364 M. HOLDEN 1961acetone), than found in earlier investigations. Mayer(1930) used a concentration of 66% for his deter-minations, but this is definitely inhibitory, thoughlese so for insoluble preparations than for soluble.The enzyme is remarkably stable before it

becomes soluble and the activity of soluble pre-parations is retained well in the cold. The optimumtemperature for enzyme action is much lower thanfor many enzymes, presumably because thetemperature at which the enzyme is inactivated iscomparatively low, particularly in the presence ofacetone.The rise in chlorophyllase activity when dark-

grown pea seedlings are transferred to light issimilar to that found by Hageman & Arnon (1955)for triphosphopyridine nucleotide-linked glycer-aldehyde phosphate dehydrogenase. Marcus (1960)induced the formation of triose phosphate dehydro-genase, without the development of chlorophyll, inkidney-bean leaves by brief exposure to red light.Booth (1960) found that a carotene-destroyingenzyme system was not normally present in potatoesand carrots, but exposure to light leading tochlorophyll formation in the surface layers alsocaused the enzyme to appear.

SUMMARY

1. Soluble chlorophyllase preparations weremade from sugar-beet leaves, which have thehighest activity of all species tested.

2. For the partly purified enzyme the optimumconditions for enzyme action are pH 7-7 at 250 andan acetone concentration of 40 %.

3. When dark-grown pea seedlings with verylow chlorophyllaae activity were transferred todaylight the activity rose in 48 hr. to a level com-parable with that of light-grown seedlings.

REFERENCES

Ardao, C. & Vennesland, B. (1960). Plant Phy8iol. 85, 368.Amnon, D. I. (1949). Plant Phy8iol. 24, 1.Booth, V. H. (1960). J. Sci. Fd Agric. 11, 8.Estermann, E. F., Conn, E. E. & McLaren, A. D. (1959).

Arch. Biochem. Biophy8. 85, 103.Fischer, H. & Lambrecht, R. (1938). Hoppe-Seyl. Z. 253,

253.Gage, R. S. & Aronoff, S. (1956). Plant Phy8iol. 31, 477.Hageman, R. H. & Amnon, D. I. (1955). Arch. Biochem.

Biophy8. 57, 421.Krossing, G. (1940). Biochem. Z. 305, 359.Marcus, A. (1960). Plant Phy8iot. 35, 126.Mayer, A. M. & Friend, J. (1960). Nature, Lond., 185, 464.Mayer, H. (1930). Planta, 11, 294.Peterson, P. D. & McKinney, H. H. (1938). Phytopath-

ology, 28, 329.Sironval, C. (1954). Phy8iol. Plant. 7, 523.Weast, C. A. & Mackinney, G. (1940). J. biol. Chem. 133,

551.Willstatter, R. & Stoll, A. (1910). Liebig8 Ann. 378, 18.Quoted in Inye8tigation8 on Chlorophyll (1928). Transl.by Schertz, F. M. & Merz, A. R. Lancaster, Pa.: SciencePress Printing Co.

Biochem. J. (1961) 78, 364

Pyridoxamine Phosphate-Oxidase and Pyridoxal Phosphate-Phosphatase Activities in Escherichia coli

BY J. M. TURNER* AND FRANK C. HAPPOLDDepartment of Biochemi8try, Univermity of Leeds

(Received 29 April 1960)

Beechey & Happold (1955) first demonstratedthe conversion of pyridoxamine phosphate into thealdehyde form by cell-free extracts of Escherichiacoli. The nature of the product was establiBhed byits ability to activate apotryptophanase, anenzyme also present in the aqueous extracts.Optimum conditions for enzyme assay werestudied, and simple kinetic studies were carried outby spectrophotometric methods, by Beechey &Happold (1957), who found that E. coli extracts

catalysed the disappearance of pyridoxal phos-phate and that light-absorption at the wavelengthcharacteristic of pyridoxamine phosphate in-creased during the reaction. The reaction involvedthus appeared to be reversible, although only theconversion of pyridoxamine phosphate into thepyridoxal ester required Mg2+ ions for maximumactivity.The possibility that more than one enzyme was

involved was recognized, but the apparent re-versibility of the reaction, and other data, led theseworkers to assume that a transamination reactionwas involved. Unlike Meister, Sober & Tice (1951),

* Present address: Department of Chemistry andLawrence Radiation Laboratory, University of California,Berkeley, California, U.S.A.

Vl78ENZYME ACTIVITY IN EXTRACTS OF E. COLIwho presented evidence for transamination betweenpyridoxamine phosphate and pyruvic acid inextracts of Ciostridiutm welchii, Beechey & Happold(1957) were unable to identify amino-groupacceptor-donors in their system.The present paper suggests that the apparent

interconversion of pyridoxam,ine phosphate andpyridoxal phosphate in extracts of E. coli is due tothe presence of two separate enzymes. Theseenzymes have been identified as pyridoxaminephosphate oxidase and a pyridoxal phosphatephosphatase. Pogell (1958) has reported theexistence of similar enzyme activities in mam-malian tissues, also in yeast and Streptococcusfaecali.

