THE JOURNAL OF BIOLOGICAL CHEMLWRY Vol. 243, No. 23, Issue of December 10, pp. 6291-0299, 1968

hinted in U.S.A.

Size Restriction on Peptide Utilization in Escherichia coZi*

(Received for publication, July 18, 1968)

JOHN W. PAYNET AND CHARLES GILVARG

From the Program in Biochemical Sciences, Frick‘Chemical Laboratory, Princeton University, Princeton, Nezo Jersey 08540

SUMMARY

It has been shown by measuring the growth response of appropriate amino acid auxotrophs to several series of homologous oligopeptides that higher oligomers were unable to enter Escherichia co& The exclusion did not always occur at the same chain length of peptide, for the oligomer at which exclusion was tist observed varied with the nature of the amino acid residues. To investigate the possibility that over-all size may be important in regulating peptide entry, the relative hydrodynamic volumes of the peptides used in the above studies were measured by gel filtration on Sephadex G-15. These studies suggested that to be able to enter E. coli the hydrodynamic volume of a peptide could not exceed a certain critical value. To test the generality of this observa- tion, the heterogeneous mixture of peptides present in the enzymatic protein hydrolysate Neopeptone was separated according to size on Sephadex G-15. On the basis of the previous measurements it proved possible to exactly predict which of the Neopeptone peptide fractions could enter the cell and which would be excluded. Lack of competition between peptides that can enter and peptides that are excluded, sug- gests that the exclusion may occur external to the peptide transport system. The speculation is raised that the cell wall may act as a molecular sieve preventing molecules above a certain size from crossing the cell envelope, the critical hydrodynamic volume being some measure of the pore size of such a sieve.

In previous studies of peptide transport in bacteria, attempts were made to ascertain those structural features of a peptide molecule that are essential for it to be handled by a peptide trans- port system (l-3). Initially, several homologous series of oligo- peptides were used in these studies. The ability of any member of these series to enter a bacterium was measured from the ability of the peptide to support the growth of the appropriate amino acid auxotroph. With the use of this criterion with auxotrophs

* This work was supported by Research Grant AM-10336 from the National Institutes of Health.

$ Postdoctoral Fellow of the North Atlantic Treaty Organiaa- tion.

of Escherichia coli, it was demonstrated that for an oligopeptide to be transported the presence of a free a-amino group is essential (1,2), the terminal carboxyl group performs no vital role (3), and the nature of the amino acid side chains is relatively unimportant (1, 3). These conclusions were based upon the following ob- servations. Acetylation of the a-amino group of a utilizable peptide destroyed the nutritional value of the peptide (1, 2). Lysylcadaverine oligopeptides, which are lysine oligopeptides that lack a terminal carboxyl group, could satisfy the growth re- quirement of a lysine auxotroph (3). And finally, studies with mutants deficient in oligopeptide transport, and experiments demonstrating competition for entry among diverse peptides, indicated that structurally dissimilar oligopeptides use a common transport system to enter E. coli (4).

During these studies, it was observed that higher members of several oligopeptide series could satisfy the structural require- ments for oligopeptide transport described above, and yet be unable to enter E. coli. For example, with lysine and arginine oligopeptides, nutritional effectiveness ceased abruptly with the tetrapeptides; the pentapeptides and higher homologues were all unable to support growth (1). The reason did not appear to lie in a critical contour length of the peptide backbone, for the homo- logue at which exclusion occurred varied among different oligo- peptide series. In the lysylcadaverine series the tetrapeptide was excluded (3)) with lysine and arginine oligopeptides the exclusion began with the pentapeptides (l), while in the lysylalanine and glycine series exclusion was not yet observed with the penta- and hexapeptides respectively (4). It was also shown that the ability to utilize the lower peptides did not arise because the cells excreted peptidases able to hydrolyze peptides only up to a certain size limit, for in fact, no peptidase activity was detectable in the culture fluid during growth. And furthermore, the dis- crimination against the higher peptides was not caused by the absence of peptidases, for the existence of intracellular activity able to cleave the non-utilizable peptides was demonstrated (1).

Following these observations, the work described here was designed to explore the possibility that the hydrodynamic volume of a peptide might dictate whether or not the peptide will be able to enter E. coli. A measure of the relative Stokes’ radii of known peptides was obtained by the use of gel filtration with Sephadex G-15. The Stokes’ radii were assessed from the volumes required for elution of the peptides from a column of the gel. These volumes lay between the elution volume (I’,) for the marker blue dextran 2996 (which can enter only the interstitial

6291

by guest on October 3, 2020

http://ww

w.jbc.org/

Dow

nloaded from

6292 Peptide Utilization Vol. 243, No. 23

space between the gel particles) and the elution volume (V3 for bromide ion (which can completely penetrate the gel matrix). After subtraction of the void volume (V,) from The elution volume for any peptide, the resulting volume was expressed as a fraction, or Kd value, of that additional space (V t - V,) accessible to bromide but not to blue dextran 2000. It is shown that for several oligopeptide series a regular size increment occurs with the addition of each amino acid residue, the actual incre- ment being dependent upon the nature of the amino acid. The results indicate that the tetrapeptide, trilysylcadaverine, which is excluded from E. coli, is larger than tetralysine which can enter, and furthermore, lysyltetraalanine and hexaglycine which enter E. coli are both significantly smaller than tetralysine. From these data, an approximate Kd value was obtained that corre- sponded to the size at which discrimination against peptide entry commenced. To test the generality of this relation between size and ability to enter E. coli, Neopeptone, a heterogeneous mixture of peptides obtained from an enzymatic protein hydrolysate, was also fractionated on Sephadex G-15. The fractions were tested for their abilities to support the growth of several auxotrophs; fractions with Kd values smaller (i.e., fractions that contain pep- tides of larger size) than that referred to above gave no growth; growth commenced with the fraction corresponding to the critical

