Study on the phosphorylation of NAD+-specific isocitrate dehydrogenase in the pathogenic bacteria Stenotrophomonas maltophilia and Xanthomonas sacchari Authors: Qian Tian, Yongmei Zhang, Chunxu Pan, Linxiao Wang, Na Qin, Zhen Jiang, Chenxi Li, Zhengyu Cao, Guoping Zhu, Peng Wang Published: 2025 ABSTRACT AbstractIsocitrate dehydrogenases (IDHs) regulate the distribution of carbon flux between the TCA cycle and glyoxylate shunt through reversible phosphorylation that influences pathogen virulence. Current studies only indicate that NADP+-specific IDHs (NADP-IDHs) can be phosphorylated. Whether NAD+-specific IDHs (NAD-IDHs) are susceptible to phosphorylation remains unknown. In this study, two NAD-IDHs and their regulation by phosphorylation from Stenotrophomonas maltophilia and Xanthomonas sacchari were characterised for the first time. Ser80 was identified by mass spectrometry as the phosphorylation site in SmIDH, which was functionally validated through site-directed mutagenesis. Acetate induction led to an approximately 78% decrease in the ratio of IDH/ICL specific enzyme activity, consistent with phosphorylation-mediated regulation. By modifying key recognition regions in XsIDH and XsAceK, the phosphorylation efficiency of XsIDH was improved, revealing evolutionary insights. It may enable further investigations for the new antibacterial drug targets in S. maltophilia and X. sacchari.Keywords: Isocitrate dehydrogenase, Stenotrophomonas maltophilia, Xanthomonas sacchari, Phosphorylation, glyoxylate shunt Introduction Isocitrate dehydrogenase (IDH) is a significant member of the β-decarboxylase dehydrogenase family and one of the crucial rate-limiting enzymes in the tricarboxylic acid (TCA) cycle. It catalyses the oxidative decarboxylation of isocitrate (ICT) to produce α-ketoglutarate and CO2, and reduces NAD(P)+ to NAD(P)H. The carbon scaffold for cellular anabolism is supplied by the generated α-ketoglutarate primarily, and the reducing power for that is supplied by the generated NADPH, which is crucial for maintaining the redox balance within cells. In our previous research1, we have redefined the phylogenetic classification of IDHs into three subfamilies: Type I, Type II, and Type III. And given that IDH can utilise NAD+ or NADP+ as the coenzyme, this protein family can also be categorised into two types: NAD+-dependent IDH (NAD-IDH, EC 1.1.1.41) and NADP+-dependent IDH (NADP-IDH, EC 1.1.1.42)2. The phylogenetic analysis showed that NAD-IDH is an ancient phenotype, while the NADP-IDH phenotype is an adaptation to the energy-poor sources (such as acetate and aliphatic acid, etc.)3. In the process of evolution, bacteria need a large amount of reducing power, NADPH, to reduce the highly oxidised carbon source to maintain survival. Under such selective pressure, the preference of IDH coenzyme specifically evolved from NAD+ to NADP+3. The activity of IDH can be regulated by reversible phosphorylation in vivo, which is specifically catalysed by IDH kinase/phosphatase (AceK)4, and thus regulates the distribution of carbon flux in the TCA cycle and glyoxylate shunt in cells5. At present, the phosphorylation of EcIDH has been studied in detail6–11. When E. coli grows on poor carbon sources, its IDH competes with isocitrate lyase (ICL) to bind their common substrate ICT. As ICL has a much lower affinity for the substrate than IDH does, the activity of IDH must be reduced, allowing for an increase in intracellular ICT concentration before carbon flux flows to the glyoxylate shunt. And phosphorylation is one way to reduce the activity. EcIDH can be phosphorylated by AceK at the Ser113 site near the active pocket. When it happens, the insertion of a negatively charged phosphate group into the active site of EcIDH will induce the electrostatic repulsion between EcIDH and the substrate ICT, and the binding of them will be impaired, resulting in the loss of activity6,7. In addition to EcIDH, other prokaryotic IDHs can also be phosphorylated by AceK12–15 or other kinds of kinases and phosphatases14,16. The eukaryotic IDHs have been confirmed can also be phosphorylated17–23 under certain circumstances. However, they’re unexceptionally all NADP-IDHs. A previous study has reported an in vitro simulated phosphorylation of the NAD-IDH from Streptococcus mutans (SmmIDH), which confirmed that the phosphorylated Ser102 leads to the perishing of SmmIDH24. However, no coding gene homologous to aceK has been found in its genome. Therefore, even though it has been more than 40 years since the first report of IDH phosphorylation11, there is still no report of the real phosphorylation of NAD-IDH. It raises two questions: Does the NAD-IDH phosphorylation exist in nature? And if so, is there any difference between the phosphorylation of NAD-IDH and NADP-IDH? It was found that AtIDH, different from EcIDH, was a homodimeric NAD-IDH, which has a highly similar structure to EcIDH. But there is