NOTES 41 1

21: 197-203. SHUMAN, R. D., N. NORD, R. W. BROWN, and G. E.

WESSMAN. 1972. Biochemical and serological character- istics of Lancefield groups E, P and U streptococci and Streptococcus uberis. Cornell Vet. 62: 540-568.

STRAUS, D. C., S. J . MAITINGLY, T. W. MILLIGAN, T. I. DORAN, and T. J. NEALON. 1980. Protease production by clinical isolates of type I11 group B streptococci. J. Clin. Microbiol. 12: 421 -425.

WESSMAN, G. E. 1982. Growth of group E streptococci and production of antigens in a chemically defined medium. Can. J . Microbiol. 28: 106- 110.

WESSMAN, G. E., R. D. SHUMAN, R. L. WOOD, and N. NORD. 1971. Swine abscesses caused by Lancefield's group E streptococci. IX. Comparison of the precipitin,

hemagglutination and agglutination tests for their detec- tion. Cornell Vet. 61: 400-415.

WESSMAN, G. E., and R. L. WOOD. 1979. lmmune response in swine given soluble antigens from group E streptococ- cus. Am. J . Vet. Res. 40: 1553- 1557.

WESSMAN, G. E. , R. L. WOOD, and N. NORD. 1977. De- tection of antibody to the antiphagocytic factor produced by group E streptococci. Cornell Vet. 67: 8 1-9 1.

WOODS, R. D., and R. F. Ross. 1975. Purification and sero- logical characterization of a type-specific antigen of Strep- tococcus equisimilis. Infect. Immun. 12: 881 -887.

WOOLCOCK, J. B. 1974. Purification and antigenicity of an M-like protein of Sireptococcus equi. Infect. lmmun. 10: 116-122.

Escherichia coli growth on lactose requires cycling of P-galactosidase products into the medium

R. E. HUBER AND K. L. HURLBURT Division of Biochemistry, Departrnent of Chemistry, University Biochemistry Group, University of Calgary,

Calgary, Alta., Canada T2N IN4

Accepted October 2 1 , 1983

HUBER, R.E., and K. L. HURLBURT. 1984. Escherichia coli growth on lactose requires cycling of P-galactosidase products into the medium. Can. J. Microbiol. 30: 41 1 -415.

Growth on lactose was found to be restricted in an Escherichia coli strain deficient in its ability to transport glucose and galactose. If the latter sugars were removed from the medium as they were being produced, a wild-type strain grew only poorly, - . .

while the transport-deficient strain did not grow at all. These results suggested that all of the products of P-galactosidase action on lactose are released into the medium before being metabolized. This contention was strongly supported by the finding that the appearance of products in the medium was equal to lactose disappearance at three limiting lactose concentrations and by an experiment which showed that essentially all of the label from added lactose ([l-'4C]glucose) was found in the medium as glucose when chased with unlabelled lactose.

HUBER, R.E., et K. L. HURBLURT. 1984. Escherichia coli growth on lactose requires cycling of P-galactosidase products into the medium. Can. J. Microbiol. 30: 41 1-415.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . La croissance sur du lactose s'est avCrCe restreinte pour une souche de Escherichia coli dCficiente en capacitC de transport . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . - . du glucose et du galactose. Si ces derniers sucres sont enlevCs du milieu dts qu'ils sont produits, une souche de type sauvage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

fait une faible croissance tandis que la souche dCficiente en capacitC de transport ne fait pas de croissance. Ces rCsultats . . . . . . . . . . . . . . . . . . . . . . . . . . . .

suggtrent que toutes les substances provenant de 1'activitC de la P-galactosidase sur le lactose seraient libCrCes dans le milieu . . . . . . . . . . . . . . . . . . . . . ' avant qu'elles soient mCtabolisCes. Cette hypothtse est fortement appuyke par I'observation du fait que l'apparition de ces

substances dans le milieu correspond a une disparition du lactose i trois concentrations limitantes de lactose ainsi que par une expCrience qui dCmontre que, essentiellement, tout produit marquC dCriv.6 d'un ajout de lactose ([l-'4~]glucose) se retrouve dans le milieu sous forme de glucose lorsque ce produit est remplacC par du lactose non marquC.