EXPERIMENTAL

Material8The preparation and storage of cell-free extracts of

acetone-ether-dried E. coli were carried out as describedby Beechey & Happold (1957). Glutamic-aspartic trans-aminase was prepared from heart muscle by the method ofCammarata & Cohen (1951), the fraction 'Ppt. WFT-II'being used. An alkaline-phosphatase preparation (In-testinal Phosphatase) was obtained from Armour Labora-tories, Chicago, Ill., U.S.A., and a soluble enzyme reagent(Polidase-S), known to contain acid phosphatase, wasobtained from Schwartz Laboratories Inc., New York,N.Y., U.S.A. Pyridoxal phosphate and pyridoxaminephosphate were from Hoffmann-LaRoche andCo. Ltd., ketoacids from British Drug Houses Ltd. and adenosine tri-phosphate (ATP) from L. Light and Co. Ltd.

Assay of vitamin B. compounds andassociated enzyme activities

The assay of vitamin B. compounds was carried outspectrophotometrically with a Unicam spectrophotometer,model SP. 500. Preliminary experiments were carried outto determine maximum absorption wavelengths forpyridoxamine phosphate and pyridoxal phosphate, alsopyridoxal under the conditions in which these compoundswere to be assayed.

Composition of standard reaction mixture. For intercon-versions of vitamin B6, compounds, reactions were carriedout in a total volume of 4 0 ml. The standard compositionwas: buffer, 1.0 ml.; 0-02M-MgSO4, 0 5 ml.; concentratedcell extract, approx. 10-0 mg. of protein/ml., 2-0 ml.:vitamin B*, compound (200 pmg./ml.) 0-5 ml., i.e. approx.04 ,umole of pyridoxamine phosphate and pyridoxalphosphate. Water replaced substrate in the blank andreplaced MgSO4 solution when pyridoxal phosphate wasused as substrate. When the effect of various additions wasinvestigated, cell extract was made up in buffer, and theadditions were contained in 1.0 ml. Where necessary,reaction-mixture components were adjusted to the pH ofthe reaction.

Measurement of the conversion of pyridoxamine phosphateinto the aldehyde form. The buffer used was 0 05M-NaHCO8-Na,C0., pH 10.1 (Delory & King, 1945). Blank, andreaction mixture without substrate, were equilibrated at370 in 1 cm. quartz cells, the latter being placed in a

thermostatically controlled cell-holder. Addition of sub-strate to the equilibrated reaction system, with stirring,was timed. Extinction values were then read alternativelyat 312*5 and 390 mj&, at timed intervals over the period ofthe reaction. Overall changes in extinction were calculatedafter extrapolation of readings to zero time.As the thermostatically regulated cell-holder had posi-

tions for only two cells, the complete reaction mixture wasused as a blank when the activity of reaction mixture plusaddition was of interest. In this case inhibition or activa-tion was related to enzyme activity measured immediatelybeforehand as described above. When inhibition was ob-served, readings were made with reaction mixture plusinhibitor in the blank position.

Alternatively, reaction mixtures minus substrate weremade up in thin-walled test tubes and equilibrated in awater bath at 37°. Substrate was then added at timedintervals to all but the controls, and at the end of the re-action 1.0 ml. of 2-5N-NaOH was added to all tubes.Substrate was then added to controls. Pyridoxaminephosphate is stable during incubation under the above-mentioned conditions in the absence ofenzyme preparation.Precipitated protein was removed by centrifuging at10 000 rev./min. for 30 min. (12 000g at the bottom ofthetube). The extinction of tube contents was read against thecorresponding blank in order to assay the product formed.In later work, the extinction of the control was read againstthe reaction mixtures in order to assay the substrateutilized. The wavelengths used were the wavelengths ofmaximum absorption (A.), determined for pyridoxaminephosphate and pyridoxal phosphate in 0-5x-NaOH, i.e.307 5 and 390 m. respectively.

Measurement of the dephosphorylation of pyridoxal phos-phate. The same spectrophotometric methods were used asdescribed for the measurement of the conversion of pyrid-oxamine phosphate into the aldehyde form at pH 10.1.Here, however, product formation was measured by theincrease in extinction at 300 mju, i.e. A,. for pyridoxal atpH 10.1.Methods used for converting measured changes in

extinction into changes in vitamin B.-compound concen-tration are described in the Results section. Specificactivities of the preparations are expressed in terms ofpmoles/30 min./mg. of protein/ml. of reaction mixture.

Isolation and identification of vitamin B. compounds.A Dowex-1 formate column, described for the separation ofpyridoxine metabolites by Rodwell, Volcani, Ikawa &Snell (1958), was used to isolate vitamin B. compoundsfrom reaction mixtures. Deproteinization was by theaddition of 100% trichloroacetic acid. Eluate sampleswere examined spectrophotometrically at two pH values,and vitamin B, compoundswere detected by the blue colourproduced with 2:6-dichloroquinonechloroimide reagent(Snyder & Wender, 1953). Identification was carried outby paper chromatography (Rodwell et al. 1958; Fasella &Baglioni, 1956).

Measurements of acid- andalkaline-phosphatase activities

The same composition of reaction mixture was used asfor the spectrophotometric assay for interconversion ofvitamin B. compounds. The protein concentration of E.coli extracts used was 2 mg./ml. and that of phosphatasepreparations from other sources was usually 1 mg./ml.