no report about the phosphorylation of AtIDH currently. At the same time, Bacillus subtilis IDH (BsIDH), as a homodimeric NADP-IDH like EcIDH, though, is also a poor substrate for AceK25. Studies12,26 have shown that it is their C-d loop and α-helix region that prevent the recognition of IDH by AceK, disabling the phosphorylation. Therefore, from another perspective, whether homodimeric IDH can be phosphorylated by AceK may not be correlated with the coenzyme specificity, but mainly depends on whether the IDH structure can be recognised by AceK. That said, the homodimeric NAD-IDH has the possibility to be phosphorylated as well. The glyoxylate shunt is a metabolic branch widely existing in prokaryotes. IDH acts as the key enzyme at the branch point between the TCA cycle and the glyoxylate shunt. By regulating the reversible phosphorylation of IDH, the distribution of carbon sources between the TCA and the glyoxylate shunt can be controlled. Thus, phosphorylation is a key factor in regulating bacterial virulence27. A study28 on Mycobacterium tuberculosis has found that when surviving in poor carbon sources or oxygen-deprived environments, it will express ICL to initiate the glyoxylate shunt. And the glyoxylate shunt is confirmed to play a significant role in the pathogenicity of pathogenic bacteria29–32. By inhibiting the activity of glyoxylate shunt-related enzymes, the infection ability of pathogenic bacteria can be reduced or even lost30. Previous investigations into glyoxylate shunt and bacterial pathogenicity have predominantly focused on the enzymes malate synthase (MS) and ICL. In contrast, IDH and AceK, functioning as the gatekeeping enzymes at the metabolic branch point between the glyoxylate shunt and TCA cycle, remain underexplored in this context. In this study, both of the pathogenic bacteria have the conditions to form the glyoxylate shunt. Stenotrophomonas maltophilia and X. sacchari are both non-fermenting gram-negative pathogens, which have brought about concern worldwide. S. maltophilia can produce various extracellular enzymes, which enable it firmly adhere to medical implants and catheters to form biofilms and be resistant to host immune defense and different antibacterial drugs33. It has led to infections such as bacteraemia34, respiratory infections35, liver infection36, medical implant infections37 and so on. Therefore, it is of great significance to develop new drug targets against this pathogen. X. sacchari belongs to the genus Xanthomonas, which has caused diseases in various plant hosts and substantial economic losses worldwide38–40. They can invade host vascular and mesophyll tissues, leading to lesion formation and subsequent diseases such as bacterial wilt and black rot, ultimately resulting in host plant mortality41. Current management strategies predominantly rely on chemical control methods, which often cause irreversible damage to local ecosystems, ultimately leading to ecological degradation42. Our work may be helpful to explore their drug resistance mechanism and pathogenic factors by studying the mechanism of carbon distribution at the branch point, and thus to find new therapies for pathogenic bacterial infections. Herein, the biochemical properties and phosphorylation regulation of IDHs from S. maltophilia and X. sacchari were identified. This work aims to further understand the mechanism of IDH phosphorylation and to explore the evolution of IDH and AceK preliminarily. Results Bioinformatics analyses To illustrate the evolutionary relationship among SmIDH, XsIDH, and other IDHs, as well as AceKs, the phylogenetic trees were constructed (Figure 1). The results showed clearly that SmIDH and XsIDH together with several other NAD-IDHs, such as AtIDH and Xanthomonas campestris pv. campestris IDH1 (XccIDH1), belong to the Type I NAD-IDHs subfamily. SmAceK and XsAceK bear a significantly closer phylogenetic relationship with XccAceK than with EcAceK, suggesting some differences may occur in the action mechanism of AceK between S. maltophilia or X. sacchari and E. coli. Phylogenetic analysis of IDHs (A) and AceKs (B) from different species. SmIDH and XsIDH are indicated by triangles (▲), SmAceK and XsAceK are indicated by stars (★). The multiple sequence alignment (Figure 2) showed that XccIDH1 shared the highest sequence identity with XsIDH (93.4%), followed by SmIDH (86.8%), while the others only shared no more than 40% sequence identity with SmIDH or XsIDH. Furthermore, their corresponding sites to the phosphorylation site Ser113 of EcIDH were all Ser-conserved. And an insert region of about 13 amino acids of AtIDH or BsIDH can be easily seen in the graph compared with the others. Another alignment (Figure 3) showed that 13 sites were not conserved in the SRL compared with EcAceK, which implied the possible differences in phosphorylation mechanism between them. Sequence alignment of SmIDH and XsIDH. The high-resolution structures