# Source paper archive - Source title: Structural determinants underlie the essentiality of the m-type thioredoxin TrxA in Synechocystis - Author(s): Manuel J Mallen-Ponce, Jesus Moron-Lopez, Francisco J. Florencio, Maria-Jose Huertas - Source landing URL: https://www.biorxiv.org/content/10.64898/2026.11.05.764636v1 - PDF URL: https://www.biorxiv.org/content/10.64898/2026.11.05.764636v1.full.pdf - PDF SHA-256: 0448b9adbd86a1b9c8b34e17c1b803b11863eb4e8feef172d322374444b22de8 - Converter: pymupdf4llm 0.0.27 - Limits: Automated PDF text extraction; reading order, equations, tables and formatting may be incomplete or incorrect. Images are omitted, and no OCR was performed. Markup and links are neutralized for safe downloading. Consult the original PDF for the authoritative version. --- [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 1 **Structural determinants underlie the essentiality of the** _**m**_ **-type thioredoxin TrxA in** 2 _**Synechocystis**_ 3 4 **Manuel J Mallén-Ponce** **[1,2]** ***; Jesús Moron-López** **[3]** **; Francisco J Florencio** **[1,2]** **; María José** 5 **Huertas** **[1,2]** ***** 6 7 1 Departamento de Bioquímica Vegetal y Biología Molecular, Facultad de Biología, Universidad 8 de Sevilla, Profesor García González, 1, 41012, Sevilla, Spain. 9 2 Instituto de Bioquímica Vegetal y Fotosíntesis (Universidad de Sevilla, Consejo Superior de 10 Investigaciones Científicas), Américo Vespucio 49, 41092, Sevilla, Spain. 11 3 School of Sustainable Engineering and the Built Environment, Arizona State University, 660 S 12 College Ave, Tempe, AZ 85287, USA. 13 14 *Corresponding authors: Manuel J. Mallén-Ponce. Email: m mallen@us.es ; María José Huertas 15 mjhuertas@us.es. 16 17 The author responsible for distribution of materials integral to the findings presented in this article 18 in accordance with the policy described in the Author guidelines 19 (https://academic.oup.com/plphys/pages/author-guidelines) is Manuel Jesús Mallén Ponce 20 (mmallen@us.es) and María José Huertas (mjhuertas@us.es). 21 22 **ABSTRACT** 23 Thioredoxins are central regulators of thiol-based redox signaling in photosynthetic organisms, 24 where multiple isoforms with partially overlapping functions coexist. Among them, TrxA is the 25 cyanobacterial homolog of chloroplast _m_ -type thioredoxins and the only thioredoxin known to be 26 essential in cyanobacteria. However, it is unclear whether this essentiality is due to its high 27 intracellular abundance or to its specific molecular properties. Here, we evaluated the ability of 28 different thioredoxins to functionally replace TrxA and used TrxA/TrxB chimeras to identify the 29 structural determinants underlying its functional specificity. Overexpression of TrxB, TrxQ or 30 TrxC did not support complete segregation of a disrupted _trxA_ locus, although the proteins 31 accumulated to levels comparable to those of endogenous TrxA. Similarly, heterologous 32 expression of NTRC did not rescue the phenotype associated with reduced TrxA levels. 33 Comparative structural analysis revealed differences in surface electrostatics and structural 34 features surrounding the catalytic site of TrxA compared with the other thioredoxins. Analysis of 35 TrxB/TrxA chimeras showed that determinants required for functional complementation are 36 primarily located within the C-terminal region of TrxA. Chimeric variants also differed in their 37 ability to activate cyanobacterial FBP/SBPase and pea chloroplast FBPase _in vitro_, supporting a [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 38 role for this region in target-specific interactions. Together, these results demonstrate that the 39 essential function of TrxA cannot be explained by protein abundance alone and requires molecular 40 determinants located primarily within its C-terminal region. 