205698a0Nature2054972196502136986990028-0836196510.1038/205698a0ukNatureNatureNATUREnatureNature is a weekly international journal publishing the finest peer-reviewed research in all fields of science and technology on the basis of its originality, importance, interdisciplinary interest, timeliness, accessibility, elegance and surprising conclusions. Nature also provides rapid, authoritative, insightful and arresting news and interpretation of topical and coming trends affecting science, scientists and the wider public./nature/journal/v205/n4972issueJournal homeArchiveCurrent issueAdvance online publicationPrivacy policySubscribeNature Publishing GroupCurrent issue205698a0Inhibition of Cell Division in Escherichia coli by Electrolysis Products from a Platinum Electrode
AU  - ROSENBERG, BARNETT
AU  - VAN CAMP, LORETTA
AU  - KRIGAS, THOMASBiophysics Department, Michigan State University, East Lansing.IN an investigation of the possible effects of an electric field on growth processes in bacteria, we have discovered a new and interesting effect. In E. coli, the presence of certain group VIIIb transition metal compounds in concentrations of about 1-10 parts per million of the metal in the culture medium causes an inhibition of the cell division process. The bacteria form long filaments, up to 300 times the normal length, which implies that the growth process is not markedly affected.Experiments were performed in a continuous culture apparatus in a chamber of special design. The nutrient medium was the chemically defined SC' medium of Roberts et al.1 used with 1 g of glucose/l and 0.026 g/1. MgCl2. The chamber was maintained at 37  [plusmn] 0.1  C, E. coli B and K-12 were used to inoculate the chamber. Compressed air was bubbled through a glass frit bubbler. Two half-cylindrical platinum mesh electrodes were built into the chamber. Once the bacterial population reached a steady state (about 24 h) as determined by turbidity measurements, it maintained its constant value ([plusmn] 2 per cent transmission for a 5-fold dilution of the effluent) for a period of 10 14 days. All experiments were performed within 10 days after inoculation.
The resistance of the chamber was approximately 6 ohms. An audio oscillator was the source of sinusoidal voltages of 50 105 c/s. This was fed into a conventional audio power amplifier and then to the platinum electrodes. The voltage and current were monitored with a dual beam oscilloscope.
On turning the electric field on at 1,000 c/s, 2 amp (peak to peak) through the chamber, the turbidity began to decrease after 1 h. (Platinum was chosen for the electrode material because of its well-known chemical inertness, and 1,000 c/s was chosen to eliminate electrolysis effects and electrode polarization. As we will show, both are mistaken ideas which led, via serendipity, to the effects described in this communication.) After 2 h, 'washout' of the bacteria was imminent and the voltage was turned off. After 8 h the population density had returned to its previous value. This process could be repeated. Microscopic examination of the effluent from the chamber showed that the E. coli ceased dividing (disappearance of 'waists') within 1 2 h after the current was turned on, and began to elongate. Within a few hours, all bacteria were in the form of long filaments, the length of which increased rapidly with time. For 1 2 h after the removal of the voltage, the cells continued to increase in length. Thereafter, however, the fairly uniform appearance of the bacteria under the phase contrast microscope changed. The darker material segregated into normal sized segments over the full length of the filament, leaving small, light grey gaps between segments, 'Waists' began to form in the gaps, and cell division occurred shortly thereafter over the length of the filament.
A frequency of the applied voltage of 500 c/s is very effective in causing filamentous growth. The efficiency decreases as the frequency increases, until at 6,000 c/s no filaments can be seen and no change in turbidity occurs over 24 h of application. From 6,000 c/s to 105 c/s no effects can be detected within 24-h trials. Increasing the current at a given frequency increases the size of the filaments appearing within a given time and causes a faster decline of the turbidity. We have found that oxygen is required to cause the effect; with nitrogen or helium bubbled through the cell, no effect of the electric field can be detected at all.
A number of physical and chemical agents are known to cause filamentous growth, that is, inhibit cell division but not growth. Among these are dyes such as methylene blue2 and penicillin3, near ultra-violet irradiation4, transfer to an unaccustomed medium, osmotic pressure changes, temperature changes5, and magnesium deficiency or excess6.
We have, with a variety of tests, eliminated the possibilities that ultra-violet light, temperature, pH, and magnesium concentration are involved in this effect of the electric field. We have also found that adaptive mechanisms and mutation effects are not involved.
It is possible that a chemical change is produced in the medium by electrolysis and that these new chemical products are the causative agents. To test this the nutrient was pumped into one chamber with electrodes in it (electrolysis chamber) before pumping it into a second chamber without electrodes which had been inoculated (bacterial chamber). The electric current was passed through the electrolysis chamber, No electric current was present in the bacterial chamber. If a new chemical species is created in the electrolysis chamber and it has a sufficiently long life it will appear in the bacterial chamber. The test conclusively showed that one or more long-lived new chemical species was created by the electric current and is responsible for the bacterial elongation. It caused elongation if oxygen was present in the electrolysis chamber regardless of what gas was bubbled through the bacterial chamber. Helium bubbled through the electrolysis chamber produced no elongation in the bacterial chamber. The active agent was not an insoluble gaseous product, since transferring the gas bubbling through the electrolysis chamber into the bacterial chamber caused no effect, if the medium was not also transferred.