of EcIDH (PDB code 9ICD) and BsIDH (PDB code 1HQS) were downloaded from the PDB database. The conserved phosphorylation site is indicated by a star (★). Structure-based alignment of EcAceK, XsAceK, and SmAceK amino acid sequences. The high-resolution structure of EcAceK (PDB code 3EPS) was downloaded from the PDB database. Expression, purification, and identification of wild-type and mutant enzymes The recombinant enzymes were heterologously expressed in E. coli Rosetta (DE3) and then purified. All of the purified enzymes exhibited a single band respectively (Figures 4, S1, and S2), consistent with their theoretical molecular weights. It was further revealed from the graph (Figure 4) that SmIDH was eluted as a single and symmetric peak corresponding to a molecular weight of approximately 60 kDa, and that of XsIDH was 56.4 kDa. Therefore, they may exist as homodimers in solution. And the slightly smaller molecular mass was probably due to the extremely tight binding between the two subunits. Molecular mass determinations of SmIDH (A) and XsIDH (B) by gel filtration chromatography and SDS-PAGE. The flow rate of the gel filtration chromatography was 0.7 ml min−1, and the absorbance of the proteins in the fractions was monitored at 280 nm. The evolution volume (Ve) of the recombinant SmIDH and XsIDH was 14.76 ml and 14.73 ml, respectively. The right inset panel shows the protein purity detected by 12% SDS-PAGE. The biochemical properties SmIDH and XsIDH are both NAD+-dependent enzymes without detectable NADP+-dependent activity (Table 1), so the subsequent biochemical properties were determined using NAD+ as a coenzyme. The optimal pH values of SmIDH (Figure 5A) and XsIDH (Figure 5B) were similar to the optimal pH of EcIDH (8.0, Mg2+)43 and XccIDH1(8.5, Mg2+)44. SmIDH exhibited a relatively high optimal temperature at 50–60 °C (Figure 5C), while XsIDH exhibited the highest activity at only 40 °C (Figure 5D). To further investigate the thermostability of these enzymes, the heat-inactivation studies were conducted. The activities of SmIDH (Figure 5E) and XsIDH (Figure 5F) dropped rapidly above the temperatures, which were slightly higher than that of EcIDH43, suggesting the better thermostability of them. The effects of metal ions and their combinations on the activities of SmIDH and XsIDH were measured (Table S1). Their favourite activator was Mn2+, followed by Mg2+ without exception. The recombinant SmIDH and XsIDH exhibited a complete preference for NAD+, consistent with the previous bioinformatic analysis. As is shown in Table 1, their catalytic efficiencies (kcat/Km) for NAD+ were similar but slightly higher than those of other type I NAD-IDHs, for instance, O. tauri IDH45 and S. mutans IDH24. Table 1.Coenzyme and substrate kinetic parameters of SmIDH, XsIDH, and other IDHs.EnzymeNAD+NADP+IsocitrateKm(μM)kcat(s−1)kcat/Km(μM−1 s−1)Km(μM)kcat(s−1)kcat/Km(μM−1 s−1)Km(μM)kcat(s−1)kcat/Km(μM−1 s−1)SmIDH (Mn2+)141.2 ± 12.178.6 ± 8.30.56–––25.7 ± 2.035.2 ± 2.21.37SmIDH (Mg2+)100.1 ± 7.145.0 ± 4.90.45–––70.0 ± 1.349.6 ± 1.70.71XsIDH (Mn2+)131.7 ± 11.691.5 ± 1.40.69–––5.48 ± 0.11100.9 ± 6.0818.24XsIDH (Mg2+)152.5 ± 7.1149.67 ± 0.950.33–––1.93 ± 0.02942.52 ± 2.4722.03OtIDH (Mn2+)45226590.2613354170.0051–––SmuIDH (Mn2+)24154560.36–––751241.65EcIDH (Mg2+/Mn2+)46,4747003.220.000691780.54.724.251.82.1BsIDH (Mg2+)25–––14.5 ± 2.2––5.9 ± 0.9––AbIDH1 (Mn2+)48–––46.6 ± 6.160.1 ± 3.31.350.5 ± 3.250.41.0AbIDH2 (Mn2+)1–––94 ± 636.9 ± 1.20.3921 ± 339.2 ± 2.11.9PmIDH (Mn2+)49––––––33.53 ± 1.54102.98 ± 5.823.07CgIDH (Mn2+)50190009.30.00054.087.022.058717.4“–”: No detectable activity/No related parameters were described. Data are the mean ± SD of at least three independent measurements. Coenzyme and substrate kinetic parameters of SmIDH, XsIDH, and other IDHs. “–”: No detectable activity/No related parameters were described. Data are the mean ± SD of at least three independent measurements. The optimal reaction pH (A,B), temperature (C,D), and thermal stability (E,F) of SmIDH and XsIDH. Data are the mean ± SD (n = 3 biological replicates). Given the fact that phosphorylation inactivates IDH primarily by affecting its binding to the substrate ICT, we measured the substrate kinetic parameters subsequently. Surprisingly, XsIDH displayed about 10-fold higher catalytic efficiency than most of the other reported enzymes. It can be attributed to the relatively lower Km values. As shown in Table 1, the Km values of XsIDH for ICT were similar to BsIDH25, but much lower than SmIDH as well as the other Type I IDHs like AbIDH148 and EcIDH46, indicating the extremely high affinity of XsIDH to ICT. Phosphorylation of the wild-type proteins The bioinformatics analyses above indicated that SmIDH and XsIDH may be phosphorylated due to the absence of the insertion sequence, thus, the phosphorylation experiments in vitro were performed subsequently. The residual activity after the phosphorylation reaction was