Protein:Protein Interactions in the Cytoplasmic Membrane Influencing Sugar Transport and Phosphorylation Activities of the E. coli Phosphotransferase System

The multicomponent phosphoenolpyruvate-dependent sugar-transporting phosphotransferase system (PTS) in Escherichia coli takes up sugar substrates and concomitantly phosphorylates them. We have recently provided evidence that many of the integral membrane PTS permeases interact with the fructose PTS (FruA/FruB) [1]. However, the biochemical and physiological significance of this finding was not known. We have carried out molecular genetic/biochemical/physiological studies that show that interactions of the fructose PTS often enhance, but sometimes inhibit the activities of other PTS transporters many fold, depending on the target PTS system under study. Thus, the glucose, mannose, mannitol and N- acetylglucosamine permeases exhibit enhanced in vivo sugar transport and sometimes in vitro PEP-dependent sugar phosphorylation activities while the galactitol and trehalose systems show inhibited activities. This is observed when the fructose system is induced to high levels and prevented when the fruA/fruB genes are deleted. Overexpression of the fruA and/or fruB genes in the absence of fructose induction during growth also enhances the rates of uptake of other hexoses. The β-galactosidase activities of man, mtl, and gat-lacZ transcriptional fusions and the sugar-specific transphosphorylation activities of these enzyme transporters were not affected either by frustose induction or fruAB overexpression, showing that the rates of synthesis and protein levels in the membrane of the target PTS permeases were not altered. We thus suggest that specific protein-protein interactions within the cytoplasmic membrane regulate transport in vivo (and sometimes the PEP-dependent phosphorylation activities in vitro of PTS permeases) in a physiologically meaningful way that may help to provide a hierarchy of preferred PTS sugars. These observations appear to be applicable in principle to other types of transport systems as well.

247 and NAG uptake rates 4-5-fold, but increased αMG and 2DG uptake rates 7 and 11-fold, 248 respectively, compared to the control strain bearing the empty pMAL plasmid (Fig 2A). When 249 the same experiment was conducted in the wild type genetic background, qualitatively similar 250 results were obtained, but enhancement of the uptake rates were diminished except for the 251 mannose system, which was increased (Fig 2B). Only in the WT background did overexpression 252 of fruA or fruB increase the 2DG uptake rate. Surprisingly, in both cases, Tre and Gat uptake 253 rates were not appreciably depressed. When fruB was overexpressed in either the triple mutant 254 (TM; Fig 2C) or the WT genetic background (WT; Fig 2D)  The three transport systems showing largest responses to fructose induction in wild type 280 cells were the systems encoded by the mannitol (mtl), galactitol (gat) and mannose (man) 281 operons (see Fig 1). Therefore, in order to determine if these effects reflected changes in operon 282 transcription or EII activities, transcriptional lacZ fusions were constructed to operons encoding 283 these three PTS systems (mtlA-lacZ; gatY-lacZ and manXYZ-lacZ). These were used in studies to 284 determine the effects of fructose in the growth medium prior to transport rate determinations.
285 The results are presented in Table 3. It can be seen that the presence of fructose in the growth 286 medium, or the overexpression of specific fru operon genes, had no apparent effect on the 287 induction of mtl, gat, and man operon expression. These results imply that the effects observed in 288 Figs 1 and 2 reflect the enzyme II activities and not their syntheses. This conclusion is 289 substantiated by the results obtained when the transphosphorylation reactions were studied (see 290 Table 5 below).

320
The mtlA gene was inactivated with a kanamycin resistance gene insertion or was deleted 321 (∆mtlA), and the crude extracts were compared with the wild type. Crude extracts were assayed 322 for [ 14 C]mannitol phosphorylation after growth with or without fructose, and the activity was 323 reduced to a few percent of the wild type level in the insertion or deletion mutant (S22 Table), 324 showing that mannitol cannot be phosphorylated at an appreciable rate by other PTS Enzyme II 325 complexes under these conditions (data not shown).
326  348 Moreover, while mannitol phosphorylation responded dramatically to the inclusion of FruB in 349 the assay mixture, a lesser activation was observed when N-acetylglucosamine was the sugar 350 substrate (Fig 4). Activation by FruB but not HPr showed that FruB activation is not due to the 351 activity of the FPr domain of FruB but is dependent on the presence of FruA (Table 4).  In view of the in vitro phosphorylation results described above and presented in Table 4 369 and Fig 4, Table).   Table 2). Because these PTS enzyme-transporters can be assayed in vitro as well as in vivo, 439 we initially attempted to investigate the consequences of these interactions before moving on to 440 examine the consequences of interactions between other PTS enzymes, and finally, between PTS 441 and non-PTS integral membrane proteins. Most striking was the observation that many of the 442 PTS permeases interact with the fructose PTS (FruA/FruB), more than with any other PTS 443 Enzyme II complex (Table 2) Table), digested by 539 appropriate restriction enzymes, then ligated into the same sites of pUT18 individually. In each 540 of these resultant plasmids, the target structural gene with no stop codon is inserted immediately 541 upstream of the N-terminus of T18 in pUT18, creating a single hybrid gene that encodes the 542 target protein at N-terminus and T18 domain at C-terminus. These plasmids are denoted as 543 pUT18-fruA, pUT18-fruB, pUT18-gatC, pUT18-nagE, pUT18-treB, pUT18-mtlA, pUT18-galP, 544 pUT18-lacY and pUT18-pheP (SX Table).

545
Similarly fruA, fruB, mtlA and nagE were fused to the N-terminus of T25 domain in 546 pKNT25 individually. Each of these resultant plasmids carries a single hybrid gene encoding the 547 target protein at N-terminus and T25 domain at C-terminus. These plasmids are denoted as 548 pKNT25-fruA, pKNT25-fruB, pKNT25-mtlA and pKNT25-nagE, respectively (SX Table). 558 mtlA were generated from the parental strain (E. coli K-12 strain BW25113) using a standard 559 method as described in [53]. Briefly, to construct each mutant, a kanamycin resistance gene (kn), 560 flanked by the FLP recognition site (FRT), amplified from the template plasmid pKD4 using a pair 561 of specific mutation oligos (SY Table), was first substituted for the target gene or operon. Where 562 indicated, the kn gene was subsequently eliminated (leaving an 85-bp FRT sequence) using 563 plasmid pCP20 that bears the FLP recombinase. The replacements of the target genes/operons by 564 the FRT-flanking kn gene and the subsequent removal of the kn gene were confirmed by colony For the uptake assays, the stock solution of each radioactive substrate was used at 5 648 µCi/µmole for 14 C while radioactive tritium was used at 30 µCi/µmole, and in all cases, the 649 substrate concentration of the stock solution was 1 mM. Whenever stated, uptake assays were 650 performed in the presence and absence of certain added non-radioactive sugars at the specified 651 test concentrations and conditions. For all uptake assays, the prepared bacterial suspension of the 652 test strain was used within a time period not exceeding 12 h. The assay mixture contained 900 µl 653 of a bacterial cell suspension of specified OD 600nm , 50 µl of 1 M arginine (pH 7), and 20 µl of the 654 1 mM stock radioactive substrate. The volume was brought to 1 ml with a cold sugar solution 655 and/or 50 mM Trizma-maleate buffer, pH 7, containing 5 mM MgCl 2 .

656
The final concentration of the radioactive sugar substrate in the assay mixture was 20 µM 657 for uptake assays and 10 µM for phosphorylation assays unless otherwise stated. The uptake 658 assays were carried out in a shaking water bath at 37 o C for 5-10 min followed by immediate