41 42 **INTRODUCTION** 43 Thioredoxins (Trxs) are small soluble oxidoreductases widely distributed across all 44 domains of life (Schürmann and Buchanan, 2008). Characterized by a conserved Trx fold and a 45 catalytic active site containing the canonical WC(G/P)PC motif, Trxs play a central role in cellular 46 redox homeostasis by catalyzing dithiol-disulfide exchange reactions that regulate numerous 47 cellular processes (Holmgren et al., 2005). In oxygenic photosynthetic organisms, Trxs have 48 diversified into several phylogenetic and functional types. Three major lineages are conserved 49 between cyanobacteria and photosynthetic eukaryotes, including the cyanobacterial TrxA, TrxB 50 and TrxQ, which are related to chloroplast _m_ -, _x_ - and _y_ -type Trxs, respectively (Mallén-Ponce et 51 al., 2022). TrxA is the only Trx present in all cyanobacteria and photosynthetic eukaryotes, which 52 also contain multiple TrxA or _m_ -type Trx isoforms. In cyanobacteria, the duplication of TrxA in 53 many lineages and the high similarity of these TrxA paralogues suggest that these functions were 54 originally fulfilled by a single copy. Its duplication may subsequently serve as a backup of 55 essential functions. 56 In oxygenic photosynthetic organisms, Trxs are mainly reduced by the ferredoxin 57 thioredoxin reductase (FTR) system, which links photosynthetic electron transport to thiol 58 dependent redox regulation. FTR is a heterodimeric enzyme composed of a catalytic subunit 59 containing a [4Fe-4S] cluster (FTRc) and a variable subunit (FTRv), with ferredoxin serving as 60 the electron donor (Balsera et al., 2013). FTR is widely distributed among cyanobacterial lineages 61 and is also present in chloroplasts, although it is absent in some cyanobacteria, such as 62 _Prochlorococcus_ and _Gloeobacter_, where alternative ferredoxin-dependent reductases have been 63 identified (Buey et al., 2017b; Mallén-Ponce et al., 2022). 64 Flavin-thioredoxin reductases constitute another group of Trx reductases and include 65 NAD(P)H-thioredoxin reductases (NTRs), which use NADH or NADPH as electron donors 66 (Balsera and Buchanan, 2019). Cyanobacteria also contains ferredoxin flavin-thioredoxin 67 reductase (FFTR), an NTR-related enzyme that lacks the NADPH-binding site and is reduced by 68 ferredoxin. FFTR is the only identified Trx reductase in _Prochlorococcus_ and _Gloeobacter_ (Buey 69 et al., 2017a). Many cyanobacteria and chloroplasts also contain NADPH thioredoxin reductase 70 C (NTRC), a bifunctional enzyme comprising an NTR module fused to a Trx domain (Serrato et 71 al., 2004). NTRC does not appear to interact with Trxs, but it efficiently reduces 2-Cys 72 peroxiredoxin (2-Cys Prx) in both cyanobacteria and chloroplasts (Pérez-Ruiz et al., 2006; 73 Pascual et al., 2011). [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 74 Trx-based redox regulation controls key cellular processes including carbon assimilation, 75 antioxidant response and protein synthesis. Early biochemical studies in photosynthetic 76 organisms identified several Calvin-Benson (CB) cycle enzymes as Trx targets, including 77 phosphoribulokinase (PRK), glyceraldehyde 3-phosphate dehydrogenase (GAPDH) or fructose 78 1,6-bisphosphatase (FBPase) (Michelet et al., 2013; Mallén-Ponce et al., 2022). In cyanobacteria, 79 redox regulation of carbon metabolism involves both conserved and lineage-specific targets. PRK 80 retains redox regulation across the green lineage, while cyanobacteria also possess a bifunctional 81 fructose-1,6-bisphosphatase/sedoheptulose-1,7-bisphosphatase (FBP/SBPase) enzyme that is 82 regulated by TrxA (Gerbling et al., 1986; Mallén-Ponce et al., 2021). CP12, a conserved small 83 intrinsically disordered protein, is also subject to Trx-dependent redox regulation and forms an 84 inhibitory complex with PRK and GAPDH in its oxidized state (Wedel et al., 1997; Gurrieri et 85 al., 2021). Proteome-wide measurements in _Synechocystis_ further indicate that CB enzymes 86 operate at distinct redox states _in vivo_ ; consistent with a hierarchical redox network shaped by 87 cellular context and and Trx interactions (Tanaka et al., 2026). 