[Traduit par le journal]

Escherichia coli strains K12 and JM1098 (ptsF ptsM ports these sugars only slowly (Henderson et al. 1977), ptsG galP mgl) are both inducible for the lac operon, whereas K12 is wild type. (Strain JM1098 was a gift have roughly equal amounts of lac operon enzymes, from Dr. Jones-Mortimer, Cambridge.) Strain JM1098 and release P-galactosidase products into the medium at grew somewhat more slowly than K12 in minimal glyc- equal rates (Huber, Pisko-Dubienski et al. 1980). The erol medium; the doubling time for JM 1098 was 1 1 1 latter strain is almost entirely deficient in the major min and that for K12 was 9 3 min. However, JM1098 transport systems for glucose and galactose and trans- grew much more poorly than did K12 in minimal lac-

Can

. J. M

icro

biol

. Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IV C

HIC

AG

O o

n 11

/10/

14Fo

r pe

rson

al u

se o

nly.

FIG

. I.

Gro

wth

of

K12

and

JM

1098

on

min

imal

lac

tose

(0.

2%)

med

ium

in

the

pres

ence

and

abs

ence

of

ML

308-

225.

An

al~

quot

(0.1

niL

) of

a d

ilut

ion

of c

ultu

re

was

spr

ead

on m

inim

al l

acto

se p

late

s w

hich

wer

e su

pple

men

ted

as r

equi

red

for

the

part

icul

ar s

trai

n. T

he u

pper

row

sho

ws

a di

lutio

n an

d th

e bo

ttom

row

sh

ows

a di

luti

on o

f th

e sa

me

cult

ure.

Col

umns

lab

elle

d L

acto

se s

how

the

str

ains

pla

ted

on m

inim

al l

acto

se m

ediu

m a

lone

and

col

umns

labe

lled

Lac

tose

+

ML

308-

225

show

the

stra

ins

plat

ed o

n m

inim

al l

acto

se p

late

s w

hich

had

fir

st b

een

spre

ad w

ith 0

. I

mL

of

a sa

tura

ted

cult

ure

of M

L30

8-22

5. T

he a

rrow

s in

dica

te

halo

es o

r sh

adow

s w

here

the

gro

wth

of

ML

308-

225

on t

he d

iffu

sed

prod

uct

has

beco

me

visi

ble.

Can

. J. M

icro

biol

. Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IV C

HIC

AG

O o

n 11

/10/

14Fo

r pe

rson

al u

se o

nly.

NOTES

TABLE 1. Rates of lactose disappearance from the total culture and product appearance in the medium of strain ML308 when incubated with lactose at different growth-limiting concentrations

I mM lactose 0.5 mM lactose 0.25 mM lactose

Rate of glucose appearance +0.07220.009 +0.05320.013 +0.022t0.004 Rate of galactose appearance +0.073t0.009 +0.052k0.011 +0.033k0.005 Rate of allolactose appearance +0.05920.006 +0.04620.020 t0.02420.008

Total rate of appearance of products +0.204t0.014 +O. 151 k0.026 +0.079?0.010

Rate of lactose use -0.19620.034 -0.154t0.020 -0.078?0.010

NOTE: All values are given in ~nicromoles hexose per minute per milligram dry weight of cells. The rates of glucose and galactose appearance were measured when they became stable (i.e.. after a lag). This only took a short time at I mM lactose but about 2 min at 0.25 m M lactose.

tose medium, with the doubling time for the former strain being 408 min as compared with 50 min for K12. The results were even more dramatic with a plating experiment. A series of plating dilutions of JM1098

source and thus severely restricting the growth of K12. When JM1098 was plated rather than K12, no growth was observed because JM1098, already impaired in its ability to grow on lactose, was unable to compete for

showed that colonies formed on minimal lactose plates from only 2% of the viable cells present. Thus, impair- ment of the ability to transport glucose and galactose

the monosaccharides. (The few colonies observed around the edge of the dilution plate were probably due to incomplete spreading of the ML308-225 lawn to

resulted in a significant impairment of the ability to grow when lactose was the sole carbon source.

