**4. The Intricate Process of Urease Activation – Much More Than Structural Genes**

The biosynthesis of metalloenzymes usually depends on the participation of several dedicat‐ ed proteins that are essential for the correct assembly of their active sites, and ureases are no exception. The role of these accessory proteins consist on the stabilization of the apoenzyme in a certain conformation that allows the correct insertion of the metal ion in the active site, dissociating afterwards and releasing the mature enzyme [42]. This process has been fairly studied for bacterial ureases, but the activation of plant ureases still demands more atten‐ tion. Thus it is described here based mostly on what is known for bacterial ureases.

The activation of ureases require two essential steps: the carbamylation of a lysine residue, that will be responsible for bridging and, consequently, holding the two Ni2+ ions into place within the active site; and the actual incorporation of the two Ni2+ ions in the active site. The best char‐ acterized system so far is that of *Klebsiella aerogenes* urease, for which four accessory proteins, namely UreD, UreE, UreF and UreG, are required for complete activation [2]. Although many efforts have been put on characterizing these proteins, the precise role of each one in the com‐ plex process of urease activation is still not clearly understood. These proteins appear to be well conserved among urease producing bacterial organisms and even for higher organisms as plants. In *K. aerogenes*, the accessory proteins bind consecutively to the apourease (apU) form‐ ing the following complexes: apU-UreD, apU-UreDF, apU-UreDFG [2, 43]. UreE, a Ni2+ bind‐ ing protein, is the last accessory protein proposed to bind to the complex apU-UreDFG and deliver the metal to the enzyme active site [44]. The process of urease activation requires the hydrolysis of GTP by UreG, after which the accessory proteins dissociate from the active en‐ zyme. A preformed UreDFG complex could be isolated from bacteria lacking UreE [45]. If the accessory proteins act on apU as a preformed complex or in a sequential binding fashion, is a matter that demands further investigation. However, the sequential binding model is to date the most accepted and the one supported by most evidence.

Although the exact role of each accessory protein has not been clearly assigned, some gener‐ al lines can be traced for their individual actions. The current sequential model assumes that UreD is the first accessory protein to interact with urease. UreD is yet the least characterized protein and it seems to serve as an adapter for the other accessory proteins since neither UreF or UreG are able to bind urease directly in the absence of UreD [2]. UreF interacts di‐ rectly with UreD and it has been proposed that UreF would be responsible for promoting a conformational change in apU, providing better access to the active site of the protein and allowing the next steps of the process to take place [46]. Recently, a structural model of UreF has indicated that this protein shares structural similarities to some GTP activating proteins (GAP) [47]. UreG is an intrinsically disordered GTPase, as reported for organisms such as *Helicobacter pylori*, *Bacillus pasteurii*, *Micobacterium tuberculosis* and also *Glycine max* [48-53]. UreG binds to UreF and, through the cleavage of GTP, provides the necessary energy for the activation process. It is also postulated that GTP cleavage in the presence of CO2 could form carboxyphosphate, an excellent carbamylation agent [46, 54]. UreG GTPase activity, when detectable, seems to be very low in comparison to other GTPases. It has been proposed [46] that the intrinsically disordered structure of this enzyme serves as a regulatory mechanism of its activity, which would be only maximized when, inserted in the activating complex, it acquires its fully folded state. It is also hypothesized that UreF would act on UreG as a GAP, enhancing its enzymatic activity, although, this hypothesis still waits to be tested.

Available sequences of bacterial urease accessory proteins led to the search of potential or‐ thologs in plants, and the identification of UreG (Eu3) in soybean. This was the first evi‐ dence of accessory proteins in plants [41]. Soybean cDNAs for UreD and UreF proteins were also identified later [39], but none of them were assigned as the *Eu2* gene. To date, no UreE equivalent in plants has been identified and sequence analyses supports the hypothesis that its functionality is incorporated into UreG which, in plants but not in bacteria, presents an extended N-terminal rich in aspartic acid and histidine residues. The property of soybean UreG to bind metal ions has been demonstrated [41, 53]. A study in *A. thaliana* [55] has shown that UreD, UreF and UreG are necessary and sufficient to activate urease, since knockout plants for any of the accessory proteins genes lack urease activity. They also showed that simultaneous co-expression of all accessory proteins together with *A. thaliana* urease structural gene was able to generate urease activity on *Escherichia coli* cells. Also the *Oriza sativa* urease could only be activated when all of its three accessory proteins were cotransformed in tobacco [56].