measured to fit the curves (Figure 6). It is evident that SmIDH was phosphorylated rapidly by SmAceK, and that no more than 10% of the activity was left within 2 min, whereas the residual activity in the control groups (without phosphorylation) didn’t change apparently. In contrast, XsIDH-XsAceK showed relatively lower phosphorylation efficiency than SmIDH-SmAceK. After being phosphorylated for 2 min, XsIDH still retained 40% of the activity, and no more than 10% of the activity was left after 5 min. The phosphorylated enzymes were further detected by 8% Phos-tag™ SDS-PAGE. The band (phos-SmIDH), presented as the phosphorylated enzyme, could be seen easily between the bands of SmAceK and SmIDH (Figure 6A), as the phosphate groups51 enabled it to migrate more slowly than SmIDH. The phos-XsIDH band showed a similar result (Figure 6B). To test whether the phosphorylation events also occur in vivo, we measured the specific activities of XsIDH and XsICL in crude extracts of X. sacchari grown under glyoxylate-shunt-inducing (acetate as sole carbon source) and non-inducing (LB) conditions. After induction with acetate, the specific activity of XsIDH dropped sharply, whereas that of XsICL rose by ∼50% (Figure 7A,B). Consequently, the IDH/ICL activity ratio fell by ∼78%, from ∼52.6 to 12.5, consistent with phosphorylation-mediated regulation. Phosphorylation of SmIDH, XsIDH and XsIDHT in vitro. (A) Phosphorylation inactivation curves of SmIDH treated with SmAceK was obtained over 10 min using data at regular intervals. SmIDH untreated with SmAceK was used as a control. (B) Phosphorylation inactivation curves of XsIDH treated with XsAceK was obtained over 5 min using data at regular intervals. XsIDH untreated with XsAceK was used as a control. (C) Phosphorylation inactivation curves of XsIDHT treated with XsAceKT was obtained over 5 min using data at regular intervals. XsIDHT untreated with XsAceKT was used as a control. The bottom inset panel shows the separation of phos-IDH and IDH by 8% Phos-tag™ SDS-PAGE. Data are the mean ± SD (n = 3 biological replicates). Comparison of the specific enzyme activities of XsIDH (A) and XsICL (B) from the crude extract after culturing in LB and acetate. Data are mean ± SD (n = 3 biological replicates). Significance was assessed by t tests (**p < 0.01, ***p < 0.001). Identification and functional study of the phosphorylation site in SmIDH The band of phos-SmIDH in Phos-tag SDS-PAGE was collected for mass spectrometry to detect specific phosphorylation sites. A phosphorylated peptide was detected in phos-SmIDH. There was a phosphorylation modification of Ser at the 13th site in “SPLTTPVGGGFTSINVSLR” (Figure 8). It was determined by multiple sequence alignment that the phosphorylated site in phos-SmIDH was Ser 80 (Figure 2). Mass spectrum identification of phosphorylation site of SmIDH. As a classic approach, serine (Ser) is generally replaced by aspartate (Asp) or glutamate (Glu) to mimic the continuous phosphorylation state24,52,53. To explore the role the Ser80 site played in the function of SmIDH, it was substituted with Asp80 and Glu80, successively. And it was then replaced by Tyr80, Thr80, Ala80 or Gly80 to investigate the effects that the side chains of different amino acids have on the activity of SmIDH. The activities of those variants were then measured as a comparison with the wild-type SmIDH (Figure 9). There was a remarkable decrease in the enzyme activity of all mutants. The activities of S80D and S80E were completely lost. While the other four mutants of S80A, S80G, S80T, and S80Y only retained 1.8%, 3.7%, 9.3%, and 0.05% of their original activity, respectively. It is similar to the findings of a previous study6 where the corresponding mutants also showed a decrease in enzyme activity. Comparison of the enzyme activities of SmIDH and Ser80 mutants. Data are mean ± SD (n = 3 biological replicates). Significance was assessed by one-way ANOVA with Dunnett’s post-hoc test (****p < 0.0001). The kinetic parameters of the mutants were measured subsequently, compared with the wild-type SmIDH (Table 2). The Km value of SmIDH for ICT (25.7 ± 2.0 μM) was apparently lower than that of S80A (363 ± 19.3 μM), S80G (275 ± 13.5 μM), and S80T (126 ± 1.1 μM), indicating a significant decline in affinity to ICT. And even if the catalytic efficiency (kcat/Km) of S80T reached the highest 0.028 (μM−1s−1) among the three mutants, it was only 2% catalytic efficiency of the wild-type SmIDH. The Km values for NAD+ displayed a similar trend to that for ICT. Together, these results implied that the phosphorylation of Ser80 would greatly affect the catalytic activity of SmIDH. Table 2.Kinetic parameters of SmIDH and mutants of Ser80.EnzymeNAD+IsocitrateKm(μM)kcat(s−1)kcat/Km(μM−1 s−1)Km(μM)kcat(s−1)kcat/Km(μM−1 s−1)SmIDH141.2 ± 12.178.6 ± 8.30.5625.7 ± 2.035.2 ± 2.21.37S80A1513 ± 69.91.0 ± 0.090.00067363 ± 19.30.81 ± 0.030.0022S80G1268 ± 1052.2 ± 