Vol. 78 365

J. M. TURNER AND F. C. HAPPOLDI)E. coli extracts were dialysed overnight against a largevolume of the appropriate buffer, before use, to removetraces of inorganic phosphate. In all cases the substratewas disodium phenyl phosphate (0-1M). Deproteinizationwas accomplished by addition of 100% (w/v) trichloro-acetic acid until no further precipitation ofprotein occurred,and, after centrifuging to remove protein, 1 ml. of super-natant was used for phosphate assay. The method ofGomori (1941-42) was used to determine inorganic phos-phate.

Measurements of tryptophanaseactivity and indole assay

These were carried out as described by Beechey &Happold (1957).

suggestion that the reaction involved was a trans-amination, the effect of various potential amino-group acceptors on the activity of cell extracts wasinvestigated. Glyoxylate, shown by Metzler,Ikawa & Snell (1954) to function as an amino-group acceptor in a number of transaminationsystems, inhibited the above-mentioned reactionwhen used as the sodium salt at a number of con-centrations (Table 1). Dialysis of extracts againstNaHCO3-Na2CO3 buffer, pH 10-1, decreased theiractivity by between 41-0 and 42-5%. This activityloss was calculated by comparison with the activityof an extract sample allowed to stand under the

RESULTS

Conversion of pyridoxatnine phosphateinto pyridoxal phosphate

Progress of the reaction and product identity.Spectrophotometric data showing the disappear-ance of pyridoxamine phosphate used as substrate,and the formation of product, are shown in Fig. 1.The absorption spectrum of the product, with peakabsorption at 390 mu (Fig. 2), confirms previousenzymic evidence on the identity of the product aspyridoxal phosphate (Beechey & Happold, 1957).Owing to the high light-absorbency of the cellextracts used, it was not possible to make measure-ments at wavelengths less than those indicated.

Effect of variou8 compounds, also dialy8si, onenzyme activity. In view of the previously reported

B315.5 E3900.54 0-Il

0-2 X

0'300 320 340 360 380 400 420

Wavelength (mg)Fig. 2. Absorption spectrum of reaction product withpyridoxamine phosphate as substrate. Reaction conditionswere as described in Fig. 1. A\, Difference spectrum ofproduct; *, pyridoxamine phosphate (approx. 25jug./mJ.)incubated under the same conditions but without enzyme;0, pyridoxal phosphate (approx. 25 ug./ml.). All spectrawere plotted after an incubation time of approx. 120 min.

0'44. . . . .0 10 20 30 40 50 60

Time (min.)

Fig. 1. Conversion of pyridoxamine phosphate into pyrid-oxal phosphate. Mixtures containing 1-0 ml. of 0-05M-NaHCO3-Na,C03 buffer, pH 10-1, 0-5 ml. of 0-02M-MgSO4,2-0 ml. of cell extract (10 mg. of protein/ml.) and 0-5 ml. ofsubstrate (200jsg./ml.) were incubated at 37°. At theindicated times, values of E at 312-5 mp (@) and 390 m,u(0) were read against a control in which water replacedsubstrate.

Table 1. Effect of sodium glyoxylate on the con-version of pyridoxamine phosphate into pyridoxalphosphate

Reaction mixtures contained: 1-0 ml. of 0-05M-NaHCO3-Na,CO3 buffer, pH 10-1; 0-5 ml. of sodium glyoxylatesolution in buffer; 0-5 ml. of 0-02M-MgSO4; 2-0 ml. of cellextract (10 mg. of protein/ml.); 0-5 ml. of pyridoxaminephosphate (200 pg./ml.; 0-4 pmole). Buffer replacedsodium glyoxylate solution in the control, and waterreplaced substrate in the blank. At the end of incubationfor 30 min. at 370, 0-5 ml. of 2-5N-NaOH was added, pre-cipitated protein was removed by centrifuging and E at390 m,u was read against the blank, in which cell extractwas omitted until after the addition of NaOH.

Concn. ofsodium

glyoxylate(mm)

0-250-631-25

Increase in ActivityE at 390 mju (%)

0-046 1000-037 800-039 850-038 83

1961366

ENZYME ACTIVITY IN EXTRACTS OF E. COLI

same conditions of buffer concentration and pH,but without dialysis; this sample was finally dilutedto the same protein concentration as the dialysedpreparation. The effect of a number of potentialamino-group acceptors on the activity of dialysedcell extracts is shown in Table 2. With such extractsglyoxylate did not cause inhibition. Although theuse of boiled cell-extract concentrates suggestedthat activity could be partially restored to dialysedcell extracts, the results were of little significance.Concentrated diffusate from the extract, observedto have an absorption maximum at 260 mA atpH 7-0 and to exhibit green fluorescence in u.v.light, did not restore activity. The diffusate wasobtained by dialysis of extracts against distilledwater and was then concentrated by freeze-drying.

Table 2. Effect of potential amino-group acceptorson the conversion of pyridoxamine phosphate intopyridoxal phosphate

Reaction mixture and assay procedure details were as

described for Table 1, except that dialysed cell extract wasused where appropriate, and incubation was for 90 min.