Confirmation of this result was provided by a double-

the edges of the plate.) he shadows observed around some of the K12 colonies were most likely due to the growth of ML308-225 on the diffused glucose and galactose becoming visible.

These experiments strongly suggested that all of the glucose and galactose produced from lactose is first released into the medium and then taken up again for metabolism. Further evidence for this hypothesis was obtained when strain ML308 (lac[) (Fox et al. 1967)

plating experiment. (This experTment could only be done on plates because in liquid culture it would be impossible to determine which part of the observed growth was due to K12 or JM1098 and which part was due to ML308-225.) Minimal lactose plates were first spread with a lawn of ML308-225 ( l ad IacZ), a strain which does not have an active P-galactosidase, and was fed lactose at three different growth limiting con-

centrations. The appearance of glucose and galactose in the medium was measured by coupled enzyme assays (Huber, Pisko-Dubienski et al. 1980) and the amount of lactose in the total culture and allolactose in the medium was monitored by gas chromatography (Huber, Lytton et al. 1980). Table 1 shows that the rate of appearance of products in the medium was equal to the rate of disappearance of lactose from the total culture in all three cases. Note that the concentrations of lactose used

which, therefore, cannot grow on this medium (Fox et al. 1967); it can, however, use both glucose and galactose as carbon sources for growth. A series of plating dilutions (including the one shown in Fig. I) shows that, for K12, the number of colonies on the ML308-225 lawn was only about 2% of the number observed on plates without~the lawn. It was also found that some of the K12 colonies developed haloes or shadows (see arrows). The results were even more dra- matic for JM1098: few, if any, JM 1098 colonies were were growth limiting in every case (especially at

0.25 mM). The stoichiometry did not decrease with decreases in lactose concentrations. If the release of

observed on the lawn, even when an estimated 6 x lo5 viable cells were plated (lo-' dilution). A control ex- periment in which ML35 (1ac.I lacy) was used as the products was a phenomenon due to excess lactose, then

at these growth-limiting lactose concentrations, the stoichiometry should decrease. The fact that it did not shows that the phenomenon does not occur only at

lawn gave the same results. The strain does not have an - active lactose permease and therefore does not accumu- late lactose.Thus, depletion of the carbon source from the medium by the lawn is not a factor in the observed excess lactose concentrations. Preliminary studies kith

several other E. coli strains suggested that this result is common to all strains of this organism.

results. These results are best explained as follows. The wild-

type strain (K12) hydrolysed lactose and released the It should be noted here that at low levels of lactose there was a definite time lag between introduction of latose and the point at which the steady-state rates re- ported in the table were found. This delay was about 2

resultant glucose and galactose into the solid medium. These monosaccharides were scavenged by ML308- 225 or ML35, causing depletion of the available carbon

Can

. J. M

icro

biol

. Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IV C

HIC

AG

O o

n 11

/10/

14Fo

r pe

rson

al u

se o

nly.

414 CAN. 1. MICROB IOL. VOL. 30, 1984

min FIG. 2. Lactose chase of JM1098 and uptake of lactose by

K12. The circles indicate nanomoles lactose per milligram dry weight (wt.) of cells in the cellular fraction. 0, Amount of lactose if no chase solution is added; 0, amount'of lactose in the cells after 30 mM unlabelled lactose was added (time of addition is indicated by the arrow); A, amount of glucose found in the medium before and after the chase (nanomoles glucose per milligram dry weight of cells); 0, control uptake of 0. I mM lactose by K12 cells.

min at 0.25 mM lactose. The significance of this delay will be discussed below.