A suspicion that an oxidizing agent was being generated by electrolysis was investigated by using a potassium iodide-starch test solution. Ordinary medium gave no reaction. The electrolysed medium gave a definite positive test the blue colour developing after about 5 min and going through yellow and orange intermediate states. The time course of the development of the concentration of this new agent in the electrolysed medium was strikingly parallel to the time course of the elongation processes. The concentration was proportional to the electric current. It decreased as the frequency increased from 500 to 6,000 c/s, at which frequency it was not detectable. All these measurements strongly implicate this new chemical as the causative agent.
Given the chemically defined 'C" medium, a number of oxidizing agents are conceivably created during electrolysis. The following ions or molecules were looked for with appropriate sensitive qualitative tests: C1O , ClO 2, CIO 3, CIO 4, H2O2, NH2OH and S2O1=8 None of these was detected. Some oxidizing agents were tested by inoculation directly into the bacterial continuous culture chamber, These were: H2O2, NH2OH, C1O , N2O (gas), NO (gas) and NO 2 all in appropriate quantities below that at which killing occurred. None of these chemicals caused elongation,
Each component of the C  medium was then made up in an appropriate concentration and electrolysed for 6 amp-h. The resultant solution was tested with the potassium iodide-starch solution. Negative results were obtained with the following: PO4; SO=; PO 4 + glucose; PO 4+ SO =+glucose; Na2SO4; Na2CO3, Positive tests were obtained with the following: (NH4)2SO4; (NH4)2CO3; NH4C1 and chlorides. A faint positive response was obtained with NaCl. Only solutions with chlorides showed the intermediate yellow to orange to blue transitions. It is known that platinum electrodes can be attacked by an acidified chloride solution during electrolysis (to form [PtCle]=) (ref. 7), We therefore suspected that a soluble platinum salt was the active agent. A solution of (NH4)2PtCl6, when tested with the potassium iodidestarch test, produced an exact duplication of the results with the electrolysed medium, if the platinum was present in a concentration of greater than 100 p.p.m. Inoculating the bacterial continuous culture chamber with a solution of (NH4)2PtCl6 so as to maintain a concentration in the chamber of Pt (IV) 10 p.p.m. over 2 h caused filaments to appear. This verified that the platinum salt was indeed an active agent. Quantitative tests for platinum IV in the electrolysed medium (SnCl2 test) showed that under our standard test condition (1,000 c/s, 3 amp) the steady-state concentration of the metal was 8 p.p.m. and was reached in about 2 h.
We have now tested a number of platinum and other group VIIIb compounds to determine the most effective anions and cations for this effect. The chemicals were inoculated so as to maintain a concentration of 8 p.p.m. of the metal in the nutrient medium of the cell for a period of 2 h. The effects fell into three categories: the bacteria were killed; there was no apparent change; a minimum of 20 per cent of the bacteria were forced into a filament form. These results are given in Table 1. The generalizations to be derived from these results are minimal. Platinum is not the only metal capable of inhibiting cell division, and in fact we have found that rhodium is as effective as platinum. We cannot at this time specify the relative effectiveness of the various metallic oxidation states or the effects due to various ligands and their spatial arrangements about the metal.
Table 1. EFFECTS OF GROUP VIII6 TRANSITION METAL COMPOUNDS ON BACTERIAL GROWTH (CONCENTRATIONS OF METAL IONS MAINTAINED FOR 
2 H AT 8 P.P.M. IN THE CONTINUOUS CULTURE CHAMBER) 
A. Caused bacterial 	B. Caused no 	C. Caused 
death 	change 	elongation 
CoClz 	[Co(NH3)6]Cl3 	K+,KH4+,H+ [PtCl6]= 
(NH4)8IrCl6 	K2Ir(NO,)6 	(NH4),PtBr6 
NiCla 	[M(NE3)6]C12 	CNH4)2PtI6 
(KH4)8OsCl6 	 	[Pt(en)3]Cl4 
CNH4)8PdCl4 	Cis and trans 	BhCla 
 	[B.h(en)iCl,]N08 	 
[Rh(NH3)5Cl]Cla 	 	(N'H4)3RhCl6 
PdCla 	 	[Bu(NH3)4C10H]Cl 
We believe the evidence is fairly conclusive that various group VIIIb transition metal ions in concentrations of 1 10 p.p.m. can inhibit cell division in E. coli while not apparently interfering with growth. Many questions arise from this work that we are now attempting to answer, such as: What is the mechanism of action of these metal ions ? Where is the locus of action in the bacterial cell ? How does the effect of these metal ions relate to the actions of the other causative agents of filamentous growth is there a weak link that all operate on ? Can these metal ions inhibit cell division in other bacteria, or cells ?
We thank Prof. Jaroslav Drobnik of Charles University, Prague, and Prof. Harold Sadoff of Michigan State University, for their advice, and Mr. Alan Stemler and Mr. Paul Vignola for their very able assistance with the experiments. This work was wholly supported by U.S. Public Health Service grant GM 10890.Roberts, , R. B., et al., Studies of Biosynthesis in E. coli, Carnegie Institution of Washington Publication 607 (Washington D. C., 1955).Ainley Walker, , E. W., and Murray, , W., Brit. Med. J., ii, 16 (1904).Gardner, , A. D., Nature, 146, 837 (1940).ChemPortHollaender, , A., J. Bact., 46, 531 (1943).ChemPortHinshelwood, , C. N., The Chemical Kinetics of the Bacterial Cell, 234 (Oxf. Univ. Press, London, 1946).Webb, , M., J. Gen. Microbiol., 3, 410 (1949).ISIChemPortKolthoff, , I. M., and Sandell, , E. B., Textbook of Quantitative Inorganic Analysis, third ed., 406 (Macmillan Co., New York, 1952).