Additions todialysed extract

(0-02 mM)

Pyruvatea-OxoglutarateOxaloacetateGlyoxylateNon-dialysed extract(no additions)

Increase inE at 390 m,u

0-0360-0200-0030-0080-0360-061

Activity(%)

583551358100

Table 3. Effect of anaerobic conditions on the con-

version of pyridoxamine phosphate into pyridoxalphosphate

The reaction mixture contained in the main arm ofThunberg vessels: 1-0 ml. of 0-05M-NaHCO3-Na2CO3buffer, pH 10-1; 0-5 ml. of 0-02M-MgSO4; 2-0 ml. of un-

dialysed cell extract (approx. 10 mg. of protein/ml.).Either 0-5 ml. ofwater or pyridoxamine phosphate (200 jAg./ml.; 0-4 ,umole) was contained in the side arm. Afterrepeated evacuation, and flushing with either air or

nitrogen, vessels and contents were equilibrated at 37°.Reactions were started by tipping side-arm contents intothe main arm. Incubation was for 90 min. Reactions were

stopped by the addition of NaOH, and E at 390 mju wasread as described for Table 2.

Conditionsof reaction

(atmosphere)AirNitrogenAirNitrogenAirNitrogen

Increase inE at 390 mu

0-0530-0110-0490-0150-0620-018

Activity(%)100191003010029

Nature of the reaction. Since the conversion ofpyridoxamine phosphate into the aldehyde formcould be partly due to an oxidation process, theeffect of anaerobic conditions was investigated(Table 3). At least 70-80% of the pyridoxal phos-phate was formed by an oxidation process.Previous indications suggested that the processmight require a diffusible cofactor, but neitherriboflavin nor riboflavin phosphate (0-01 mm).restored activity. The antimalarial mepacrine[6-chloro-9-(-4-diethylamino-1-methylbutylamino)-2-methoxyacridine], however, at a concentration of0-53mx inhibited the activity ofundialysed extractsby about 20%. Higher mepacrine concentrationscould not be employed at pH 10-1 owing to its in-solubility. Various flavin-dependent oxidases havebeen shown to be inhibited by this compound(Hellerman, Lindsay & Bovamick, 1946; Nygard &Thellwell, 1955; Doisy, Richert & Westerfield, 1955).

Method evolved to calculate changes insubstrate and product concentration

In order to correlate observed changes inextinction with changes in substrate and productconcentration, certain properties of the respectiveabsorption spectra were found to be of importance.Thus, at pH 10-1, whereas pyridoxamine phos-phate does not absorb light at wavelengths above370 mi, pyridoxal phosphate absorbs both at390 mi (Am) and 312-5 m,i. Thus during theenzymic oxidation of pyridoxamine phosphate toits aldehyde form, any decrease in E observed at312-5 mu would be partly masked by an increasedue to pyridoxal phosphate formation. Generalformulae were therefore constructed, relatingobserved E (I cm. light path) to the concentrationof both pyridoxamine phosphate (Pam.P) andpyridoxal phosphate (Pal.P) in terms of molecularextinction coefficients. During substitution ofconstants into such formulae, complications due tooverlapping of absorption spectra at the wave-lengths used are automatically allowed for. Thegeneral formulae are:

6390 Elcm - Ep12-5 Elem.

[Pam-P] = -390 312-6- 3123590Ep&l.p Ep&m.p EPal.P EPam.Pc312-5 Eicm._ 90 1lcm.

[Pal.fl _ Pa._ 390 (Pam . 812-5*Pal.P = 812-5 390 - 390 312-5Pam. Pal.P EPM.P EPl.P

It was found that, at pH 10-1,312-5 = 1000; 39O = 6000;

(312- = 5700; e390 = 0.6PAMP p&MpResults supporting these values are included inTable 4, lines 1 and 5. Substituting these values inthe general formulae,

[Pam.P] = (1-75 El'm- 0-29 ElC) x 10-4M[Pal.P] = 1-67 El *- x 10-4M.

Vol. 78 367

J. M. TURNER AND F. C. HAPPOLD

In order to check the validity of the formulae,mixtures of standard solution of pyridoxaminephosphate and pyridoxal phosphate were preparedat two concentrations. The known concentrationvalues for each component were then comparedwith values calculated from the values of Eobserved; good agreement was obtained.

Stoicheiometry of the reaction. By following theprogress of the reaction now deduced to be anoxidation of pyridoxamine phosphate, with time,as shown in Fig. 1, it was found that the relation-ship between the amounts of substrate utilized andproduct formed was variable. This non-equivalencewas particularly marked when extracts of E. colicells grown on a medium including 0-2% of glucosewere used (Fig. 3). A quantitative examination ofthe results illustrated in Figs. 1 and 3, with thespecially derived formulae, suggested a more rapidremoval of pyridoxal phosphate from the systemwhen extracts prepared from cells grown on glucosecontaining media were used. The calculated results

0-66 X

0-64

0-62

0.60

0.58

0-56

0.54

0^52

*0.12

.0-09

* 0-07 0

-0-05 0O 0

- bb.onwY m~~~~~~~sX-. .a .*

0 10 20 30 40 50 60 70 80 90 100Time (min.)

Fig. 3. Conversion of pyridoxamine phosphate into pyrid-oxal phosphate by extracts of cells grown on a mediumcontaining 0-2% of glucose. Reaction conditions were asdescribed in Fig. 1. *, Measured at 312-5 mp; 0,measured at 390 m&.