88 In addition to CB cycle regulation, TrxA also links photosynthetic redox control to the 89 oxidative pentose phosphate pathway by indirectly inactivating glucose-6-phosphate 90 dehydrogenase (G6PDH) through reduction of its allosteric activator OpcA. Structural and redox 91 proteomic studies have shown that OpcA regulates G6PDH through allosteric conformational 92 changes and redox-sensitive cysteine modifications near the protein-protein interface (Doello et 93 al., 2024). Together, these findings illustrate the broad role of TrxA in coordinating redox control 94 of central metabolic pathways in cyanobacteria, extending beyond the classical regulation of CB 95 cycle enzymes. 96 Despite the high conservation of the Trx fold and catalytic motif throughout evolution, 97 several studies have provided evidence for functional specificity among Trx types. In 98 chloroplasts, redox analyses of some CB cycle enzymes revealed a higher affinity for _f_ -type Trxs 99 than for _m_ -type Trxs (Wolosiuk et al., 1979; Michelet et al., 2013; Yoshida et al., 2015). Other 100 Trxs, such as _x_ - and _y_ -type Trxs, appear to regulate the oxidative stress responses (Collin et al., 101 2003; Collin et al., 2004; Vieira Dos Santos et al., 2007). Recent protein-engineering work in 102 _Chlamydomonas_ further supports the idea that Trx specificity can be altered by modifying 103 selectivity determinants outside the catalytic cysteines, highlighting the importance of surface 104 features and target-recognition regions in defining Trx function (Lemaire et al., 2025). In 105 cyanobacteria, previous studies have also suggested functional differentiation among Trx 106 isoforms, although this aspect remains largely unexplored (Pérez-Pérez et al., 2009). However, 107 most reports have been primarily based on _in vitro_ biochemical studies, where even different Trxs 108 can reduce the same target protein. Subsequent _in vivo_ studies, including the generation of 109 knockout or knockdown mutant strains for one or two Trxs, suggested a high degree of overlap 110 between the target proteins of the different Trxs. Indeed, most Trx mutants in plants and [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 111 cyanobacteria display phenotypes only under specific conditions (Geigenberger et al., 2017; 112 Mallén-Ponce et al., 2022). In contrast, TrxA ( _m_ -type Trx) mutants exhibit phenotypes even under 113 standard growth conditions (Wang et al., 2013; Okegawa and Motohashi, 2015; Mallén-Ponce et 114 al., 2021). 115 Many studies have not considered that TrxA ( _m_ -type Trx) is the most abundant Trx in 116 photosynthetic organisms. Analyses of Trx levels in _Synechocystis_ and _Anabaena_ revealed that 117 TrxA represents about 80% of the total Trx pool (Florencio et al., 2006; Mihara et al., 2020), 118 whereas in _Arabidopsis_ it constitutes approximately 70% of the total Trx pool (Okegawa and 119 Motohashi, 2015). These observations, together with the inability to obtain mutant strains 120 completely lacking this Trx (Muller and Buchanan, 1989; Navarro and Florencio, 1996), 121 suggested that TrxA essentiality could derive mainly from its high intracellular abundance. 122 Recent work with a conditional TrxA mutant has shown that reduced TrxA levels impair protein 123 synthesis and alter carbon/nitrogen partitioning, reinforcing the idea that TrxA controls central 124 cellular processes beyond individual metabolic enzymes (Mallén-Ponce et al., 2024). 125 Together, these observations raise the question of whether TrxA essentiality depends 126 primarily on its abundance or on intrinsic molecular properties that determine target recognition. 127 Here, we show that TrxA essentiality cannot be explained by intracellular abundance alone. 128 Instead, genetic, structural and biochemical analyses of TrxA/TrxB chimeras reveal that 129 conserved surface regions downstream of the active site are required for functional interaction 130 with redox regulated targets. 