0.140.0017275 ± 13.51.9 ± 0.180.0068S80T1575 ± 1125.1 ± 0.130.0033126 ± 1.13.6 ± 0.030.028S80Y––––––S80D––––––S80E––––––“–”: No detectable activity/No related parameters were described. Data are the mean ± SD of at least three independent measurements. Kinetic parameters of SmIDH and mutants of Ser80. “–”: No detectable activity/No related parameters were described. Data are the mean ± SD of at least three independent measurements. Identification and functional study of the phosphorylation sites in XsIDH and XsAceK As the phosphorylation of EcIDH has been studied quite extensively, we modified the phosphorylation sites of XsIDH and XsAceK to the counterparts in E. coli and detected their phosphorylation in vitro afterwards. The Phos-XsIDHT bands migrated more slowly than the XsIDHT band (Figure 6C), suggesting that the mutants XsIDHT and XsAceKT still retained the ability for phosphorylation. The residual activity of XsIDHT after the in vitro phosphorylation reaction was measured (Figure 6C). It is shown that XsIDHT retained 100% activity steadily without the XsAceKT, while it decreased quickly in the presence of XsAceKT, and no more than 5% of the activity was left within 5 min, which was quite different from that of the wild type. Compared to XsIDH, XsIDHT showed slightly lower residual activity at each reaction time point (Figure 10), indicating a higher phosphorylation efficiency of XsIDHT -XsAceKT than that of the wild type. Comparison of phosphorylation efficiency of the wild types and mutants. Data are mean ± SD (n = 3 biological replicates). Significance was assessed by two-way ANOVA with Šídák’s post-hoc test (*p < 0.05, ***p < 0.001; ns, p > 0.05). Discussion AceK is one of the key enzymes in the glyoxylate shunt, by which the activity of IDH can be regulated through reversible phosphorylation and thus regulates the distribution of carbon flux in cells. Although the phosphorylation of IDH has been studied intensively in NADP-IDHs, there is no report about the real phosphorylation of NAD-IDH so far. Previous studies12,26 have shown that it is the α-helix formed by the extra insertion sequence in the small domain that hinders the recognition by AceK. EcIDH, together with SmIDH, XsIDH, and XccIDH1, didn’t have the extra insertion sequence (Figure 2), suggesting the possibility of being phosphorylated. To investigate the possible phosphorylation of NAD-IDH, we first selected the SmIDH and XsIDH for characterisation. It was shown that SmIDH and XsIDH could be phosphorylated by SmAceK and XsAceK in vitro (Figure 6) aligned with our previous speculation. We reasoned that if the phosphorylation also occurs in vivo under inducing conditions, it should lead to a corresponding and measurable change of enzyme activity in the crude extract. It is showed that the ratio of IDH/ICL activity decreased by approximately 78%. The specific enzyme activity of XsIDH declined significantly after culture in acetate, whereas that of XsICL increased by about 50%, consistent with phosphorylation-mediated regulation. The phosphorylation site was determined in SmIDH, which was later proved by mass spectrometry to be Ser80, the homologous site of EcIDH Ser113. As the mechanism of IDH phosphorylation has been extensively investigated in E. coli6,7, it is reasonable to hypothesise that the Ser80 in SmIDH is phosphorylated in the similar way. The site is involved in the binding of SmIDH to the substrate ICT, and its hydroxyl group interacts with the carboxyl group of ICT to catalyse the reaction. When being phosphorylated, the Ser80 site was recognised by SmAceK and bound to it, introducing a negatively charged phosphate group into the active site of SmIDH so that the binding of substrate ICT was impaired. As a result, it lost catalytic activity. To study the functional effects of the Ser80 site on SmIDH, six mutants were designed. The mutants S80D and S80E were constructed to mimic the permanent inactivation of Ser80 by phosphorylation. The acidic amino acids Asp and Glu introduced the carboxyl group into the substrate catalytic pocket of SmIDH, inactivating it consequently. The mutants S80Y, S80T, S80A, and S80G reduced the binding affinity to ICT by removing hydrogen bond donors or introducing significant steric hindrance and thus affected their catalytic activity. To be specific, compared with the wild type, the substitution of Ser with Thr retained the original β-hydroxyl group for hydrogen bonding with ICT. Therefore, even though the side chain of S80T introduced a small methyl group, causing some degree of steric hindrance, it still retained the highest residual activity among those mutants. In contrast, despite the absence of a hydroxyl group as the hydrogen bond donor, Gly and Ala still exhibited somewhat residual activity, less than S80T, though. This can be attributed to the relatively smaller side chains of Gly (a single hydrogen atom) and Ala (a methyl group). As for S80Y, the