Kd. The results support the idea that if the over-all size of a peptide exceeds a certain critical value the peptide may be pre- vented from entering E. coli. The site at which this discrimina- tion is exerted is not known, but as other experiments indicate that large excluded peptides do not compete for entry with the smaller utilizable peptides it is suggested that the discrimination may occur before the peptides interact with the peptide transport system. A preliminary report upon part of this work has ap- peared earlier.’

EXPERIMENTAL PROCEDURE

Materials

Peptides

Lysyltrialanine and lysyltetraalanine were a gift from Dr. I. Schechter of the Weizmann Institute. All peptides containing ornithine residues were a gift from Dr. Y. Levin, also of the Weizmann Institute. Lysine oligopeptides and lysylcadaverine oligopeptides were prepared as described previously (3). The synthesis of cr-acetyl trilysine was reported earlier (1). Glycine peptides were purchased from Cycle Chemical Corporation (Los Angeles, California), and all other peptides were purchased from Mann. Neopeptone was obtained from Difco.

Bacterial Strain-s

The following strains derived from E. coli W (ATCC 9637) were used. Strain M-26-26 is a lysine auxotroph that lacks the de- carboxylase that converts meso-diaminopimelic acid to L-lysine (5, 6). Strain M-123, a glycine-serine auxotroph, was described earlier (4). The bacterial cultures were grown in the Medium A of Davis and Mingioli (7). Strain M-26-26 was grown in Klett tubes without agitation. Strain M-123 was grown in Nephelo flasks with shaking. All cultures were incubated at 37”.

1 PAYNE, J. W., AND GILVARG, C., Fed. Proc., 26, 393 (1967).

Methods

Chromatography on Sephadex G-l 6

Measurement of Peptide Size-The dry gel (Pharmacia Fine Chemicals, Ltd.) was allowed to swell for 12 hours at room temperature in water containing approximately 0.1 M sodium chloride. Fines were removed by decantation. The column (110 x 1.4 cm) was packed under gravity. This column was used for all measurements of the hydrodynamic volumes of the peptides. Before applying a sample the column was equilibrated with the eluent; in most cases the eluent used was 0.1 M sodium chloride in 0.01 M sodium phosphate, pH 7. The peptide (ap- proximately 1 mg) was dissolved in the solution used for elution (1 ml), to which had been added blue dextran 2000 (0.1 r;O, w/v) and sodium bromide (0.2%, w/v). Individual peptides were applied to the column in a volume of 0.4 ml. The rate of flow was 24 ml per hour. The effluent was monitored continuously at 220 rnp using a Gilford model 2000 spectrophotometer equipped with flow cells. The concentrations were selected so that when monitored at this wave length similar absorbances were recorded for blue dextran 2000, bromide ion, and the peptide; these ab- sorbances were plotted out automatically. Under the standard conditions of constant flow rate and constant chart speed, the distance of the chart paper between the maximum of the ab- sorbance peak for blue dextran 2000 (corresponding to the void volume, VO) and the maximum of the absorbance peak for bro- mide ion (corresponding to the total penetrable volume, V t) is a measure of the column’s inner volume (Vi). The fraction (Ka) of this inner volume accessible to any peptide was then calculated directly from the position of the maximum of the peptide ab- sorbance peak between these two markers. The results for this column were reproducible to 0.5%. This automated procedure obviated the need for assaying materials in effluent fractions and expressing the results as volumes. Nevertheless, it was initially confirmed that identical results were obtained with this latter procedure. In these experiments it was shown that the void volume, VO, was 55.3 ml and the total penetrable volume, V,, was 126.6 ml. When tritiated water and bromide ion were simultaneously added to the column, and the fractions were assayed for both bromide absorbance and radioactivity, the latter marker gave a volume of 126.0 ml for V t.

Fractionation of Neopeptone-To achieve an adequate frac- tionation of a large amount (approximately 700 mg) of Neopep- tone, a larger Sephadex G-15 column was used. The column (110 x 2.4 cm), prepared from the same batch of Sephadex G-15, was packed and equilibrated as described above. The Neopep- tone was dissolved in 2.5 ml of the eluent, 0.1 M sodium chloride in 0.01 M sodium phosphate, pH 7. The rate of flow was 38 ml per hour. Fractions of 3 ml were collected.

When the column was calibrated with blue dextran 2000 and bromide ion, a volume of 146 ml was found for VO, and 351 ml for V t. For comparison with the smaller column, the Ka values for lysine oligopeptides were determined in a way analogous to that described above.