With the exception of dimeric ureases described for some plants, such as canatoxin from *C. ensiformis* [5], and the ureases from *Morus alba* [57] and *Momordica charantia* [58], all reported ureases share a basic trimeric state that may aggregate to form larger oligomers, as hexam‐ ers in the case of most plant ureases, in which each monomer carries one active site. For bac‐ teria, each unit of this trimer is usually itself a heterotrimer composed by the subunits UreA, UreB and UreC which co-align with plant ureases with over 50% identity. Thus the single type polypeptide chain of plant ureases corresponds to the collinear fusion of the bacterial subunits (UreA-UreB-UreC). It is curious to notice that despite the high similarity between plant and bacterial ureases, some of the accessory proteins do not share the same degree of sequence identity. Plant and bacterial UreG are very similar, with about 40% of sequence identity, but plant UreD and UreF share only about 20% of sequence identity with their bac‐ terial counterparts [56]. Considering this discrepancy, a recent work using *in silico* structure prediction tools, has shown that despite the sequence disparities, UreF proteins are very conserved at the structural level [56]. Plant UreD, although sharing some structural similari‐ ties, possess some marked differences when compared to their bacterial counterparts. This different overall structure of plant and bacterial UreD is hypothesized to reflect distinct re‐ quirements for interaction with the bacterial heterotrimeric or the plant single polypeptide urease unit [56].

hydrolysis of GTP by UreG, after which the accessory proteins dissociate from the active en‐ zyme. A preformed UreDFG complex could be isolated from bacteria lacking UreE [45]. If the accessory proteins act on apU as a preformed complex or in a sequential binding fashion, is a matter that demands further investigation. However, the sequential binding model is to date

A Comprehensive Survey of International Soybean Research - Genetics, Physiology, Agronomy and Nitrogen

Although the exact role of each accessory protein has not been clearly assigned, some gener‐ al lines can be traced for their individual actions. The current sequential model assumes that UreD is the first accessory protein to interact with urease. UreD is yet the least characterized protein and it seems to serve as an adapter for the other accessory proteins since neither UreF or UreG are able to bind urease directly in the absence of UreD [2]. UreF interacts di‐ rectly with UreD and it has been proposed that UreF would be responsible for promoting a conformational change in apU, providing better access to the active site of the protein and allowing the next steps of the process to take place [46]. Recently, a structural model of UreF has indicated that this protein shares structural similarities to some GTP activating proteins (GAP) [47]. UreG is an intrinsically disordered GTPase, as reported for organisms such as *Helicobacter pylori*, *Bacillus pasteurii*, *Micobacterium tuberculosis* and also *Glycine max* [48-53]. UreG binds to UreF and, through the cleavage of GTP, provides the necessary energy for the activation process. It is also postulated that GTP cleavage in the presence of CO2 could form carboxyphosphate, an excellent carbamylation agent [46, 54]. UreG GTPase activity, when detectable, seems to be very low in comparison to other GTPases. It has been proposed [46] that the intrinsically disordered structure of this enzyme serves as a regulatory mechanism of its activity, which would be only maximized when, inserted in the activating complex, it acquires its fully folded state. It is also hypothesized that UreF would act on UreG as a GAP,

enhancing its enzymatic activity, although, this hypothesis still waits to be tested.

transformed in tobacco [56].

Available sequences of bacterial urease accessory proteins led to the search of potential or‐ thologs in plants, and the identification of UreG (Eu3) in soybean. This was the first evi‐ dence of accessory proteins in plants [41]. Soybean cDNAs for UreD and UreF proteins were also identified later [39], but none of them were assigned as the *Eu2* gene. To date, no UreE equivalent in plants has been identified and sequence analyses supports the hypothesis that its functionality is incorporated into UreG which, in plants but not in bacteria, presents an extended N-terminal rich in aspartic acid and histidine residues. The property of soybean UreG to bind metal ions has been demonstrated [41, 53]. A study in *A. thaliana* [55] has shown that UreD, UreF and UreG are necessary and sufficient to activate urease, since knockout plants for any of the accessory proteins genes lack urease activity. They also showed that simultaneous co-expression of all accessory proteins together with *A. thaliana* urease structural gene was able to generate urease activity on *Escherichia coli* cells. Also the *Oriza sativa* urease could only be activated when all of its three accessory proteins were co-

