Posttranslational modification of 6-phosphofructo-1-kinase as an important feature of cancer metabolism

Abstract

Background: Human cancers consume larger amounts of glucose compared to normal tissues with most being converted and excreted as lactate despite abundant oxygen availability (Warburg effect). The underlying higher rate of glycolysis is therefore at the root of tumor formation and growth. Normal control of glycolytic allosteric enzymes appears impaired in tumors; however, the phenomenon has not been fully resolved.

Methodology/principal findings: In the present paper, we show evidence that the native 85-kDa 6-phosphofructo-1-kinase (PFK1), a key regulatory enzyme of glycolysis that is normally under the control of feedback inhibition, undergoes posttranslational modification. After proteolytic cleavage of the C-terminal portion of the enzyme, an active, shorter 47-kDa fragment was formed that was insensitive to citrate and ATP inhibition. In tumorigenic cell lines, only the short fragments but not the native 85-kDa PFK1 were detected by immunoblotting. Similar fragments were detected also in a tumor tissue that developed in mice after the subcutaneous infection with tumorigenic B16-F10 cells. Based on limited proteolytic digestion of the rabbit muscle PFK-M, an active citrate inhibition-resistant shorter form was obtained, indicating that a single posttranslational modification step was possible. The exact molecular masses of the active shorter PFK1 fragments were determined by inserting the truncated genes constructed from human muscle PFK1 cDNA into a pfk null E. coli strain. Two E. coli transformants encoding for the modified PFK1s of 45,551 Da and 47,835 Da grew in glucose medium. The insertion of modified truncated human pfkM genes also stimulated glucose consumption and lactate excretion in stable transfectants of non-tumorigenic human HEK cell, suggesting the important role of shorter PFK1 fragments in enhancing glycolytic flux.

Conclusions/significance: Posttranslational modification of PFK1 enzyme might be the pivotal factor of deregulated glycolytic flux in tumors that in combination with altered signaling mechanisms essentially supports fast proliferation of cancer cells.

Figure 1. PFK1 activities after limited proteolytic degradation of native rabbit PFK1 by Proteinase K.

Activities of the native PFK1 isolated from rabbit muscle after limited proteolysis by Proteinase K (dark) and untreated native enzyme (light) as measured in a system containing 5 mM citrate. Data are representative of three independent measurements and are presented as means ± standard deviation.

Figure 2. Western blots of homogenates of four tumorigenic cell lines, HEK immortalized cells and lymphocytes immunostained with PFK-M anti-bodies.

Western blots of four tumorigenic cell lines (above) showed the presence of fragments of different lengths, with a fragment of 47 kDa regularly present, while no native 85-kDa PFK1 could be observed. In non-neoplastic cell lines (below) (HEK cells), the native PFK1 forms were predominant, while in normal lymphocytes isolated from peripheral human blood only a single protein band was detected corresponding to 85 kDa native PFK1. In the Western blot of lymphocytes two volumes of cell lysate were applied to the gel: 10 µl (left), and 20 µl (right).

Figure 3. Western blots of B16-F10 cells growing as a tissue culture and B16-F10 cells that formed a tumor in mouse, immunostained with PFK-M anti-bodies.

No native PFK1 enzyme was detected in the cells growing in a tissue culture, while in a tumor, a strong signal corresponding to the native enzyme was present. Shorter fragments were detected in both homogenates with a 47 fragment present in individually growing cells and a 45 kDa fragment present in a tumor tissue.

Figure 4. Growth of two E.coli transformants encoding two different human shorter PFK-M fragments.

Two E.coli transformants encoding Fragment 4 (⧫) and Fragment 9 (□) were able to grow in supplemented glucose minimal medium. No growth of the parental strain, RL 257, carrying the pALTER-Ex-1 plasmid with no inserted gene (•) could be detected. Data are representative of three independent measurements and are presented as means ± standard deviation.

Figure 5. PFK1 activity of a recombinant shorter PFK-M fragment and native PFK-M enzyme with respect to some inhibitors.

In figure 5A relative specific PFK1 activities detected in the homogenate of the transformant encoding Fragment 9 (□ ) and with native PFK-M enzyme isolated from E.coli transformant (○) are shown, that were measured at increasing concentrations of ATP. In figure 5B specific PFK1 activities were measured in the homogenate of the transformant encoding Fragment 9 without inhibitor (□), in the presence of 5 mM Na₃-citrate (◊), and with 5 mM Na-lactate (Δ). All measurements were conducted with 0.5 mM of F6P. Data are representative of at least three independent measurements and are presented as means ± standard deviation.

Figure 6. PFK1 activity of a recombinant shorter PFK-M fragment and native PFK-M enzyme with respect to F-2,6-BP as an activator.

In figure 6A F6P saturation curves of the isolated native PFK-M enzyme with (○) and without (◊) 4 µM of F-2,6-BP are presented. In figure 6B F6P saturation curves detected in the homogenate of the transformant encoding Fragment 9 with (□) and without (Δ) 4 µM of F-2,6-BP are shown. The measurements were conducted with 1 mM of ATP. In both graphs relative specific activities are shown. Data are representative of at least three independent measurements and are presented as means ± standard deviation.

Figure 7. Growth, glucose consumption and lactate excretion by the stably transfected HEK cells synthesizing PFK1 Fragment 9, the native PFK-M and HEK cells with empty plasmid.

In figure 7A growth of Flp-In T-Rex HEK 293 cells with integrated hpfkMFrg9 (hpfkMFrg9/HEK - ▪) encoding the Fragment 9; integrated hpfkM (hpfkM/HEK - •) encoding the native PFK-M enzyme; and cells with integrated empty plasmid (HEK - ◊) are presented in logarithmic mode. In figure 7B glucose consumption by transfectants normalized to 1 million cells is shown. In figure 7C lactate production recalculated to 1 million cells is presented. Identical symbols for individual transfectants are used in all figures. Data are representative of three independent measurements and are presented as means ± standard deviation.