Hydrolysis of Neopeptone FractionsSamples (0.5 ml) were removed from even numbered fractions and added to aliquots (0.5 ml) of 12 N HCl. The fractions were hydrolyzed by auto- claving in sealed tubes at 15 pounds of pressure for 11 hours. The hydrolysates were taken to dryness at 50” under reduced pressure. To remove any residual HCl the tubes were stored in a vacuum over CaClz and KOH for 36 hours. The residues were

by guest on October 3, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Issue of December 10, 1968 J. W. Payne and 6. Gilvarg 6293

redissolved in water (0.5 ml) and the solutions were sterilized by autoclaving.

RESULTS

It is generally accepted that molecular size is the dominating factor that governs the separation of substances by gel filtration. At the present time, various theories have been advanced and various physical models have been proposed in attempts to ex- plain the basis for the fractionation observed (8-11). A common feature of these models is that the size (more specifically effective diameter) of a molecule governs the extent to which it may penetrate the gel matrix. Ideally therefore, diverse molecules of equal size should be able to penetrate the gel matrix to equal extents. In the present case therefore, by measuring the ex- tents to which different peptides are able to penetrate the gel a measure of their relative sizes should be obtained. Such meas- urements are most conveniently made by determining the elution volume of the peptide on passage through a column of the gel. Unfortunately, the idealized situation described above does not always pertain in practice, and complications may arise through certain classes of compound (for example, charged compounds and those with aromatic character) interacting with the dextran gel (12-14).

E$ect of Added Electrolyte on El&on Behavior of Lysine Pep- tides--It was therefore anticipated that interactions with the gel material would be found for the peptides studied here. To test this, the column was first equilibrated and eluted with distilled water, and the peptides were also applied to the column in dis- tilled water. Data for the lysine peptides are shown in Table I. In Column 2 it can be seen that the peptides were eluted in the reverse order to that expected from their relative sizes, i.e., hexalysine was eluted last, not first. The effect presumably arises through interaction of the charged peptides with the free carboxyl groups introduced into the gel during its manufacture (9, 14). I f the above effects are caused through electrostatic binding they should be counteracted if an eluent of high ionic strength is used. To test this, 0.1 M NaCl was used as eluent, and as shown in Table I, Column 3, the anticipated elution se-

quence was observed. That the above ionic strength is sufficient to vitiate the electrostatic effects is shown by the results in Table I, Column 4, which indicate that the Kd values remained essentially unchanged when 0.5 M NaCl was substituted for 0.1 M NaCl as the eluent. In subsequent studies, the NaCl eluent was buffered at pH 7 with 0.01 M sodium phosphate in an attempt to maintain a constant charge on the peptides throughout the elution procedure. Again, this caused no significant change from

TABLE I

Effect of added electrolyte on elution behavior of lusine peptides on Sephadex G-16

Kd values were determined as described under “Methods.”

Kd values

Peptide

Distilled water AquecmIl M Aqueo;;;” Y

Dilysine 0.758 0.328 Trilysine . 0.800 0.245 0.258 Tetralysine 0.838 0.201 0.19B Pentalysine 0.885 0.105 Hexalysine . 0.910 0.126

TABLE II

Kd values for elution of various peplide series on Sephadex G-16

All samples were eluted with 0.1 M NaCl in 0.01 M sodium phos- phate buffer, pH 7. Kd values were determined as described under “Methods.”

Peptide Kd Peptide Kd

Dilysine Trilysine Tetralysine Pentalysine Hexalysine

0.325 Lysylcadaverine 0.252 Dilysylcadaverine 0.195 Trilysylcadaverine.. 0.162

0.367 0.251 0.183

0.126 Lysylalanine. 0.396 Lysyldialanine. .

Diornithine Triornithine . Tetraornithine Pentaornithine. Hexaornithine .

0.365 Lysyltrialanine . 0.286 Lysyltetraalanine.. 0.224

0.284 0.244

0.175 Lysyldiornithine. 0.140 Diornithyllysine.

Ornithyllysylornithine 0.503

0.278 0.276 0.286

Diglycine Triglycine Tetraglycine Pentaglycine Hexaglycine

0.457 Alanyldiornithine.. 0.418 Alanyltriornithine 0.375 Diornithylalanine.. 0.347

0.321 0.243 0.331

a-Acetyl trilysine 0.250

the & values of the lysine peptides determined in 0.1 M NaCl alone.

Measurement of Peptide Size-In Table II are given the & values for several peptide series. The values for the three oligopeptides lysyldiornithine, diornithyllysine, and ornithyl- lysylornithine lend strong support to the view that the elution patterns are indeed a reflection of relative peptide size. These oligopeptides would be expected to carry essentially equal charges; their molecular weights are identical. However, from observations with molecular models of various oligopeptides, it seems that when a lysine residue occupies a terminal position it is likely to increase the hydrodynamic radius of a peptide to a greater extent than when it constitutes an internal amino acid residue; and this feature would seem to be most significant in the case of charged peptides in which intramolecular repulsive forces may limit the mutual approach of the amino acid side chains causing the peptide to adopt a generally extended conformation. In accord with this expectation, it is seen that the & value of ornithyllysylornithine remains equal to that of triornithine, whereas the smaller & values for diornithyllysine and lysyldi- ornithine reflect their anticipated increased size, a size which lies between that of triornithine and trilysine. The significant changes in Kd values with the latter two peptides that result from the introduction of a single methylene residue illustrates the sensitivity of the procedure to small changes in size.