With the exception of dimeric ureases described for some plants, such as canatoxin from *C. ensiformis* [5], and the ureases from *Morus alba* [57] and *Momordica charantia* [58], all reported ureases share a basic trimeric state that may aggregate to form larger oligomers, as hexam‐ ers in the case of most plant ureases, in which each monomer carries one active site. For bac‐

the most accepted and the one supported by most evidence.

Relationships

322

Despite the fact that the presence of the set of accessory proteins is enough to get ureases activated, there seems to be more to it concerning regulation. In bacteria, UreF and UreD are expressed in very low levels, and it has been shown that over expression of these proteins can hamper urease activation [59, 60]. It has been proposed that differential splicing generat‐ ing aberrant mRNA could reduce UreD production in plants [55]. Cao and co-workers [56] reported for *ureF* an intron in the 5'leader conserved among 16 plant genomes and they no‐ ticed that in almost every case the spliced transcript would be free of AUG codons upstream of the start codon. Special attention was given to the two *ureF* genes from soybean which, although both are spliced at the 5' leader, only the paralog in chromosome 2 (Gly‐ ma02g44440) has all the nonstart AUG codons removed. The transcript of the paralog from chromosome 14 (Glyma14g04380) has an out of frame AUG codon upstream of the start co‐ don and therefore was postulated to be ineffectively translated and consequently non func‐ tional. In the same work, a low splicing efficiency of *AtureF* was observed, which led to the conclusion that for some plants the 5' leader sequence may have a regulatory role in reduc‐ ing the amount of *ureF* mRNA either by differential splicing (*Arabidopsis*) or translational in‐ hibition (soybean). Limited expression may be required to ensure that UreF and UreD dissociate from urease after activation to release the active enzyme and that the putative GTPase activating protein UreF does not trigger UreG activity in the absence of urease [56]. Both UreD and UreF have been observed to be very unstable proteins either in plants [56] or bacteria [59], and this intrinsic instability possibly contributes to the regulation of their activ‐ ities *in vivo* [56].

As mentioned above, soybean genome contains two UreF genes. The one in chromosome 14 UreF (Ch14UreF) has previously been characterized and demonstrated to activate the urease from *S. pombe* [39], but the UreF encoded by chromosome 2 (Ch02UreF) remains to be char‐ acterized. Until recently, the product of *Eu2* gene was not identified, and it had been pro‐ posed that *Eu2* could represent the functional paralog of UreD or UreF. Analysis of soybean *Eu2* mutants revealed that missense mutations resulted in the expression of an altered form of Ch02UreF protein [40]. Interestingly, these mutants presented no urease activity, even

though the Ch14UreF is present and is supposed to activate urease. Another mutant, in which the expression of Ch02UreF was impaired, presented 5-10 % of the wild type urease activity. The authors presented two possible explanations for the results. The first one con‐ sidered that Ch02UreF could spoil activation by Ch14UreF because of a higher affinity for the activation complex. The second explanation, favoured in their work, proposed that Ch02UreF is more abundant than Ch14UreF, which would be less efficiently translated. Therefore, in *Eu2* mutants, the missense mutants of Ch02UreF block the access of Ch14UreF to the urease activation complex, preventing activation [40].

As pointed out here, although accessory proteins differ widely according to their source, the process of urease activation seems to be very well conserved. Among plants, the urease acti‐ vation complex seems to be structurally very similar, since accessory proteins from different plants are able to functionally complement each other. Rice urease, for instance, can be acti‐ vated by *Arabidopsis* bulk of accessory proteins, and UreD and UreG from rice can replace the native accessory protein in mutants of Arabidopsis [56]. The similarities of the activation process seem also to break the kingdom barrier. Soybean UreF has been shown to comple‐ ment the *Saccharomyces pombe* accessory set [39], and UreG from potato (*Solanum tuberosum*) complements the *K. aerogenes* operon [61].