bulky side chains of Tyr may have impeded the binding to ICT, resulting in the complete loss of activity like S80D and S80E. Thus, we have first identified in vitro the possible phosphorylation of NAD-IDH that may exist in nature, and they are inactivated in a manner analogous to that of NADP-IDHs. Another open question is whether there is any difference between the phosphorylation of NAD-IDH and NADP-IDH? Given the phylogenetic analysis results and the low sequence identity that XsIDH and XsAceK shared with EcIDH and EcAceK, XsIDH and XsAceK appear to have evolved earlier than EcIDH and EcAceK, respectively. The phosphorylation efficiency of XsIDH-XsAceK could be successfully improved by modifying the phosphorylation sites to the counterparts of EcIDH and EcAceK, further corroborating the supposition that the phosphorylation of XsIDH-XsAceK may be a more ancient form of IDH phosphorylation. According to the previous study3, NAD-IDH is an ancient phenotype, while NADP-IDH is the result of adaptation to poor carbon sources. It is the NADPH that supplies the reducing power for living on the oxidised carbon sources. Whereas, IDH phosphorylation also enables some bacteria to exploit host carbon sources (like fatty acids and lipids) by activating the glyoxylate shunt to avoid the loss of two molecules of CO254. Despite the difference in regulation, NAD-IDH phosphorylation may be another strategy that bacteria utilise to adapt to poor carbon sources. Besides, recent studies have proven that M. tuberculosis can regulate its carbon flux by acetyltransferase, which can acetylate IDH55 or by ICL, which produces glyoxylate to metabolically cross-activate IDH13,56. Therefore, the other organisms harbouring NAD-IDH but lacking AceK may utilise the above strategies instead. However, the evolutionary sequence of these methods with IDH still needs further investigation. As is known to all, IDH will be inactivated after being phosphorylated, and the glyoxylate shunt will be activated at the same time, resulting in a carbon flux from the TCA cycle into the glyoxylate shunt. An earlier study57 has shown that the glyoxylate shunt plays a significant role in the pathogenicity of X. campestris. Two key enzymes in the glyoxylate shunt, ICL and MS, are potential drug targets when controlling the virulence of some animal and plant pathogenic bacteria or fungi58–60. However, the role of IDH-AceK at the glyoxylate shunt branch point during the process of pathogenic bacterial infection still remains unclear and is worth further exploration. X. campestris, as one of the significant pathogenic bacteria from Brassicaceae plants, has the condition to form the glyoxylate shunt like S. maltophilia, suggesting a similar mechanism of phosphorylation between them. This unique kind of phosphorylation may be closely related to the pathogenicity of pathogenic bacteria, though its specific mechanism still needs further exploration. The XsIDH-XsAceK phosphorylation efficiency could be easily changed by modifying the ARS and SRL of XsIDH and XsAceK, indicating the two regions played key roles in the phosphorylation of X. sacchari. They may be the potential drug targets to treat the black rot disease caused by X. sacchari. In conclusion, this study identifies the phosphorylation of NAD-IDH by AceK and its biochemical functions for the first time, shedding new light on the regulationof IDH. Further analyses for crystal structures of SmIDH and XsIDH are needed to uncover the structural basis of NAD-IDH phosphorylation in homodimers. And although preliminary in vivo studies have confirmed changes in enzyme activity, definitive proof of in vivo phosphorylation and the corresponding physiological roles still need to be verified. In the future, through gene editing technology, IDH and AceK deficient strains of S. maltophilia and X. sacchari will be constructed. This may facilitate the discovery of novel drug targets in these pathogenic bacteria. Materials and methods Bacterial strains, plasmids, reagents, and growth conditions The S. maltophilia strain was purchased from the BeNa Culture Collection (BNCC, Beijing, China). The X. sacchari strain (Accession Number: CGMCC 1.15149) was purchased from the China General Microbiological Culture Collection Centre (CGMCC, Beijing, China). The Escherichia coli strains, DH5α and Rosetta (DE3), and the pET-28b (+) expression vector were preserved at low temperature in our laboratory. The gene encoding SmAceK (Gene ID: WP_065181545.1), after codon optimisation, was synthesised by Generay Biotech Co., Ltd. (Shanghai, China). The genes encoding XsIDH (Gene ID: AJC45683.1) and XsAceK (Gene ID: AJC46993.1), after codon optimisation, were synthesised by GenScript Biotech Co., Ltd. (Nanjing, China). PrimeSTAR Max DNA Polymerase was purchased from TaKaRa (Dalian, China). Restriction enzymes (NdeI, XhoI, and DpnI), T4 DNA ligase and protein molecular weight standards were obtained from