Further support of the hypothesis that the sugars are totally effluxed before being metabolized was provided by a chase experiment. Strain JM1098 was grown to midlog phase in minimal glycerol containing isopropyl- P-D-thiogalactoside (lPTG), after which the culture was centrifuged and the cells were resuspended in minimal medium without a carbon source or IPTG. Radio- labelled lactose ([I -'4C]glucose) was added to a concen- tration of 0. I mM; 2.5 min later, 30 mM unlabelled lactose was added as a chase. At various times, aliquots of the incubation mixture were centrifuged through sil- icone oil (75% Dow Corning 550, 25% Dow Corning 51 0, by volume) to separate the medium from the cells (Chalfie et al. 1976). Samples of the !ayers, including the pelleted cells, were counted in liquid scintillation cocktail (4 g Omnifluor per litre of scintillation-grade toluene). ~ a h ~ l e s of the medium were also subjected to gas- liquid chromatography and the glucose peaks were collected by condensation on a Pasteur pipette (Burchfield and Storrs 1962). The pipettes were rinsed with 10 mL of cocktail directly into scintillation vials which were counted to determine the [14C]glucose con- tent. A control experiment was performed with ML35 (lac1 lacy). a strain which doesnot have an active lac . . . . permease and which neither binds nor transports lactose (Fox et al. 1967). Values obtained for radioactivity in the cellular fraction of this strain were subtracted

from values obtained for JM1098 to account for non- specific binding and for label clinging to the cells as they sedimented.

Figure 2 shows that in the absence of a chase the radioactivity in the JM 1098 cells increased steadily for about 2 min and then levelled off. When the chase solution (30 mM) was added at 2.5 min, the amount of label in the cellular fraction decreased nearly to zero within a short time. The chase caused a rapid rise in the amount of glucose in the medium, reaching a maximum at values approximately equal to the amount of radio- activity that had been released from the cells by the chase. In other words, the chase caused the label in the lactose taken up by the cells to appear in the medium as glucose. This experiment strongly suggests that glucose is released into the medium before being metabolized, since if the sugar were being immediately metabolized, the label would become an integral part of the cell very rapidly and therefore could not be chased into the me- dium. Indeed, if the first fate of glucose is metabolism, there is no reason why it should be chased out of the cell by lactose at all.

Another control experiment was carried out to ensure that the results with strain JM1098 were not due to unusual lactose transport resulting from the effects of pts mutations. In Fig. 2 uptake of 0.1 mM ['4C]lactose by K12, a strain which is wild type with respect to the pts genes, is compared with uptake of the same com- pound by JM1098. It can be seen that the uptake is essentially the same for both strains. Thuspts mutations were not a factor in the results observed in the chase experiment.

In,each case in the chase experiment, the amount of lactose in the cells rose rapidly for 2 min before reaching a constant level. Presumably this reflected processes such as crossing the outer membrane or dif- fusing across the periplasmic space and becoming at- tached to the lac permease, which had to occur before a steady state could be achieved. As mentioned above, a delay in the efflux of glucose and galactose (Table 1) was also seen. This delay was short at high lactose concentrations and was attributed to the time required for accumulation of products in the coupled enzyme assay (Rudolph et al. 1979). However, at low lactose concentrations the lag was longer and was most likely due to time required for lactose to diffuse to the lac permease and to be processed there.

Overall the experiments described in this study show that under the conditions of this study the products of E. coli P-galactosidase action on lactose are all cycled into the medium and that re-uptake of these products is necessary for growth on lactose as the sole carbon source to occur. When this process was first discovered (Huber, Lytton et al. 1980), it was thought that only excess products, not immediately needed for metabo-

Can

. J. M

icro

biol

. Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IV C

HIC

AG

O o

n 11

/10/

14Fo

r pe

rson

al u

se o

nly.

NOTES 415

lism were released, but the results strongly suggest that this is not the case.

The reason that E. coli excretes the products of P-galactosidase action on lactose is far from obvious. ~ o i s i b l ~ there is not enough room inside the cells to store the large amounts of galactose and glucose which can be produced. Another possibility is that the excre- tion may be energy conserving, in that if glucose is excreted and taken up by the pts system it is not neces- sary for E. coli to have hexokinase. Indeed, the hexo- kinase content of E. coli is known to be low (Tanaka et al. 1967). The process may be important for chemo- taxis. It may even be possible that metabolites are shared to help the overall growth of the culture. Efflux of metabolites is not uncommon in microbes (Cooper e t al. 1978; Reizer and Panos 1980). The questions as to how and why this unexpected phenomenon occurs are currently being investigated in-our laboratory.

BURCHFIELD, H. P., and E. E. STORRS. 1962. Biochemical applications of gas chromatography. Academic Press, New York. pp. 124-137.