On the basis of theoretical or empirical considerations, various proposals have been made in attempts to relate elution volumes or Kd values to molecular parameters. Perhaps the most com- monly employed is the relationship between log molecular weight and Kd (15). Andrews showed (16) that this relationship was linear for a number of proteins, and Carnegie also demonstrated (17) a linear relationship for several peptides in the molecular weight range 400 to 2000. Other proposals that in certain cases lead to linear plots have been made (8, 10, 11). The data shown in Table II were plotted (not shown) according to these various

by guest on October 3, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Peptide Utilizatim Vol. 243, No. 23

proposals and no one relationship proved to be superior. All relationships gave fairly good linear plots; in general, the lysine, ornithine, and lysylalanine oligopeptides all fell on a single line, whereas the glycine oligopeptides were somewhat disjplaced from this line. In view of the feeling that with the peptides studied here, hydrodynamic volume is likely to be a more meaningful parameter than molecular weight, and on the further assumption that within each homologous series the hydrodynamic volume is likely to increase in a regular manner with the addition of each amino acid residue, the data of Table II were plotted as a function of log Kd versus number of amino acid residues. Certain of the data is shown in Fig. 1 from which it can be seen that a linear relationship exists for each series. Log K,J was selected for this plot to make some allowance for the possible distribution of pore sizes within the gel matrix, but it should be mentioned that a plot of the cubic root of K,J versus number of amino acid residues was also found to give good linear relationships. These results strongly suggest that in all series the regular addition of identical amino acid residues produced a correspondingly regular increase in peptide size. The absence of any discontinuity further sug- gests that no dramatic structural modifications occur within these series in the size range examined

E$ect of pH on Peptide Elution-To assess the importance of the net charge of a peptide in determining its Kd value (and thus presumably its over-all conformation), the ornithine peptides were also examined at pH 6 and 8. Over this pH range the de- gree of protonation of the a-amino groups changes significantly (18). It was necessarily assumed that the change in pH would affect the peptide only, and not the gel structure. When the data in Table III are compared with those for the ornithine peptides shown in Table II, it can be seen that only small changes in Kd values were found. No readily interpretable pattern emerged. In general, the peptides appeared to be larger at pH 7 than at either pH 6 or 8.

Relationship between Peptide Size and Ability to Enter E. coli- As discussed earlier, on the basis of studies upon the growth of E. coli upon various homologous peptides, it appeared that certain peptides could enter and others could not (l-3). Neither

-0.4 -

y”

$0.6-

-0.6 -

LYSYLCADAVERYL

-1.0 I I I I I I 2 3 4 5 6

AMINO ACID RESIDUES FIG. 1. Log K,J as a function of number of amino acid residues

for several peptide series. Kd values are taken from Table II.

TABLE III Efect of pH on elution behavior of ornithine peptides

on Sephadex G-16

All samples were eluted with 0.1 M NaCl in sodium phosphate buffer of the appropriate pH. Ka values were determined as described under “Methods.”

Peptide

Diornithine . . Triornithine . Tetraornithine . . Pentaornithine Hexaornithine . . .

Kd values

PH 6 PH 8

0.380 0.375 0.292 0.308 0.223 0.223 0.181 0.184 0.151 0.145

trilysylcadaverine nor pentalysine could enter, but hexaglycine and lysyltetraalanine could enter. Inspection of Table II indicates that the cut off appears to occur at a size corresponding to a Ka value in the region of 0.185 to 0.195. However, the peptides used in the growth studies and examined by gel filtra- tion were of a rather specific type and in general were charged. The question therefore arose whether a similar size exclusion might be exerted for all peptides. Thus, if a heterogeneous mixture of peptides were to be fractionated on Sephadex G-15, would it be possible to predict on the basis of Kd values whether or not the peptides in any fraction could enter E. co&’

Fractionation of Neopeptone-To test this an enzymatic protein hydrolysate, Neopeptone, was fractionated as described under “Methods.” For this fractionation a larger Sephadex G-15 column (110 x 2.4 cm) was used. The absorbance of the effluent was measured at 220 rnp, and the elution profile is shown in Fig. 2. The smooth profile testifies to the heterogeneity of the peptides. The fractions were checked for ninhydrin-positive material and a positive result was obtained with Fraction 3 and succeeding fractions. Before the Neopeptone was fractionated, the column was first calibrated with lysine peptides and & values of 0.247,0.185, and 0.157 were found for tri-, tetra-, and pentaly- sine, respectively. Reference to Fig. 2 indicates that these Kd values correspond approximately to Fractions 20, 16, and 14, respectively.

Growth of Lysine Auxotroph on Neopeptone Fraction+--The fractions were next tested for ability to support the growth of the lysine auxotroph M-26-26. Fig. 3 shows that a cut ofi is observed and essentially no growth is obtained with the initial fractions, which on the basis of the fractionation procedure should contain the large peptides. Significant growth com- menced at Fraction 15. When fraction numbers were converted to & values, it was seen that on the basis of the proposed critical & of about 0.19 growth should not be possible with peptides present in about the first 14 fractions. Therefore, the expecta- tion and the observation were closely similar.