Thermo Fisher Scientific (Shanghai, China). The TALON Metal Affinity Resin used for protein purification was purchased from TaKaRa. The DL-isocitrate and NADP+ were purchased from Sigma–Aldrich (Shanghai, China). And the TIANamp Bacteria DNA Kit was purchased for genomic DNA extraction from TIANGEN (Beijing, China). The other reagents (e.g. kanamycin, chloramphenicol, NAD+ and so on) were obtained from BBI (Shanghai, China). Escherichia coli, S. maltophilia and X. sacchari strains were grown routinely for pre-cultures in LB medium at 37 °C, 30 °C, and 30 °C, respectively. For experiments to induce glyoxylate shunt in X. sacchari, a seed was first incubated in 5 ml Glucose-MOPS medium (20 mM glucose, 0.29 mM K2SO4, 1.32 mM K2HPO4, 1× MOPS buffer, adjusted to pH 7.5) at 30 °C and 225 rpm overnight. This culture was then inoculated at a 1:100 ratio into Succinate-MOPS medium (with glucose replaced by 20 mM sodium succinate) and subsequently into Acetate-MOPS medium (with glucose replaced by 20 mM sodium acetate) for acclimation. Next, 2 ml of the seed culture was washed twice with phosphate‐buffered saline (PBS, pH 7.5) and incubated in 200 ml of Acetate-MOPS medium in a 500 ml flask at 30 °C and 225 rpm for 3 days. Bioinformatics analyses The gene sequences used for analysis and gene synthesis were all downloaded from the NCBI database (National Centre for Biotechnology Information). The phylogenetic trees were constructed by the neighbor-joining (NJ) distance algorithm using the MEGA 7 software and then visualised using the Chiplot online tool (https://www.chiplot.online/)61. And the multiple sequence alignments were constructed using the ESPript 3.0 online biotool (https://espript.ibcp.fr/ESPript/cgi-bin/ESPript.cgi)62 based on the secondary structures downloaded from the PDB database (Protein Data Bank). Cloning and plasmid construction of SmIDH gene Briefly, the genomic DNA, which functioned as a template, was isolated from the S. maltophilia cultivated earlier using a DNA extraction reagent. And the SmIDH gene was amplified by PCR with the forward primer 5′- GGGAATTCCATATGACGCAGAAAATTACGGTCATCCGC-3′ (EcoRI site underlined) and the reverse primer 5′- CCGCTCGAGTTACAGGCGGCTGGCGATGGCCTGGGC-3′ (XhoI site underlined). Then, the polymerase chain reaction (PCR) was run at 95 °C for 3 min, and 35 cycles of 95 °C for 50 s, 60 °C for 50 s, and 72 °C for 15 s. Following this, the PCR products were purified, digested with NdeI and XhoI, and then ligated into the pET-28b(+) to obtain the recombinant plasmids (pET-SmIDH), which were subsequently verified by DNA sequencing. Recombinant protein overexpression and purification Firstly, the recombinant vectors were introduced into E. coli Rosetta (DE3) strain, respectively, which were then inoculated in 5 ml LB medium with 30 µg/ml kanamycin and chloramphenicol. After overnight culturing at 37 °C, the precultures were inoculated into 200 ml fresh LB medium containing the same antibiotics and incubated at 37 °C and 225 rpm. Once the optical density at 600 nm (OD600) reached 0.4–0.6, isopropyl-β-D-thiogalactopyranoside (IPTG) was added to the cultures to a final concentration of 0.1 mM, followed by an incubation at 20 °C and 180 rpm for 20 h. After that, the cells were harvested by centrifugation, resuspended in lysis equilibration washing buffer (LEW buffer: 50 mM NaH2PO4 and 300 mM NaCl, pH 7.5) and disrupted by sonication. The cell debris was then removed by centrifugation at 10 000 rpm and 4 °C for 20 min to obtain the supernatant, from which the 6 × His-tagged recombinant enzymes were purified using TALON Metal Affinity Resin according to the manufacturer’s instructions. SDS-PAGE and gel filtration chromatography The protein purity and the subunit(s) molecular mass were determined using the 12% SDS-PAGE, while the polymeric state was estimated by gel filtration chromatography, which was carried out by the ÄKTA purifier protein purification system equipped with a Superdex 200 (10/300) Increase column (GE Healthcare Life Sciences, Pittsburgh, PA, USA). The column was equilibrated with the LEW buffer previously mentioned. The protein standards were Thyroglobulin (669 kDa), Ferritin (440 kDa), Aldolase (158 kDa), Conalbumin (75 kDa), and Ovalbumin (44 kDa), according to which the standard curves were drawn to calculate the molecular mass of the recombinant enzymes. Enzyme assays A method63, after modification, has been introduced to measure the activity of SmIDH and XsIDH. A 1-ml standard reaction system was prepared, which contained 50 mM Tris-HCl (pH 7.5), 2 mM MnCl2/MgCl2, 1 mM trisodium DL-isocitrate, and 0.5 mM NAD+/NADP+. Using a Cary 300 UV-Vis spectrophotometer (Agilent, Santa Clara, CA, USA), the increase in NADPH or NADH was able to be monitored at 340 nm. One unit (U) of activity was defined as 1 μmol NADPH or NADH formed per minute at 25 °C (the standard assay condition). The specific enzyme activity (U/mg) was