Table 4. Comparison of known and calculated concentrations of pyridoxatmine phosphateand pyridoxal phosphats

Standard solutions of pyridoxamine phosphate (Pam.P) and pyridoxal phosphate (Pal.P) were prepared, eachat 0x12 mm and at one-half of this conoentration, in 12-5 M-NaHCOO-Na2CO8 buffer, pH 10-1. Mixtures of thesesolutions were made up as indicated, and values of E (1 cm. light-path) were measured at 312-5 and 390 mp.Concentration values were calculated as described in the text.

Known concn. of mixturecomponents (0-1 mm)

Pam.P Pal.P1-20 00-90 0-300-60 0-600-30 0-900 1-200-60 00-45 0-150-30 0-300-15 0-450 0-60

Content ofPal.P

in mixture(%)0

2550751000

255075100

312-5 mp0-6850-5400-4000-2650-1200-3400-2700-2000-1350-070

390 m,u00-1800-3600-5350-72000-0900-1800-2670-360

Calc. conen. of mixturecomponents (0-1 mm)

A

Pam.P Pal.P1-20 00-890 0-3000-595 0-6000-305 0-8900 1-200-5950-4450-2800-1590-017

00-1500-3000-4450-600

Table 5. Apparent stoicheiometric relationships between substrate and productduring oxidation of pyridoxamine phosphate

The reaction mixture is described in the Experimental section of the paper; 5 mg. of protein was present/ml.

Calc. changes in concn.(0-1 mm)

f -

Pam.P(a)00-100-150-1800-190-24

Pal.P(b)00-090-130-1600-070-08

Source ofdata

Fig. 1(Extract 1)

Fig. 3(Extract 2)

Reactiontime(min.)

030609003060

Ratio:A (a)/A (b)

1-111-151-12

2-703-00

1961368

E312.1 E310

.E I cm.MJA

r.. AA

ENZYME ACTIVITY IN EXTRACTS OF E. (COLIare given in Table 5. From Table 5, the specificoxidase activities of extracts 1 and 2 can be calcu-lated as 2*0 and 3 81am-moles of pyridoxaminephosphate/30 min./mg. of protein respectively.

Evidence for the removal of pyridoxal phosphatefrom reaction systems- was also obtained whenpyridoxamine phosphate was used as substrateand added apotryptophanase preparations wereused to assay pyridoxal phosphate fornation. Themethod used was that described by Beechey &Happold (1957). With either E. coli extracts ormammalian glutamic-aspartic-transaminase pre-parations it was found that an initial rise in thepyridoxal phosphate formation was followed by arapid fall in the ability of samples to activate theapotryptophanase system.

In early experiments, designed to determine theequilibrium position for the reaction convertingpyridoxamine phosphate into pyridoxal phosphate,it was observed that the initial decrease in E at312 5 mp was followed after several hours by aslow rise. A simultaneous slow decrease in E at390 mp was also observed. These results suggested,in retrospect, the conversion of pyridoxal phos-phate into some compound having an absorptionat 312-5 m,A similar to, but not identical with, thatdue to pyridoxamine phosphate.

Conversion ofpyridoxal phosphate intofree pyridoxalEstablishment of the nature of the reaction.

Pyridoxal phosphate is readily hydrolysed to thefree base and inorganic phosphate (Peterson, Sober& Meister, 1953), and it appeared likely that thedisappearance of this compound from enzymic-reaction systems could be due to enzymic hydro-lysis.

In order to distinguish between pyridoxal forma-tion and the production of pyridoxamine phos-phate as suggested by the data of Beechey &Happold (1957), the light-absorptions of reactionmixtures with pyridoxal phosphate as substratewere measured at 300, 312-5 and 390 m,, over aperiod of approx. 40 min. at 37°. At pH 10-1, thewavelengths quoted characterize the peaks ofmaximum absorption for pyridoxal, pyridoxaminephosphate and pyridoxal phosphate respectively.The results of one experiment are shown in Fig. 4.Absorption spectrum of the reaction product in asimilar experiment is shown in Fig. 5, together withthose of pyridoxal and pyridoxal phosphate incu-bated under the same conditions but in the absenceof enzyme preparation. It can be seen that the cellextracts catalysed the breakdown of pyridoxalphosphate to the free base. As the extinction ofpyridoxal at 312-5 ml& is appreciable comparedwith that at 300 mp, the apparent formation ofpyridoxamine phosphate previously reported isthus explained.

24

Column chromatography of reaction mixtures inwhich pyridoxal phosphate had been used as sub-strate and where extracts had been prepared fromcells grown on media, with or without addedglucose, failed to show any pyridoxamine phos-phate or free pyridoxamine. Only pyridoxal wasdetected in eluate samples.

Effect of various additions. In order to examinethe possibility that the dephosphorylation ofpyridoxal phosphate might be due to a reversal ofthe pyridoxal-kinase reaction (Hurwitz, 1953)

B390 E0 & 31215

0-6r0-4

Time (min.)Fig. 4. Conversion of pyridoxal phosphate into freepyridoxal. Reaction conditions were as described in Fig. 1,except that pyridoxal phosphate (approx. 200 pg./ml.) wasthe substrate, and water replaced MgSO, solution. 0,Measured at 312-5 m,u; A, measured at 300 m,u; 0,measured at 390 mp.