However, there exist a number of other possible explanations for the lack of growth in the initial fractions. First, to check whether lysine was present in these fractions, they were hydro- lyzed as described under “Methods.” To rapidly screen for the presence of lysine in the hydrolysates, even-numbered samples from Fractions 2 to 60 were examined electrophoretically on Whatman CM-82 paper as described previously (3). By this procedure, the amino acids were broadly separated into positive, negative, and neutral species. Lysine was readily detected in

by guest on October 3, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Issue of December 10, 1968 J. W. Payne and C. Gilvarg 6295

Fractions 4 to 36. The distribution of amino acids changed throughout the fractions examined and it was apparent that the nature of the parent peptides must also vary significantly. In general, the positively charged amino acids were more concen- trated in the early fractions, while the neutral amino acids pre- dominated in late fractions. Lysine samples of known concen- trations were also applied to the electrophoretogram, and the intensities of the lysine spots in the hydrolysates relative to these standards gave a measure of the unknown concentrations of lysine. The concentrations determined in this way were re- ferred to a calibration curve for growth of M-26-26 as a function of lysine concentration (1). In this way, it was estimated that for Fractions 6 through 16, 1 ml of the hydrolysates contained an amount of lysine that should give 766 to 1660 Klett units in the growth assay employed. For this reason, significantly smaller volumes than 1 ml were used in growth studies upon the hydrolysates, as is shown in Fig. 3. It can be seen that, in contrast to the results with the intact peptides, growth was obtained upon the hydrolysates from the initial fractions, and that the extent of this growth was essentially that predicted on the basis of the electrophoretic estimation of lysine concentra- tion. Lack of lysine in the peptides present in the initial frac- tions cannot therefore be the reason for their inability to support growth.

Second, the addition of l-ml samples from the initial fractions

IE DEXTRAN

FRACTION NUMBER

FIU. 2. Fractionation of Neopeptone on Sephadex G-15. Frac- tions of 3 ml were collected. Arrows indicate positions at which blue dextran first emerges from the column, and the peak of the absorbance profile for elution of bromide ion. Elution procedure was as described under “Methods.”

7 I

FIG. 3. Growth response of lysine auxotroph, M-26-26, to Neopeptone fractions before and after hydrolysis. In all cases of growth upon unhydrolyzed peptides, l-ml samples were removed from appropriate fractions. Growth was carried out in Klett tubes in a total volume of 5 ml. Incubation was for 10 hours. In a control, the addition of 1 ml of eluent, i.e. 0.1 M NaCl in 0.01 M sodium phosphate, did not affect the growth of M-26-26 on lysine- supplemented media. With hydrolyzed fractions, 0.3 ml was used from Fractions 2 and 4, and 0.1 ml from all other fractions. Data have been corrected to correspond to growth obtained with 1 ml of each fraction.

to lysine-supplemented media was without effect upon the growth of the lysine auxotroph. Therefore, the lack of growth with these fractions is not caused by their being inhibitory to the growth of M-26-26.

Third, the possibility existed that the peptides in the initial fractions could enter E. coli but perhaps because of structural specificities of the intracellular peptidases they could not be cleaved to release lysine. To check this point, samples from these fractions were incubated with crude extracts from M-26- 26 as described previously (4). After incubation at 37”, enzy- matic reaction was stopped by the addition of acid, the suspen- sions were centrifuged, and samples of the supernatant solutions were examined electrophoretically. Extensive cleavage was observed; lysine together with other amino acids was released and the electrophoretic pattern resembled that observed with the acid hydrolysates examined previously.

A final possibility that competition for entry among the many peptide constituents might discriminate against peptides containing lysine residues is considered later in another context and shown to be unlikely.

Growth of Glycine-Serine Auxotroph, M-1.23, on Neopeptone Fractions-If the above observations on the growth response of the lysine auxotroph are indeed an example of a general phenom- enon of peptide exclusion, then it would be anticipated that essentially the same pattern of growth response should be ob- tained irrespective of the nature of the amino acid auxotroph studied. To test this the growth response of a glycine-serine auxotroph on the Neopeptone fractions was also studied. Be- cause single fractions contained insufficient material, alternat- ing fractions were used to measure the growth response to un- hydrolyzed and hydrolyzed samples. Fig. 4 shows that, as with the lysine auxotroph, the initial fractions could support essentially no growth before hydrolysis, but commencing at about 15 a dramatic increase in peptide utilization occurred. For example, the total growth, in Klett units per ml of sample,

by guest on October 3, 2020

http://ww

w.jbc.org/

Dow

nloaded from

6296 Peptide Utilization Vol. 243, No. 23

was 8 for Fraction 5 before hydrolysis, 510 for Fraction 6 after hydrolysis; 19 for Fraction 13 before hydrolysis, and 109 for Fraction 14 after hydrolysis. The result therefore supports the idea that the phenomenon is general for all peptides.

Distribution of Peptide Size in Neopeptone Fractions--When hydrolyzed and unhydrolyzed samples from the Neopeptone fractions were examined electrophoretically, two features of the electrophoretograms supported the idea that separation was achieved on the basis of peptide size. First, with unhydrolyzed samples from the initial fractions, ninhydrin-positive material remained near the origin, with increasing fraction number the extent of migration increased to give an elongated ninhydrin- positive band, and with late fractions (approximately 40 and higher) the distribution of ninhydrin-positive materials ap- proached that seen for the hydrolyzed materials. And second, with the initial fractions, the amount of ninhydrin-positive material seen before hydrolysis was markedly less than that seen with hydrolyzed samples. Moreover, this disparity disap- peared in a regular manner with increasing fraction number.