131 132 **MATERIALS AND METHODS** 133 **Strains and culture conditions** 134 _Synechocystis_ sp. PCC 6803 was grown at 30 °C in BG11 supplemented with 12 mM 135 NaHCO₃ (BG11C) (Rippka R, 1988), bubbled with 1% (v/v) CO₂ in air under continuous 136 illumination (50 μmol photons m⁻² s⁻¹). Sodium arsenite (NaAsO₂; 1 mM) was added when 137 required to induce the _arsB_ promoter. 138 For inducer removal experiments, exponentially growing cells cultured with 1 mM arsenite 139 were harvested (4,200 × g, 5 min), washed twice with BG11C and resuspended at a similar cell 140 concentration in fresh medium with or without inducer. Solid BG11C contains 1% (w/v) agar. 141 Antibiotics were used at the following concentrations: nourseothricin and kanamycin, 50 μg mL⁻¹; 142 spectinomycin and streptomycin, 2.5 μg mL⁻¹ each. 143 **Generation** **of mutant** _**Synechocystis**_ **strains** 144 The _trxA_ ( _slr0623_ ), _trxB_ ( _slr1139_ ), _trxQ_ ( _slr0233_ ), and _trxC_ ( _sll1057_ ) genes were PCR 145 amplified, and chimeric TrxA/TrxB constructs were generated by overlap-extension PCR using 146 the oligonucleotides listed in Table S3. After sequence verification, chimeric constructs were 147 cloned into pglnN::ParsB via _EcoR_ I _/Not_ I restriction sites; native _trx_ genes were cloned directly [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 148 into the same vector. These constructs target the _glnN_ ( _slr0288_ ) locus and confer nourseothricin 149 resistance (Tables S2 and S5). 150 The constructs were introduced into WT _Synechocystis_ and, where indicated, the SR 151 background. To assess TrxA essentiality, strains were subsequently transformed with pGQ5.4+ 152 (Mallén-Ponce et al., 2021), which disrupts the endogenous _trxA_ gene. Fully segregated strains 153 were obtained for STXA3, STXBA and STXB48A. 154 For NTRC expression, the _ntrc_ gene ( _all0737_ ) from _Anabaena_ sp. PCC 7120 was excised 155 from p28AnNTR2 using _Nde_ I _/Not_ I and cloned into pnrsD::PcpcB, generating pBS-NTRC. This 156 construct targets the _nrsD_ ( _slr0796_ ) locus and confers kanamycin resistance. Transformation of 157 WT and STXA2 generated strains SNTRC and STXA2C, respectively. 158 **PCR verification and chromosome segregation analysis** 159 Genomic DNA from _Synechocystis_ transformants was used to verify construct integration 160 and chromosome segregation by PCR. Integration of the _arsB_ -driven constructs at _glnN_ and the 161 _PcpcB-ntrC_ construct at _nrsD_ was assessed using primers flanking the respective insertion sites. 162 Segregation of the endogenous _trxA_ locus was assessed using primers distinguishing the WT and 163 disrupted alleles. PCR products were separated by agarose gel electrophoresis and visualized by 164 DNA staining. All primers are listed in Table S3. 165 **Protein extraction and immunoblot analysis** 166 10 mL cultures were harvested and frozen at -20ºC. Cell pellets were resuspended in 50 mM 167 Tris–HCl, pH 8, 25 mM NaCl and 1 mM Phenylmethylsulfonyl fluoride. Subsequently, cells were 168 lysed using glass beads and total proteins were extracted as previously described (Reyes et al., 169 1995). Total protein concentration was determined using the Lowry assay (Markwell et al., 1978). 170 15 µg of proteins were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis 171 (SDS-PAGE), transferred to nitrocellulose membranes (Bio-Rad) and immunoblotted with the 172 indicated antibodies in each case (Table S4). Finally, blots were analyzed using the ECL Prime 173 Western Blot Detection Reagent (Bio-Rad) with secondary anti-rabbit antibody (1:20.000) 174 following the instructions of the manufacturer. 