0-5

0-4

E 0.30-2

0.1

O290 310 330 350 370 390 410

Wavelength (mP)Fig. 5. Absorption spectrum of reaction product withpyridoxal phosphate as substrate. Reaction conditionswere as specified in Fig. 4; reaction time, 5 hr. A, Differ-ence spectrum of product; 0, pyridoxal phosphate,approx. 25 pg./ml.; * pyridoxal, approx. 25jtg./ml.Both pyridoxal and Ipyridoxal phosphate were incubatedunder the same conditions as when pyridoxal phosphatewas used as substrate,- except that no enzyme preparatioxnwas present.

Bioch. 1961, 78

Vol, 78 369

-41

J. M. TURNER AND F. C. HAPPOLD

rather than a hydrolysis, the effect of ATP, at afinal concentration of 5 mm, on the dephosphoryl -ation of pyridoxal phosphate was investigated. Nounequivocal evidence was obtained that ATPaffected the rate of dephosphorylation. Spectro-photometric results were complicated by the factthat the E. coli extracts appeared to catalyse adecrease in E, due to ATP, at 300 m,u, i.e. the wave-length used to assay pyridoxal formation.. Underthe conditions previously employed for the dephos-phorylation of pyridoxal phosphate, no pyridoxal-kinase activity could be detected spectrophoto-metrically with pyridoxal and ATP as substrates,and with E at 390 m,u as the criterion of pyridoxalphosphate formation. This confirms the finding ofBeechey & Happold (1957), who used enzymicmethods for the detection of pyridoxal phosphate.A study of the effect of Mg2+ ions at 2-5 mM on

the dephosphorylation of pyridoxal phosphateshowed a consistent 8% increase in activity.Ethylenediaminetetra-acetate, however, at a con-centration of 0-62 mm, caused no inhibition whennon-dialysed cell extract was used.With either dialysed or non-dialysed E. coli

preparations, no effect of 00O1M-NH4+ ions on therate of pyridoxal phosphate utilization could bedetected. This contrasts with the apparent 60%increase in rate observed by Pogell (1958), whoreported a pyridoxal phosphate phosphatase to bepresent in mammalian-liver preparations, andsuggested the formation of an imide type of inter-mediate in the presence of NH4+ ions.

Phosphata8e activitite in extract8 of Escherichiacoli, and the effect of phosphatase preparations onpyridoxal phosphate. The fact that pyridoxalphosphate was dephosphorylated by E. coliextracts posed the question whether the enzymeinvolved was one specific for this substrate orpossibly an alkaline phosphatase of broad speci-ficity.With disodium phenyl phosphate as substrate,

and by measuring inorganic phosphate formation,cell extracts were found to have little activity atpH 10.1. At lower pH values, however, activityincreased to an optimum at pH 5-3. Maximumactivity was observed with 0-2 mM-Mg2+ ions. Itthus appeared that cell extracts contained an acidphosphatase. The respective phosphatase activitiesat pH 5-3 and pH 10 1 were 12-20 and 0-27 anolesof phosphate/30 min./mg. of protein.With an alternative source of acid-phosphatase

activity, i.e. Polidase-S, no dephosphorylation ofpyridoxal phosphate could be detected at pH 10 1.When an alkaline phosphatase was used, i.e.Intestinal Phosphatase, rapid dephosphorylationof pyridoxal phosphate occurred, even at roomtemperature. These results suggest that, althoughE. coli extracts contain an acid phosphatase, this

enzyme is not responsible for the breakdown ofpyridoxal phosphate at pH 10.1, but that analkaline phosphatase present in smaller amounts isinvolved. Attempts to demonstrate pyridoxalphosphate-phosphatase activity by assay of in-organic phosphate formed were inconsistent,probably owing to the instability of the substrateunder the conditions of the assay procedure.Method evolved to calculate changes in substrate

and product concentration. General formulaerelating observed changes in E to the respectivemolar concentrations of the two components werederived for mixtures of pyridoxal phosphate andpyridoxal. Thus:

300 l1 cm. _ 390 l1 cm.9Pal.Pal. (Pal. -'300

[Pal.P ] = -300-390 - 390 300(Pal. Epal.p 4EPal. fPal.P

30lm._ 300 171 cm.

PaalP]=PLl-300 EPaL.P " 390[Pal.] = 4E39 El m -EpE390 300 300 390

9paL.p 9pal. -pal.P (Pal.

It was found that, under the conditions used forenzymic dephosphorylation of pyridoxal phos-phate, E304 = 5000; 4390 = 1600; E3oop = 1080;

9,al.P = 6000. Thus the above formulae may besimplified to:

[Pal.P] = (1-77 E'cm--0.56 El cm-) x 10-4M[Pal.] = (2-13 El cm--0 40 Elc%m) x 10-4M.

Applying these relationships to the data illustratedin Fig. 4, the specific activities for the dephos-phorylation of pyridoxal phosphate by the cellextract obtained were 10-6 and 9-5,um-moles ofpyridoxal and pyridoxal phosphate respectively/30 min./mg. of protein.