To obtain a quantitative estimate of the average size of peptide in each fraction, measurements were made of total

FRACTION NUMBER

FIG. 4. Growth response of glycine-serine auxotroph, M-123, to Neopeptone fractions before and after hydrolysis. For un- hydrolyzed fractions, l-ml samples were used from odd-number fractions. In a control, the addition of 1 ml of eluent, i.e. 0.1 M NaCl in 0.01 M sodium phosphate, pH 7, did not affect the growth of M-123 on media supplemented with glycine. With hydrolyzed fractions, 0.2-ml samples were used from even-numbered frac- tions. In all cases, growth waz carried out in Nephelo flasks in a total volume of 5 ml. Incubation was for 24 hours. Data were corrected to correspond to growth obtained with 1 ml of each fraction, and then the amount of growth obtained before hydroly- sis was expressed as a percentage of the total amount obtained after hydrolysis. 8-8, percentage of growth relative to suc- ceeding fraction; C-0, percentage of growth relative to pre- ceeding fraction.

TABLE IV Distribution of peptide size throughout Neopeptone fractions.

Amino groups were determined by the method of Dubin (19). Concentrations are per ml of fraction. Lysine concentrations were calculated from a calibration curve of growth versus lysine concentration (1) by using growth yields of M-26-26 on hydrolyzed samples. Values for concentrations of (Y-NH* groups have been corrected for the contribution of e-amino groups of lysine. Aver- age number of amino acid residues was calculated from ratio of a-NH2 groups present after hydrolysis to those present before hydrolysis.

Fraction

4 6 8

10 12 14 16 18 22 24 23 32 36 44

Lysine

pnole/ml

0.133 0.640 0.890 0.890 0.890 0.940 0.615 0.500 0.355 0.205 0.135 0.210 0.080 0.092

-r a-NH* grouPs

Before After hydrolysis hydrolysis

rwlml 1.01 15.7 15.6 5.25 52.0 9.9 7.10 59.0 8.3 7.69 54.1 7.0 7.60 56.2 7.4 8.74 55.0 6.3 8.55 52.0 6.1 7.30 37.0 5.1 6.05 32.0 5.3 5.30 20.0 3.8 5.65 17.0 3.0 5.60 17.7 3.2 6.00 11.3 1.9 4.27 6.14 1.4

P

_-

overage number of amino acid

residues

I I OO

I I I IO 20 M 40

FRACTION NUMBER

FIG. 5. Distribution of peptide size in Neopeptone fractions. Data are taken from Table IV. ATTOW indicates point at which peptide exclusion is first detected.

a-amino content before and after hydrolysis, by using the dinitro- fluorobenzene assay of Dubin (19). Approximate peptide size was obtained from the ratio of these two values. In this assay, e-amino groups of lysine also react. To correct for this, lysine concentrations in the fractions were determined from a calibra- tion curve of growth versus lysine concentration (1) with the

by guest on October 3, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Issue of December 10, 1968 J. W. Payne and C. Gilvarg 6297

growth yields of M-26-26 on hydrolyzed samples described earlier. Because ammonia (formed by amide hydrolysis) contributed only about one-twentieth the absorbance of an equivalent amount of lysine (19), no correction was made for the possible presence of small amounts of amide groups in the pep- tides. The results are shown in Table IV. It can be seen that the average number of amino acid residues in peptide linkage decreases rather regularly with increasing fraction numbers, again indicating that separation has been achieved on the basis of size, and also testifying to the heterogeneity of the peptides. Furthermore, the distribution pattern for amino acid concentra- tions, shown in Column 4, parallels fairly well the 220 rnp absorb- ance profile illustrated in Fig. 2. Fig. 5 shows that a plot of average number of amino acid residues versus fraction number is a fairly smooth curve. It can be seen that exclusion com- mences with peptides that contain an average of slightly more than 6 amino acid residues.

Competitive Ability of Neopeptone Fractions-In an earlier report (4), evidence was presented to indicate that oligopeptides enter E. coli through a single transport system, and competition for entry was demonstrated between a number of different peptides. It was therefore of interest to see whether the heterog- enous peptides present in the initial Neopeptone fractions, although apparently unable to enter E. coli, might be able to compete at the level of the peptide transport system. Fig. 6 shows that when M-26-26 was grown in media supplemented with tetralysine, peptides in the initial fractions were completely

HOURS

FIG. 6. Effect of Neopeptone fractions on growth response of M-26-26 to tetralysine. Tetralysine was present in all cases at a concentration of 0.025 pmole per ml. Numbers refer to Neopep- tone fractions. With Fractions 7, 9, 11, and 13, 0.6-ml samples were used; with all other fractions, l.O-ml samples were used. Growth was measured in Klett tubes at 37” in total volumes of 5 ml.

without effect on the growth rate. In this system it would be anticipated that competition by these Neopeptone fractions should limit the entry of tetralysine and thus decrease the rate of growth (4). From the peptide concentrations shown in Table IV it can be calculated that in the above competition studies, the molar ratio of Neopeptone peptides to tetralysine was about 50: 1. On the other hand, it can also be seen in Fig. 6 that fractions that themselves support the growth of M-26-26, stimulate the growth observed in the presence of tetralysine. This stimulation of growth is to be expected, for the addition of these particular peptides gives an effectively enriched media that contains a large number of nutrient amino acids.