CHALFIE, M., D. HOADLEY, S. PASTAN, and R. L. PERLMAN. 1976. Calcium uptake into rat Pheochromocytoma cells. J.

I

Neurochem. 27: 1405 - 1409. COOPER, D. G., K. L. KENNEDY, D. F. GERSON, and J. E.

ZAJIC. 1978. The production of extracellular galactose by

Arthrobacter globiformrs using lactose as a substrate. J. Ferment. Technol. 56: 550-553.

Fox, C. F., J. R. CARTER, and E. P. KENNEDY. 1967. Gen- etic control of the membrane protein component of the lactose transport system of Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 57: 698-705.

HENDERSON, P. J. F., R. A. GIDDENS, and R.C. JONES- MORTIMER. 1977. Transport of galactose, glucose and their molecular analogues by Escherichia coli K12. Biochem. J. 162: 310-320.

HUBER, R. E. J. LYTTON, and E. B. FUNG. 1980. Efflux of P-galactosidase products from Escherichia coli. J. Bacte- riol. 141: 528-533.

HUBER, R. E., R. PISKO-DUBIENSKI, and K. L. HURLBURT. 1980. Immediate stoichiometric appearance of P- galactosidase products in the medium of Escherichia coli cells incubated with lactose. Biochem. Biophys. Res. Commun . 96: 656 -66 1.

REIZER, J. , and C. PANOS. 1980. Regulation of P-galacto- sidase phosphate accumulation in Streptococcus pyrogenes by an expulsion mechanism. Proc. Natl. Acad. Sci. U.S.A. 77: 5497-5501.

RUDOLF, F. B., B. W. BAUGER, and R. S. BEISSNER. 1979. Methods in enzymology. Vol. 63. Academic Press, New York. pp. 22-41.

TANAKA, S. , D. G. FRAENKEL, and E. C. C. LIN. 1967. The enzymatic lesion of strain MM-6, a pleiotropic carbohydrate-negative mutant of Escherichia coli. Bio- chem. Biophys. Res. Commun. 27: 63-67.

Spheroplasts of Rhizobium japonicum'

JAMES 0. BERRY^ AND ALAN G . ATHERLY' Department of Genetics, Iowa State Universiry, Ames, IA, U.S.A. 50011

Accepted November 4, 1983

BERRY, J . O., and A. G. ATHERLY. 1984. Spheroplasts of Rhizobium japonicum. Can. J. Microbiol. 30: 415-419. Speroplasts of Rhizobium japonicum strains 61A76, USDA 3 1, and 110 were prepared by culturing cells in the presence

of glycine, followed by treatment with lysozyme. The cells were examined by scanning electron microscopy before, during, and after becoming spheroplasts and found to be morphologically similar to the bacteroid forms found in soybean root nodules. Some similarities of spheroplast and bacteroid formation are discussed.

BERRY, J. O., et A. G. ATHERLY. 1984. Spheroplasts of Rhizobium japonicum. Can. J. Microbiol. 30: 415-419 Des sphCroplastes des souches 61A76, USDA 31 et 110 de Rhizobium japonicum ont CtC prCparCs en cultivant les cellules

en presence de glycine, ces derni2res ont CtC traitCes en suite au lysozyme. Les cellules ont CtC examinees a I'aide d'un microscope Clectronique a balayage, avant, pendant et apr2s la formation des sphCroplastes. Nous avons observC que les sphCroplastes sont morpholo~iquement semblables aux bactkroides que I'on retrouve dans les nodules des racines de la fkve sbja. d n discute des sirnilarks qui existent lors de la formation des sphCroplastes et des bactCro'ides.

[Traduit par le journal]

'~ournal Paper No. J-11026 of the Iowa Agriculture and Home Economics Experiment Station, Project 2299, supported in part by grant 59-2191-0-1-494-0 from the United States Department of Agriculture and funds from Land O'Lakes Corp., M~nneapolis, MN, U.S. A.

'present address: Department of Biology, University of Utah, Salt Lake City, UT, U.S.A. 'Author to whom all correspondence should be addressed.

Can

. J. M

icro

biol

. Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IV C

HIC

AG

O o

n 11

/10/

14Fo

r pe

rson

al u

se o

nly.