Relative pyridoxamine phosphate-oxidase andpyridoxal pho&phate-pho&phata8e activities

The relative activities of pyridoxamine phos-phatase oxidase and pyridoxal phosphatase weredetermined, with extracts prepared from a numberof different batches of E. coli cells. The cells weregrown with or without glucose. Methods used forthe calculation of substrate concentration changeswere those described above. In each case, theamount of substrate utilized was used as thecriterion of activity measurements (Table 6). Thepyridoxal phosphate-phosphatase activity of cellsgrown in a medium containing glucose is consider-ably in excess of the pyridoxamine phosphate-oxidase activity.No pyridoxamine phosphate-phosphatase activ-

ity could be detected by the assay of inorganicphosphate production in any of the extractsexamined.

DISCUSSION

The experimental measurement of pyridoxaminephosphate-oxidase and pyridoxal phosphate-phos-phatase activities is complicated by the presence of

370 1961

ENZYME ACTIVITY IN EXTRACTS OF E. COLI

Table 6. Relative pyridoxamine pho8phate-oxidae and pyridoxal phoaphate-pho8phata8e activitime

Reaction mixture composition and procedure were as described for Fig. 1, except that cell-extract concentra-tion was 1.0 mg. of protein/ml. (final concn.), and MgSO4 solution was replaced by water when Pal.P was used assubstrate. Phosphatase activity was determined immediately after oxidase activity, with the same sample ofcell extract.

Growth conditionsfor E. coli

Without glucose

With 0-2% of glucose

Relative specific activities*

Pam.P Pal.Poxidase phosphatase

(a) (b)0-28 0-200-12 0-170-48 0-980-20 0-79

Ratio:(b)/(a)0-721-422-053-95

* 10-2 mole/30 min./mg. of protein/ml. of reaction mixture.

both enzymes, in variable proportions, in the cellextracts used. If enzymic or chemical methods are

used for the assay of pyridoxal phosphate whenpyridoxamine phosphate is used as substrate, thenthe possible simultaneous breakdown of pyridoxalphosphate is difficult to allow for. On the otherhand, spectrophotometric methods allow theprecise determination of stoicheiometric relation-ships between substrates and products. Spectro-photometric methods for the simultaneous deter-mination of pyridoxamine phosphate, pyridoxalphosphate and pyridoxal concentrations in reactionmixtures are at present under development inthese laboratories. Such methods allow exactmeasurement of pyridoxamine phosphate-oxidaseactivities even in the presence of pyridoxal phos-phate phosphatase. The relative activities ofpyridoxamine phosphate oxidase and pyridoxalphosphate phosphatase in biological systems is ofintrinsic interest. Although Wada et al. (1959) claimthat phosphatase activity towards pyridoxal phos-phate can be removed from crude extracts of liverby precipitation at pH 5-0, Pogell (1958) was unableto remove this activity completely by such treat-ment.

In the present report, no explanation is given forthe decrease in pyridoxamine phosphate-oxidaseactivity during dialysis of E8cherichia coli extracts.In view of the linking of oxidase and pyridoxalphosphate-phosphatase activities, it is possiblethat the loss of some factor affecting the latter isresponsible.

It has been reported that pyridoxamine phos-phate is the predominating member of the vitaminB. group present in the cell (Rabinowitz & Snell,1947; McNutt & Snell, 1948), and the possiblesignificance of the enzymic conversion of pyrid-oxamine phosphate into pyridoxal phosphate inthe control of amino acid metabolism has beenpointed out (Beechey & Happold, 1957; Pogell,1958). Pyridoxamine phosphate-oxidase activityhas been shown to be responsible for the indirect

activation of tyrosine decarboxylase (Pogell, 1958),and is now thought to effect the observed activa-tion by pyridoxamine phosphate of tryptophanase(Beechey & Happold, 1957) and kynureninase(Saran, 1958). The role played by pyridoxal phos-phate in a large number of enzyme reactions hasrecently been reviewed by Snell (1958). The inhibi-tion of pyridoxamine phosphate oxidase by a-oxoacids commonly involved in transamination re-

actions may be part of the mechanism used by cellsfor metabolic control. Of similar significancewould be those factors affecting the activity ofpyridoxal phosphate phosphatase. Both Pogell(1958) and Wada et al. (1959) have detected phos-phatase activities for pyridoxal phosphate inrabbit-liver preparations. In the control of aminoacid metabolism by coenzyme destruction, thestrength of binding between pyridoxal phosphateand the respective apoenzymes would also be ofimportance. A number of pyridoxal phosphate-dependent enzymes lose their activity on agingunder various conditions but can be reactivated bythe addition of synthetic coenzyme (Kallio, 1951;Azarkh & Gladkova, 1952; Happold & Struyven-berg, 1954). The disappearance of bacterial lysinedecarboxylase after induction (Mandelstam, 1954)also implicates coenzyme destruction (Pardee,1958). If local intracellular control of either or bothof the oxidase and phosphate enzymes dealt withhere is possible, then differential regulation ofindividual enzymes or groups of enzymes involvedin amino acid metabolism would occur. Horiuchi(1959) has recently reported that E8cherichia colicells grown on a phosphate-deficient medium ex-

hibit an alkaline-phosphatase activity approxi-mately 100 times as great as that of cells grown on

a complete medium. In view of the fact thatalkaline phosphatase rapidly destroys pyridoxalphosphate, these findings suggest that the aminoacid economy of cells grown on deficient media maybe affected by the inactivation of enzymes respon-sible for amino acid breakdown.

