Introduction
The compound 3’-phosphoadenosine-5’-phosphosulfate (PAPS) acts as a
sulfate group donor in the production of glucosinolate, heparin,
chondroitin sulfate, and oxamniquine (Ji et al., 2020; Zhang, Lin,
Huang, & Linhardt, 2020). At present, PAPS can be produced either via
metabolic or enzymatic synthesis. In the metabolic biosynthetic pathway
(Harjes, Bayer, & Scheidig, 2005; Sekulic, Dietrich, et al., 2007), ATP
is converted to PAPS by ATP
sulfurylase (ATPS) and
adenosine-5′-phosphosulfate
(APS) kinase (APSK). However, because ATP is an energy-rich compound
with limited capacity to accumulate in cells, only 0.8–1.2 μmol PAPS is
obtained per gram of cells (Badri et al., 2021; Badri, Williams, Xia,
Linhardt, & Koffas, 2019). Enzymatic synthesis (Burkart, Izumi,
Chapman, Lin, & Wong, 2000), which includes single-enzyme catalysis and
dual-enzyme cascade catalysis, has been developed to further enhance
PAPS output.
In single-enzyme catalysis, aryl sulfotransferase (ASST) converts
3′-phosphoadenosine-5′-phosphate (PAP) and p -nitrophenyl sulfate
(PNPS) to PAPS (Jin et al., 2020; Z. Zhou et al., 2019). Over the past
decade, various ASSTs from mammalian species and bacteria, such asStreptomyces sp. (Kaysser et al., 2010) and Escherichia
coli (Malojcic, Owen, & Glockshuber, 2014; Malojcic et al., 2008),
have been identified and used to synthesize PAPS. However, ASST
expression and affinity for PAP need to be improved (Berger, Guttman,
Amar, Zarivach, & Aharoni, 2011). To increase yields, fusion with the
signal peptides Cex, YebF, and PelB has been applied, allowing for ASST
secretion and fourfold higher expression (89.67
U·mL-1) (Z. Zhou et al., 2019). To improve affinity
for PAP, random mutagenesis and molecular evolution of the PAP-binding
pocket gate loop were applied based on the enzyme’s crystal structure.
PAP affinity and ASST catalytic efficiency were thus increased by 2.48
times and 12.50 times, respectively (Z. Zhou et al., 2019). Still, some
hurdles remain, as the regeneration of PAPS requires coupling to
sulfotransferases and prevents PAPS accumulation (An, Zhao, Wei, &
Zhou, 2017; Bao et al., 2015), while the elevated cost of PAP (38
$·mg-1, Sigma) lowers its commercial appeal.
A relatively cheap dual-enzyme cascade catalysis using ATP as the
substrate was developed. Given that the price of ATP is only 1/25 that
of PAP, this strategy offers an appealing alternative (Datta et al.,
2020). The catalysis process includes two steps: first, sulfate and ATP
are converted by ATPS to adenosine-5′-phosphosulfate (APS) and the
by-product pyrophosphate (PPi) (Kang et al., 2018; Schmidt, 1977);
second, APS kinase catalyzes the conversion of APS and ATP to PAPS and
the by-product ADP (Badri et al., 2019). At present, this method has
been used to synthesize PAPS at gram-level in the preparation of
bioengineered heparin and chondroitin sulfate (Jian, Liu, Robert, &
reports, 2014; Jin et al., 2020). However, by-product inhibition and low
enzymatic activity compel a substrate conversion rate of only 47% (An
et al., 2017). To alleviate inhibition, by-product degradation and
recycling have been attempted (Michael D. Burkart, 2000; X. Zhou,
Chandarajoti, Pham, Liu, & Liu,
2011). In one such example, the
conversion rate was increased by 50%, and 5.0 g·L-1PAPS was generated when ADP was transformed back to ATP, using
phosphoenolpyruvate (PEP) as a phosphate donor (An et al., 2017).
APSK is a key factor affecting catalytic efficiency of the dual-enzyme
cascade and is characterized by a typical “lid structure” (Gay, Segel,
& Fisher, 2009). It catalyzes a sequential reaction, whereby ATP binds
ahead of APS, and then PAPS leaves before ADP is released (Lansdon,
Segel, & Fisher, 2002). If ADP is not released in time, an
APS–enzyme–ADP termination complex forms, resulting in a decrease in
enzymatic activity (Lansdon et al., 2002; Mueller & Shafqat, 2013).
Protein crystallization (Ian J. MacRae, 2000; Poyraz et al., 2015),
site-directed mutagenesis (Wang D. Z et al., 2016), and truncated
mutagenesis (Ravilious, Westfall, & Jez, 2013; Sekulic, Konrad, &
Lavie, 2007) have been used to study APSK. Secondary structure analysis
of APSK from Arabidopsis thaliana showed that
Arg93 was necessary for substrate recognition, and
affinity for ADP was 217-times lower in the APSKR93Amutant (Ravilious et al., 2013). Site-directed mutagenesis produced anOryza sativa Os APSKC36A/C69A mutant,
whose kcat was 43% lower than that in the wild-type, while theKm for APS was 1.6-fold higher (Wang D. Z et al., 2016).
These studies focused mainly on
APSK structure and its catalytic mechanism; however, attempts to improve
catalytic performance of the enzyme have yielded only limited success.
In this study, a PAPS-producing catalytic system composed of a main
module and an auxiliary module was designed and assessed in vivo .
The main module converted ATP to PAPS, and the auxiliary module
effectively eliminated by-product inhibition by hydrolyzing PPi to
regenerate ATP from ADP. APSK was identified as the limiting step in
this catalytic system due to delayed release of ADP. To further increase
conversion efficiency, a
mechanism-guided “ADP expulsion” strategy was developed to weaken the
binding affinity of APSK for ADP and expand the bottleneck caused by its
U-shaped release channel. Finally, by integrating the best variant in
the cascade pathway, 73.6 mM (37.3 g·L-1) PAPS was
synthesized, using a 3-L fermenter, with 98.1% conversion.