calculated as one unit (U) of activity per mg of protein. And the crude enzyme activity was measured from the supernatant of the cell lysate. All measurements were carried out in triplicate at least. And the concentrations of the protein were determined using the Bio-Rad protein assay kit (Bio-Rad, Hercules, CA).4.7 The in vitro phosphorylation assays The verified recombinant plasmids pET-SmAceK and pET-XsAceK were expressed in E. coli Rosetta (DE3), and the proteins were then purified in vitro. The phosphorylation was performed in a reaction mixture consisting of 25 mM Tris-HCl (pH 7.5, 100 mM NaCl), 50 mM MgCl2, 200 μM ATP, 6–10 μM AceK and IDH, using a method25 after modification. The mixture was then incubated at 37 °C for different durations (15 s, 30 s, 1 min, 2 min, 3 min, 5 min, 10 min, 30 min, 40 min, and 50 min). Their corresponding mutants were measured in the same way. And the residual enzyme activity was measured subsequently. To calculate the residual activity at different reaction time points, the value of the enzyme activity without phosphorylation was set as 100%. Measurements were performed in triplicate. Identification of phosphorylation site A Phos-tag SDS-PAGE was used to detect and separate the phosphorylated SmIDH for the subsequent phosphorylated protein analysis by using Nano LC-MS/MS. In the Phos-tag SDS-PAGE, the protein migrated faster than its phosphorylated form51,64. Therefore, the band of the phosphorylated protein could be excised from the SDS-PAGE, which was then sent to Shanghai Applied Protein Technology Co., Ltd. (APTBIO, Shanghai, China) to identify the phosphorylation site of SmIDH. Construction and analysis of mutants Site-directed mutagenesis was used to mimic the change in enzyme activity after phosphorylation so that the function of the phosphorylation site could be identified. Six forms of the mutants (S80D, S80E, S80Y, S80T, S80A, and S80G) were acquired according to the template pET-SmIDH. The primer sequences of target genes are shown in Table S2. By using Overlap Extension PCR (OE-PCR), the mutations were then introduced. The PCR was run at 95 °C for 3 min, and 35 cycles of 95 °C for 50 s, 60 °C for 50 s, and 72 °C for 2 min. Following this, a 20 μL DpnI digestion system was used to eliminate the template sequences, which contained 16 μL PCR products, 2 μL methylation-specific DpnI enzyme, and 2 μL Buffer. The digestion mixture was incubated at 37 °C for 90 min and 80 °C for 10 min. Finally, the confirmed mutants were introduced into competent bacterial strains for the subsequent heterologous expression, purification, and enzymatic analysis, similar to the wild-type one. According to the result of the sequence alignment aforementioned, the gene sequence of the phosphorylation sites (position 256–268 of the XsIDH) and the SRL (position 484–510 of the XsAceK) were modified to the corresponding sites of EcIDH and EcAceK. To substitute the gene fragment, the fusion PCR was applied. Due to the limited length of the sticky ends capable of being fused by the fusion PCR, each mutation was performed by two steps. The primer sequences of target genes are shown in Table S3. The PCR used to amplify the target DNA ran at 95 °C for 5 min, and 30 cycles of 94 °C for 45 s, 60 °C for 45 s, and 72 °C for 10 s. The target DNA was then purified using the TIANprep Midi Purification Kit. Then the purified products were used to run the PCR without adding primers at 95 °C for 5 min, 9 cycles of 94 °C for 30 s, 60 °C for 30 s, and 72 °C for 15 s, followed by another 30 cycles after adding primers (S-primer/AS-primer). Finally, the final confirmed mutated products were introduced into competent bacterial strains for the subsequent heterologous expression, purification, and enzymatic analysis, similar to the wild-type one. Supplementary Material Supplementary information_clean.docx Main Figure_clean.docx Main_Table__Clean.docx Funding Statement This work was supported by the National Natural Science Foundation of China [Grant no. 32071270], the Anhui Natural Science Foundation [Grant no. 2208085MC48], the Outstanding Innovative Research Team for Molecular Enzymology and Detection in Anhui Provincial Universities [Grant no. 2022AH010012], the Scientific Research Project of Anhui Provincial Department of Education [Grant no. 2022AH050183], and the Scientific Research Project of Anhui Provincial Department of Education [Grant no. 2023AH040023]. Disclosure statement No potential conflict of interest was reported by the author(s). Data availability statement The datasets presented in the current study are available from the corresponding author, P.W., upon reasonable request. References Associated Data Supplementary Materials Supplementary information_clean.docx Main Figure_clean.docx Main_Table__Clean.docx Data Availability Statement The datasets presented in the current study are available from the corresponding author, P.W., upon reasonable request.