3.2 Hydrolysis of LplA·Lip-AMP complex and effectors
During the quantitative analysis of the lipoylation process, we noticed that the ratio of AMP/Hlip (Fig. 3b) is often much higher than the theoretical value which would be expected to be 1.0 from the stoichiometry of the lipoylation reaction (Fig. 1), indicating that a part of Lip-AMP bound on LplA might be hydrolyzed during the transfer of Lip-AMP to Hapo. This has not been reported before and is therefore examined in this work in more detail.
Fig. 4a depicts the kinetics of the formation of Hlip, AMP and Lip-AMP in a typical assay of the lipoylation process with 10 µM initial Hapo. It is noticed that AMP increased faster than Hlip, especially at the beginning of the assay. It is also interesting to note that there is a rapid accumulation of Lip-AMP, indicating that the first reaction step is much faster than the second reaction step. This was first thought to be the main reason for the hydrolysis of Lip-AMP, but turned out to be not true.
As summarized in Table 2, the formation rate of Lip-AMP is over 400 times higher than that of Hlip in the overall reaction assay. When the two reaction steps were separately assayed, the formation rate of Lip-AMP (Step 1) is still over 245 times faster than that of Hlip (Step 2) and the hydrolysis rate is about 1.24 times of that of the Hlip formation under the experimental conditions. A separate assay of Step 1 revealed a similar fast formation rate of Lip-AMP, but a neglectable hydrolysis of Lip-AMP in comparison to that in the overall reaction, suggesting that the accumulation of Lip-AMP is not the primary reason for its hydrolysis. A separate assay of Step 2 resulted in formation rates of AMP and Hlip both about 40% higher than those in the overall reaction. This might be because of the fact that Lip-AMP used for the assay of Step 2 was obtained as the reaction product of Step 1 and contained therefore also some residual lipoic acid and ATP. The residual ATP was between 127-173 µM in the different assays and lipoate in the range of 135-182 µM. Interestingly, the ratio of the formation rates of AMP and Hlip remained at 1.24. With the increase of Hapo concentration, the ratio of the formation rates of AMP and Hlip can be further increased (data not shown).
To better assess the effect of H protein on the hydrolysis of the Lip-AMP complex, we added HK64A in the overall reaction mixture. As shown in Fig. 4b and 4c, an enhanced hydrolysis (formation of AMP) was also observed with the addition of HK64A. It is also clear from Fig. 4c and 4d that HK64A increased the formation rates of AMP and Hlip as well, but not as significantly as it to the hydrolysis of Lip-AMP. The promotion of HK64A on the hydrolysis of Lip-AMP complex is dose-dependent (Fig. 5a and 5b). Since no Hapo was added in these experiments, they represent a situation of reaction Step 1 under the influence of HK64A. While the formation of Lip-AMP nearly finished after about 3 hours of assay, its hydrolysis continued as shown by the increased formation of AMP. After an assay time of 7 hours, the experiments with 10 μM and 100 μM initial HK64A concentration produced 37% and 492% more AMP, respectively, compared to the control experiment.
The effect of Hlip on the kinetics of Hlip formation in the overall reaction was also studied. With the addition of 100 µM Hlip the formation rate of Hlip is increased from 0.066 to 0.234 µmol/min/mg (Fig. 4d). In a separate assay for reaction Step 1 (Fig. 6a), a much higher AMP formation rate was also observed under the influence of Hlip.
Taking the results together, it can be stated that the addition of various forms of H protein enhances the hydrolysis of Lip-AMP. It is noted that the hydrolysis is especially significant at low level of Hapo. As shown in Fig. 6b the amount of AMP formed at Hapo below 50 µM was much higher than Hlip formed. With the increase of Hapoto above 100 µM the ratio of AMP and Hlip is close to 1.0. This suggests that the hydrolysis of Lip-AMP competes with the lipoate transfer reaction.
