rhizobium phaseoli and Rhodobacter sphaeroides, species belonging to the alpha-group bacteria of the Proteobacteria class, have been fused in vitro to the lacZ gene of Escherichia coli. By using a mini-Tn5 transposon derivative, each of these recA-lacZ fusions was introduced into the chromosome of each of the four species, and into that of E. coli. The recA genes of three of the alpha bacteria are induced by DNA damage when inserted in A. tumefaciens, R. phaseoli or R. meliloti chromosomes. The expression of the recA gene of R. sphaeroides is DNA damage-mediated only when present in its own chromosome; none of the genes is induced in E. coli. Likewise, the recA gene of E. coli is not induced in any of the four alpha species. These data indicate that A. tumefaciens, R. meliloti and R. phaseoli possess a LexA-like repressor, which is able to block the expression of their recA genes, as well as that of R. sphaeroides, but not the recA gene of E. coli. The LexA repressor of R. sphaeroides does not repress the recA gene of A. tumefaciens, R. meliloti, R. phaseoli or E. coli.     

Cloning and characterization of a Rhizobium meliloti nonspecific acid phosphatase

Nodulated legumes require high levels of phosphorus for optimal symbiotic performance. However, the basis for this elevated phosphorus requirement is poorly understood, and very little information regarding bacteroid phosphorus metabolism is available. To develop an understanding of the relative importance of organic and inorganic phosphorus sources for bacteroids, we investigated phosphatase activity in Rhizobium meliloti. An R. meliloti plasmid library clone that complemented an Escherichia coli phosphatase mutant was isolated, and the clone was sequenced. The complementing fragment contained a 337-amino-acid open reading frame that has a potential leader sequence and processing sites characteristic of periplasmic proteins. The phosphatase activity was located in the periplasm of R. meliloti and of E. coli containing the cloned gene. The subunit molecular mass of the cloned phosphatase was 33 kDa, and gel filtration indicated the active enzyme was a 66-kDa homodimer. Lack of substrate specificity suggests the cloned gene, napD, encodes a nonspecific acid phosphatase with a pH optimum of approximately 6.5. An R. meliloti napD transposon-insertion mutant was constructed, and its symbiotic phenotype was determined to be Fix⁺ regardless of the level of phosphorus provided to the host plant. Received: 26 August 1997 / Accepted: 4 February 1998     

Characterization and Nucleotide Sequence of the Cryptic Cel Operon of Escherichia Coli K12

Wild-type Escherichia coli are not able to utilize beta-glucoside sugars because the genes for utilization of these sugars are cryptic. Spontaneous mutations in the cel operon allow its expression and enable the organism to ferment cellobiose, arbutin and salicin. In this report we describe the structure and nucleotide sequence of the cel operon. The cel operon consists of five genes: celA, whose function is unknown; celB and celC which encode phosphoenolpyruvate-dependent phosphotransferase system enzyme IIcel and enzyme IIIcel, respectively, for the transport and phosphorylation of beta-glucoside sugars; celD, which encodes a negative regulatory protein; and celF, which encodes a phospho-beta-glucosidase that acts on phosphorylated cellobiose, arbutin and salicin. The mutationally activated cel operon is induced in the presence of its substrates, and is repressed in their absence. A comparison of proteins encoded by the cel operon with functionally equivalent proteins of the bgl operon, another cryptic E. coli gene system responsible for the catabolism of beta-glucoside sugars, revealed no significant homology between these two systems despite common functional characteristics. The celD and celF encoded repressor and phospho-beta-glucosidase proteins are homologous to the melibiose regulatory protein and to the melA encoded alpha-galactosidase of E. coli, respectively. Furthermore, the celC encoded PEP-dependent phosphotransferase system enzyme IIIcel is strikingly homologous to an enzyme IIIlac of the Gram-positive organism Staphylococcus aureus. We conclude that the genes for these two enzyme IIIs diverged much more recently than did their hosts, indicating that E. coli and S. aureus have undergone relatively recent exchange of chromosomal genes.     

The Rhizobium etli FixL protein differs in structure from other known FixL proteins

The central heme-binding domain in the FixL proteins of Sinorhizobium meliloti, Bradyrhizobium japonicum, Rhizobium leguminosarum biovar viciae and Azorhizobium caulinodans, is highly conserved. The similarity with the corresponding domain in the Rhizobium etli FixL protein is considerably less. This observation prompted us to analyze the heme-binding capacities of the R. etli FixL protein. The R. etlifixL gene was overexpressed in Escherichia coli. In the presence of S. meliloti FixJ, the overexpressed R. etli FixL protein was able to enhance FixJ-mediated activation of an S. meliloti pnifA-lacZ fusion, indicating that the R.?etli FixL protein possesses an active conformation in E. coli. Subsequently, using a non-denaturing gel assay for heme, we analyzed the heme-binding capacity of the R.?etli FixL protein expressed in E. coli, taking the S.?meliloti FixL protein as a positive control. The R. etli FixL protein expressed in E. coli does not contain a heme group, in contrast to the S. meliloti FixL protein. Therefore we conclude that the R. etli FixL is a non-heme protein in the nif regulatory cascade.     