24-2

Cellsample

1234

Vol. 78 371

372 J. M. TURNER AND F. C. HAPPOLD 1961

SUMMARY

1. Evidence is presented for the presence ofpyridoxamine phosphate oxidase and a pyridoxalphosphate phosphatase in aqueous extracts ofE8cherichia coli.

2. Spectrophotometric methods have been de-veloped for the precise measurement of substrateand product concentrations in both enzymesystems. Specific activities of the order of 10-2 tu-mole/30 min./mg. ofprotein/ml. ofreaction mixtureare reported.

3. Pyridoxamine phosphate oxidase was in-hibited by dialysis, and further inhibition wascaused by pyruvate, ao-oxoglutarate and oxaloace-tate at a concentration of 0-02 mm. Inhibition ofactivity by mepacrine suggests that a flavincoenzyme is involved.

4. Pyridoxal phosphate phosphatase was alsofound to be highly active in mammalian 'alkaline'-phosphatase preparations. The E8cherichia coliextracts used exhibited high 'acid '-phosphataseand low 'alkaline '-phosphatase activities whendisodium phenyl phosphate was used as substrate.No dephosphorylation of pyridoxamine phosphatecould be detected.

5. The presence of glucose in Escherichia coligrowth medium increased the relative amount ofphosphatase compared with oxidase activity.

This work has been carried out with the support of theMedical Research Council. One of us (J.M.T.) is alsograteful for a Medical Research Council training scholar-ship. Pyridoxamine phosphate, pyridoxal and pyridoxalphosphate were gifts from Roche Products Ltd.

REFERENCES

Azarkh, R. M. & Gladkova, V. N. (1952). Dokl. Akad.Nauk, SSS1, 85, 173. Cited in Chem. Ab8tr. (1952), 40,11266h.

Beechey, R. B. & Happold, F. C. (1955). Biochem. J. 61, xx.Beechey, R. B.&Happold, F. C. (1957). Biochem. J. 66, 520.Cammarata, P. S. & Cohen, P. P. (1951). J. biol. Chem.

198, 53.Delory, G. E. & King, E. J. (1945). Biochem. J. 89, 245.Doisy, R. J., Richert, D. A. & Westerfield, W. W. (1955).

J. biol. Chem. 217, 307.Fasella, P. & Baglioni, C. (1956). Acta Vitamin., Milano,

10, 27.Gomori, G. (1941-42). J. Lab. clin. Med. 27, 955.Happold, F. C. & Struyvenberg, A. (1954). Biochem. J. 58,

379.Hellerman, L., Lindsay, A. & Bovarnick, M. R. (1946).

J. biol. Chem. 163, 553.Horiuchi, T. (1959). J. Biochem., Tokyo, 40, 1467.Hurwitz, J. (1953). J. biol. Chem. 205, 935.Kallio, R. E. (1951). J. biol. Chem. 192, 371.McNutt. W. S. & Snell, E. E. (1948). J. biol. Chem. 173,

801.Mandelstam, J. (1954). J. gen. Microbiol. 11, 426.Meister, A., Sober, H. A. & Tice, S. V. (1951). J. biol. Chem.

189, 577.Metzler, D. E., Ikawa, M. & Snell, E. E. (1954). J. Amer.

chemn. Soc. 76, 648.Nygard, A. P. & Thellwell, H. (1955). Acta chem. 8cand. 9,

1587.Pardee, A. B. (1958). In The Enzyme8, 2nd ed., vol. 1,

p. 681. Ed. by Boyer, P. D., Lardy, H. & Myrback, K.New York: Academic Press Inc.

Peterson, E. A., Sober, H. A. & Meister, A. (1953). InBiochemical Preparations, vol. 3, p. 34. Ed. by Snell,E. E. London: Chapman and Hall Ltd.

Pogell, B. M. (1958). J. biol. Chem. 232, 761.Rabinowitz, J. C. & Snell, E. E. (1947). J. biol. Chem. 169,

643.Rodwell, V. W., Volcani, B. E., Ikawa, M. & Snell, E. E.

(1958). J. biol. Chem. 233, 1548.Saran, A. (1958). Biochem. J. 70, 182.Snell, E. E. (1958). Vitam. & Horm. 15, 78.Snyder, J. Q. & Wender, S. H. (1953). Arch. Biochem.

Biophys. 40, 465.Wada, H., Morisue, T., Nishimura, Y., Morino, Y.,

Sakamoto, Y. & Ichihara, K. (1959). Proc. imp. Acad.Japan, 35, 229.

Biochem. J. (1961) 78, 372

The Sedimentation Characteristics of Deoxyribonucleic Acidfrom Human Tissues

BY P. A. BIANCHI* Aim K. V. SHOOTERChester Beatty Re8earch In8titute, In8titute of Cancer Re8earch: Royal Cancer Hospital, London, S.W. 3

(Received 21 July 1960)

Polli & Shooter (1958) have presented the dataobtained from sedimentation experiments withpreparations of deoxyribonucleic acid from normalhuman spleen and leucocytes and from leucocytes

of patients with chronic myeloid and lymphaticleukaemia. It was found that the average sedi-mentation coefficients at infinite dilution for pre-parations from leucocytes of lymphatic leukaemiawere higher than those observed for preparationsfrom normal leucocytes. Different preparations* British Council Scholar.