DISCUSSION

The studies described here sought an explanation for earlier observations that higher members of several oligopeptide series are unable to enter E. coli (1, 3). The oligomer against which discrimination is first exhibited varies with the nature of the oligopeptide series, and for this and other reasons, the specula- tion was advanced that the hydrodynamic volume of a peptide might be an important factor that controls entry to the cell. To investigate this idea, an attempt has been made by the use of gel filtration to ascertain the Stokes’ radii of the oligopeptides used in the earlier growth studies. However, before Kd values obtained by gel filtration measurements can be related to molec- ular dimensions, it is a prerequisite that factors other than size do not govern the Kd values that are observed. In the present case, the use of a buffered solution of NaCl as eluent appeared to minimize errors arising from electrostatic interactions between the gel and the peptides. And the data shown in Table II, especially those for the tripeptides lysyldiornithine, diornithylly- sine, and ornithyllysylornithine, support the idea that separation is achieved on the basis of size.

The regular elution sequences shown in Fig. 1 for several series of peptides is good evidence that no abrupt alteration in the structure occurs within the size range examined. This is a particularly important observation, for the possibility existed that with increasing chain length a point would be reached at which a particular oligomer might adopt a new conformation (20), and it might be this structural change rather than the increment in size that prevented it from entering the cell.

In the fractionation of Neopeptone, it would be expected from the known adsorption of aromatic compounds on Sephadex (12, 21) that the elution of peptides that contain aromatic residues would be retarded. Thus, while the distribution of peptide sizes within any fraction would predominantly be within narrow limits, it might be anticipated that to some extent larger retarded peptides might also be found. This possibility may in part explain the observation (Fig. 3) that with fractions that are able to support growth before hydrolysis, the growth observed is frequently less than that obtained with the same fractions after hydrolysis. However, another possibility that may con- tribute would be the inability of intracellular peptidase activity to liberate all lysine and glycine residues. In contrast to re- tardation caused through adsorption, there appears to be no way in which smaller peptides may be eluted prematurely from Sephadex. Thus, it does not seem feasible that the first frac- tions in which growth is detected contain unusually small peptides, and that these are responsible for the observed growth response.

In using Neopeptone as a means of illustrating a cut off for

by guest on October 3, 2020

http://ww

w.jbc.org/

Dow

nloaded from

6298 Peptide Utilization Vol. 243, No. 23

peptides in general, it was assumed that the fractions would contain a diverse mixture of peptides. Various observations substantiate this initial assumption, for example, the elution profile shown in Fig. 2, the calculated size distribution of the peptides shown in Table IV, and the distribution of amino acids seen in the peptide hydrolysates. Further support is obtained if data are taken from the curve shown in Fig. 5 and are replotted (not shown) in a manner similar to Fig. 1, as a function of amino acid residues against log fraction number. The relationship is essentially linear, a result compatible with a very mixed popula- tion of peptides. In this plot, deviation from linearity is observed with peptides containing about 8 or more amino acid residues, and this may be an indication of folding in the peptide backbone. It would be of interest to study the higher oligomers of the series shown in Fig. 1 to see if similar effects would be observed there, although with the glycine peptides the insolubility of the higher oligomers would present a difficulty to such studies.

With the Neopeptone fractions, the cut off which is first observed at about Fraction 14 occurs with peptides that on average contain slightly more than 6 amino acid residues (Fig. 5). This value of approximately 6 residues, corresponding to an average size based on all amino acids is reasonable in view of the earlier observations that with lysine peptides the exclusion begins with 5 residues, and that the presumed cut off with glycine peptides requires more than 6 residues.

In view of the results obtained here it is of interest that a similar type of observation has been reported by Phillips and Gibbs (22). They separated a tryptic digest of casein on Sephadex G-25 and measured the ability of various pooled fractions to stimulate the growth of Streptococcus equisimilis. They demonstrated a definite specificity in the stimulatory abilities of the various pools, and, significantly, the early fractions were poorly stimulatory. In contrast, studies by Pittman, Lakshmanan, and Bryant (23) with Bacteroides ruminicola seem to suggest that both large and small peptides can enter this organism with equal facility. However, it should be noted that B. ruminicola cannot utilize free amino acids, and therefore may have evolved special adaptations to growth on peptides.

The evidence presented in this paper supports the idea that in the process of peptide transport in E. coli discrimination may be exerted on the basis of over-all size, and the obvious question that arises is at which point in the passage across the cell envelope does this discrimination occur. An initial approach to this question has been taken in the following way. Earlier studies have shown that there appears to be only one transport system for oligopeptides in E. coli and therefore competition for entry can be readily demonstrated among many oligopeptides (4). For example, competition between trilysine and tetraglycine was shown with the observation that the growth of a glycine auxotroph on tetraglycine was markedly inhibited by the addi- tion of low concentrations of trilysine to the growth media. The inhibition was considered to arise through trilysine competi- tively limiting the entry of the essential nutrient tetraglycine. In a similar way it was shown that when tetralysine was used, it also could inhibit the growth of the glycine auxotroph on tetraglycine (4). However, in further unreported experiments, pentalysine and hexalysine (peptides that cannot enter E. coli) were tested with the glycine auxotroph and no competitive inhibition could be detected even with high concentrations. It was therefore of interest in the present studies to see whether competition would be demonstrated by the large peptides

present in the initial Neopeptone fractions. Fig. 6 shows that these large peptides are completely without effect upon the growth of a lysine auxotroph on tetralysine. If competition with the tetralysine should occur an inhibition of growth would be expected. It is just conceivable that the observed lack of effect could arise through mutually compensating effects of growth inhibition (through competition) and growth stimulation (from the supply of mixed amino acids). However, the observa- tion that these fractions cause no stimulation in the growth of the lysine auxotroph 011 lysine supplemented media makes this possibility even more remote.