3.3 A conceptual model of H protein lipoylation and competitive hydrolysis
Lipoylation of H protein plays a key role in the functionality of glycine cleavage system. Fujiwara et al . (Fujiwara et al., 2010) elaborated the crystal structures of LplA itself and its complexes with lipoate, Lip-AMP and octyl-AMP-Hapo, which contributed greatly to understanding the reaction mechanism of LplA-catalyzed H protein lipoylation as outlined in Fig.1. In general, the lipoylation process is considered to be composed of two steps: (1) the adenylation of lipoate catalyzed by LplA, forming thereby the complex Lip-AMP-LplA, and (2) the binding of Hapo to the complex, resulting in a transfer of the lipoate to Hapo which is subsequently released as Hlip. For the first time the kinetics of the overall lipoylation process and the two individual steps were studied in this work with the help of newly developed simple and efficient HPLC analytic methods to quantitatively determine the involved educts and products. The experimental results revealed unusual kinetic behavior of the overall and individual processes. Overall, no typical saturation kinetic behavior can be found with Hapo as a substrate of the lipoylation process (Fig.2). As summarized in Table 2, the specific rate of adenylation in the first step is over 420 times faster than that of lipoate transfer in the second step under the given experimental conditions. This leads to the accumulation of a high amount of the intermediate Lip-AMP during the lipoylation process in vitro . Furthermore, this study discovered two new phenomena of the H protein lipoylation process:
  1. Whereas Lip-AMP itself is quite stable under the experimental conditions, the presence of protein H either in the forms of Hapo, Hlip or HK64Acan cause the hydrolysis of Lip-AMP bound to LplA. Enzymatic assays of the individual steps clearly showed that the hydrolysis proceeds in competition with lipoate transfer in the second step of the lipoylation process.
  2. Both the rates of protein H lipoylation and Lip-AMP hydrolysis can be enhanced by Hlip and HK64A. Obviously, Hapo acts not only as a substrate, but also as an activator for its lipoylation. For this reason, it was not possible to accurately determine the so-called km andv max values of LplA towards Hapo as a substrate. In fact, the specific lipoylation rate of Hapo increased nearly linearly with Hapo concentration even up to 400 µM studied in this work.
We propose here an extended model of H protein lipoylation as presented in Fig.7 to account for these new experimental observations. We hypothesize that the presence of H protein can cause a conformational change of LplA (or LplA bound with Lip-AMP), which allows not only Hapo but also H2O to attack the Lip-AMP complex easily, therefore, leading to an accelerated hydrolysis in accompany with lipoylation. As a result, the formation of AMP increased with the addition of various forms of H protein, and the production of Hlip linearly increased with increased concentration of Hapo and Hlip, because they can help to maintain the easily attackable conformation of LplA, rather than causing a substrate inhibition or product inhibition.
When the mutant H protein HK64A is present and the C-terminal domain of LplA is opened, no amine group could attack but only the hydroxyl group of water. Based on the structure of LplA, in the first step, the C-terminal domain of LplA adopts a binding conformation (Fig. 1) which can form a Lip-AMP intermediate bound to LplA. The intermediate Lip-AMP or H protein induces a LplA rearrangement from the β3 strand to the α2 helix, including a cleavage of the charge-charge interaction between the N-terminal domain (residue Arg47) and the C-terminal domain (residue Glu291). As a consequence, the free C-terminal domain rotates by 180°, adopting a stretched conformation with a “closed” C-terminal domain (Fujiwara et al., 1992). When the C-terminal domain is in the closed form, Lip-AMP cannot combine with LplA, further protecting Lip-AMP from the attack of H2O, and yield less hydrolysis product AMP. And the open form of LplA will drive the amine group on the lysine residue of Hapo to attack the carbonyl group of Lip-AMP, yielding Hlip. At the same time, the hydroxyl group of H2O would also attack the carbonyl group competitively, yielding the hydrolysis product of lipoic acid and AMP.
This hypothesis can explain our experimental results very well. First, few AMP was produced in Step 1, since no H protein was added in the reaction system and the C-terminal domain of LplA is closed. As a result, less H2O could attack the Lip-AMP complex. Second, the hydrolysis by the attack of H2O competes with the lipoylation by the attack of Hapo. When Hapo was increased from 1 μM to 50 μM, the ratio between AMP and Hlip amplified, since more H2O could attack the carbonyl group of Lip-AMP rather than lipoylation of Hapo. However, with further increase of Hapo concentration, the attack from Hapodominates the reaction, rather than the attack from H2O. Thus, the ratio between AMP and Hlip returned to nearly 1, as Hapo increased from 50 μM to 200 μM (Fig. 6). The exact nature of the conformational change of LplA induced by protein H is not known at this stage. In view of the importance of protein lipoylation for several essential enzymes of cellular metabolism it is certainly worth further study, ideally with new experimental tools for studying protein interactions under dynamic conditions. It would be also interesting to examine if similar phenomena as reported here could be observed for relevant enzyme systems such as pyruvate dehydrogenase and 2-oxoglutarate complexes.