Expression and Regulation of the Arsenic Resistance Operon of Acidiphilium multivorum AIU 301 Plasmid pKW301 in Escherichia coli

The arsenic resistance (ars) operon from plasmid pKW301 of Acidiphilium multivorum AIU 301 was cloned and sequenced. This DNA sequence contains five genes in the following order: arsR, arsD, arsA, arsB, arsC. The predicted amino acid sequences of all of the gene products are homologous to the amino acid sequences of the ars gene products of Escherichia coli plasmid R773 and IncN plasmid R46. The ars operon cloned from A. multivorum conferred resistance to arsenate and arsenite on E. coli. Expression of the ars genes with the bacteriophage T7 RNA polymerase-promoter system allowed E. coli to overexpress ArsD, ArsA, and ArsC but not ArsR or ArsB. The apparent molecular weights of ArsD, ArsA, and ArsC were 13,000, 64,000, and 16,000, respectively. A primer extension analysis showed that the ars mRNA started at a position 19 nucleotides upstream from the arsR ATG in E. coli. Although the arsR gene of A. multivorum AIU 301 encodes a polypeptide of 84 amino acids that is smaller and less homologous than any of the other ArsR proteins, inactivation of the arsR gene resulted in constitutive expression of the ars genes, suggesting that ArsR of pKW301 controls the expression of this operon.     

Functional testing of putative oligopeptide permease (Opp) proteins of Borrelia burgdorferi: a complementation model in opp− Escherichia coli

Studies of the protein function of Borrelia burgdorferi have been limited by a lack of tools for manipulating borrelial DNA. We devised a system to study the function of a B. burgdorferi oligopeptide permease (Opp) orthologue by complementation with Escherichia coli Opp proteins. The Opp system of E. coli has been extensively studied and has well defined substrate specificities. The system is of interest in B. burgdorferi because analysis of its genome has revealed little identifiable machinery for synthesis or transport of amino acids and only a single intact peptide transporter operon. As such, peptide uptake may play a major role in nutrition for the organism. Substrate specificity for ABC peptide transporters in other organisms is determined by their substrate binding protein. The B. burgdorferi Opp operon differs from the E. coli Opp operon in that it has three separate substrate binding proteins, OppA-1, -2 and -3. In addition, B. burgdorferi has two OppA orthologues, OppA-4 and -5, encoded on separate plasmids. The substrate binding proteins interact with integral membrane proteins, OppB and OppC, to transport peptides into the cell. The process is driven by two ATP binding proteins, OppD and OppF. Using opp-deleted E. coli mutants, we transformed cells with B. burgdorferi oppA-1, -2, -4 or -5 and E. coli oppBCDF. All of the B. burgdorferi OppA proteins are able to complement E. coli OppBCDF to form a functional Opp transport system capable of transporting peptides for nutritional use. Although there is overlap in substrate specificities, the substrate specificities for B. burgdorferi OppAs are not identical to that of E. coli OppA. Transport of toxic peptides by B. burgdorferi grown in nutrient-rich medium parallels borrelial OppA substrate specificity in the complementation system. Use of this complementation system will pave the way for more detailed studies of B. burgdorferi peptide transport than currently available tools for manipulating borrelial DNA will allow.     

Succinoglycan Production by Rhizobium meliloti Is Regulated through the ExoS-ChvI Two-Component Regulatory System

The Rhizobium meliloti exoS gene is involved in regulating the production of succinoglycan, which plays a crucial role in the establishment of the symbiosis between R. meliloti Rm1021 and its host plant, alfalfa. The exoS96::Tn5 mutation causes the upregulation of the succinoglycan biosynthetic genes, thereby resulting in the overproduction of succinoglycan. Through cloning and sequencing, we found that the exoS gene is a close homolog of the Agrobacterium tumefaciens chvG gene, which has been proposed to encode the sensor protein of the ChvG-ChvI two-component regulatory system, a member of the EnvZ-OmpR family. Further analyses revealed the existence of a newly discovered A. tumefaciens chvI homolog located just upstream of the R. meliloti exoS gene. R. meliloti ChvI may serve as the response regulator of ExoS in a two-component regulatory system. By using ExoS-specific antibodies, it was found that the ExoS protein cofractionated with membrane proteins, suggesting that it is located in the cytoplasmic membrane. By using the same antibodies, it was shown that the exoS96::Tn5 allele encodes an N-terminal truncated derivative of ExoS. The cytoplasmic histidine kinase domain of ExoS was expressed in Escherichia coli and purified, as was the R. meliloti ChvI protein. The ChvI protein autophosphorylated in the presence of acetylphosphate, and the ExoS cytoplasmic domain fragment autophosphorylated at a histidine residue in the presence of ATP. The ChvI protein was phosphorylated in the presence of ATP only when the histidine kinase domain of ExoS was also present. We propose a model for regulation of succinoglycan production by R. meliloti through the ExoS-ChvI two-component regulatory system.     

Construction and Complementation of In-Frame Deletions of the Essential Escherichia coli Thymidylate Kinase Gene

This work reports the construction of Escherichia coli in-frame deletion strains of tmk, which encodes thymidylate kinase, Tmk. The tmk gene is located at the third position of a putative five-gene operon at 24.9 min on the E. coli chromosome, which comprises the genes pabC, yceG, tmk, holB, and ycfH. To avoid potential polar effects on downstream genes of the operon, as well as recombination with plasmid-encoded tmk, the tmk gene was replaced by the kanamycin resistance gene kka1, encoding amino glycoside 3′-phosphotransferase kanamycin kinase. The kanamycin resistance gene is expressed under the control of the natural promoter(s) of the putative operon. The E. coli tmk gene is essential under any conditions tested. To show functional complementation in bacteria, the E. coli tmk gene was replaced by thymidylate kinases of bacteriophage T4 gp1, E. coli tmk, Saccharomyces cerevisiae cdc8, or the Homo sapiens homologue, dTYMK. Growth of these transgenic E. coli strains is completely dependent on thymidylate kinase activities of various origin expressed from plasmids. The substitution constructs show no polar effects on the downstream genes holB and ycfH with respect to cell viability. The presented transgenic bacteria could be of interest for testing of thymidylate kinase-specific phosphorylation of nucleoside analogues that are used in therapies against cancer and infectious diseases.