It appears therefore that higher excluded peptides are unable to competitively inhibit the uptake of lower peptides. Assuming that every peptide has the potential to bind to the transport system, this suggests that the discrimination against large peptides is unlikely to be exerted at the transport system itself, for if all peptides are able to attach to the transport system but only those of a certain size can then traverse the cytoplasmic membrane, this should be demonstrated by competition. In contrast, the fact that no competition is observed suggests that peptides of a certain size may not reach the transport system. That is, there exists an additional permeability factor external to the transport mechanism. The simplest conception would envisage the bacterial cell wall acting as a molecular sieve ex- cluding molecules above a certain size. The volume at which a peptide becomes excluded is presumably some measure of the pore size of this sieve. However, there are alternatives to the above suggestion, for example, it could be imagined that the cell wall does allow the peptides to reach the transport system but the micro-environment of the oligopeptide transport system itself prevents peptides of a certain size from adopting that strict spatial conformation presumed necessary for attachment to the transport system. It should not prove difficult to dis- tinguish between these two alternatives for if the idea of a sieve is correct it would be expected that the size at which a peptide is excluded should also be the size at which all other, unrelated molecules are prevented from crossing the cell envelope. On the other hand, if the size restriction is specific for peptides, it would argue for the exclusion occurring in the immediate vicinity of the peptide transport system.

Studies with cells that have undergone treatments to specifi- cally modify their cell envelopes, and studies with mutants that, possess known cell envelope defects should be useful in trying to establish which element or elements of the envelope is responsible for the proposed size discrimination.

Acknowledgment-We wish to thank Mr. Karl Russell-Brown for his competent assistance with certain aspects of the experi- mental work.

REFERENCES

1. GILVARG, C., AND K.~TCHALSKI, E., J. Viol. Chem., 240, 3093 (1965).

2. LOSICK, R., AND GILVARG, C., J. Biol. Chem., 241, 2340 (1966).

3. P.;Y:,,J. W., .~ND GILVARG, C., J. Biol. Chem., 243, 335

4. P.~YNE, J. W., J. Biol. Chem., 243, 3395 (1968). 5. DAVIS, B. D., Nature, 169, 534 (1952). 6. DEWEY, D., AND WORK, E., Arafure, 169, 533 (1952). 7. DAVIS, B. D., AND MINGIOLI, E. S., J. Bacterial., 60, 17

(1950). 8. LAURENT, T. P., AND KILLANDER, J., J. Chromatogr., 14, 317

(1964).

by guest on October 3, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Issue of December 10, 1968 J. W. Payne and C. Gdvarg 6299

9. FLODIN, P., Pharmscia, Uppsala, Sweden, 1962. 20. GOODMAN, M., SCHMITT, E. E., LISTOWSKY, I., BOARDMAN, 10. PORATH, J., Pure Appl. Chem., 6, 233 (1963). F., ROSEN, I. G., AND STAKE, M. A., in M. A. STAAMANN 11. SQUIRE, P. G., Arch. Biochem. Biophys., 101, 471 (1964). 12. GELOTTE, B., J. Chromatogr., 3, 330 (1960).

(Editor), P&amino acide, potypeptidee, and proteins, Uni-

13. MABSDEN, N. V. B., Ann. N. Y. Acad. Sci., 126, 428 (1965). versity of Wisconsin Press, Madison, Wisconsin, 1962, p.

14. EAKER, D., AND PORATH, J., Separ. Pci., 2, 507 (1967). 195.

16. GRANATH, K., AND FLODIN, P., Makromol. Chem., 46, 160 21. BRETTHAUER, R. K., AND GOLICHOWSICI, A. M., Biochim.

(1961). Biophys. Acta, 166, 549 (1967).

16. ANDREWS, P., Biochem. J., 96, 595 (1965). Z. PHILLIPS, A. W., AND GIBBS, P. A., Biochem. J., 81, 551

17. CARNEGIE, P. R., Biochena. J., 96, 9P (1965). (1961). 18. WA-Y, S. G., AND WATSON, J., J. Chem. Sot., 475 (1963). 23. PITTMAN, K. A., LAKSBMANAN, S., AND BRYANT, M. P., J. 19. DUBIN, D. T., J. Biol. Cha., 226, 783 (1960). Bacterial., 99, 1499 (1967).

by guest on October 3, 2020

http://ww

w.jbc.org/

Dow

nloaded from

John W. Payne and Charles GilvargEscherichia coliSize Restriction on Peptide Utilization in

1968, 243:6291-6299.J. Biol. Chem. 

  http://www.jbc.org/content/243/23/6291Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/243/23/6291.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on October 3, 2020

http://ww

w.jbc.org/

Dow

nloaded from