175 **Cloning, expression and purification of recombinant chimeric Trxs** 176 Recombinant TrxA and TrxB were purified as previously described (Navarro and 177 Florencio, 1996; Pérez-Pérez et al., 2009). Chimeric Trxs and the _Synechocystis_ FBP/SBPase 178 (García-Cañas et al., 2022) and pea FBPase (Pla and Lopez-Gorge, 1981) target enzymes were 179 produced in _E. coli_ BL21(DE3) using the corresponding pET28-derived plasmids described above 180 and in Table S5. Protein expression was induced with 0.2 mM IPTG, followed by incubation for 181 24 h at 30 °C. 182 Cells were harvested, resuspended in 50 mM Tris–HCl (pH 8.0), 150 mM NaCl, 5 mM 183 imidazole and 1 mM PMSF, and disrupted by sonication. His-tagged proteins were purified by 184 Ni-NTA affinity chromatography, eluted with the same buffer containing 500 mM imidazole and [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 185 desalted using PD-10 columns equilibrated with 50 mM Tris–HCl (pH 8.0) and 150 mM NaCl. 186 Proteins recovered predominantly in the insoluble fraction were purified under denaturing 187 conditions and refolded using a urea-based protocol. 188 Proteins were further purified by size-exclusion chromatography using an ENrich SEC70 189 column connected to a Bio-Rad NGC system. Fractions containing the target proteins were 190 analyzed by SDS–PAGE, pooled and stored at −20 °C in 10% (v/v) glycerol. Protein 191 concentrations were determined by absorbance at 280 nm using a NanoDrop spectrophotometer 192 and theoretical extinction coefficients calculated with ProtParam. When required, proteins were 193 concentrated using Amicon Ultra centrifugal filters with a 10 kDa molecular-weight cutoff. 194 **Fructose-1,6-bisphosphatase activity assay** 195 Purified _Synechocystis_ FBP/SBPase and pea FBPase were oxidized by overnight 196 incubation at 4 °C with continuous agitation before the assays. Enzyme activity was determined 197 by measuring the inorganic phosphate released from fructose-1,6-bisphosphate using the Fiske– 198 Subbarow method (García-Cañas et al., 2022). The reagent contained 9 N H₂SO₄, 0.053 M 199 (NH₄)₆Mo₇O₂₄·4H₂O and 0.30 M FeSO₄·7H₂O. Absorbance was measured at 660 nm, and 200 phosphate concentrations were calculated using an inorganic phosphate standard curve. 201 Reaction mixtures contained 1 µM FBP/SBPase or pea FBPase, 1 µM recombinant Trx or 202 chimeric Trx, 15 mM MgSO₄·7H₂O, 15 mM Tris–HCl (pH 8.0) and 50 µM DTT. After 203 preincubation for 15 min at room temperature, reactions were initiated by adding fructose-1,6 204 bisphosphate to 7.15 mM and incubated at 30 °C for 8 min. Reactions were stopped with 1 mL 205 Fiske–Subbarow reagent, and absorbance was measured at 660 nm. 206 **Structural predictions and computational processing** 207 Structures of Trxs and Trx–target complexes were predicted using AlphaFold 3. The four 208 highest-ranked models for each prediction were evaluated based on the predicted template 209 modeling (pTM) score, interface pTM (ipTM) score for complexes, and predicted aligned error 210 (PAE). Models with pTM and ipTM scores above 0.7 and favorable PAE plots were selected for 211 analysis. 212 Structures were visualized using UCSF ChimeraX. Electrostatic surface potentials were 213 calculated using APBS after structure preparation with PDB2PQR (Jurrus et al., 2018), and 214 protein–protein interfaces were analyzed using PICKCLUSTER (Genz et al., 2025). 215 Protein net charges at pH 7.5 were calculated using Protein Calculator, and theoretical 216 isoelectric points were obtained using ExPASy ProtParam. 217 **Statistical analysis** 218 Data from at least three biological replicates were analyzed using GraphPad Prism. Enzymatic 219 activities obtained with each Trx or chimeric Trx were compared with the DTT-only control (No 220 Trx) by one-way ANOVA followed by Dunnett’s multiple-comparisons test. 221 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 222 **RESULTS** 223 **TrxA possesses specific functions beyond its protein abundance** 224 Several proteomic analyses have revealed a high degree of overlap among the target 225 proteins of different Trxs. However, despite this apparent functional redundancy, multiple 226 attempts to eliminate TrxA have been unsuccessful. One possible explanation is that TrxA is more 227 abundant than other Trxs (Mallén-Ponce et al., 2022). To assess whether TrxA essentiality results 228 from this higher cellular abundance, we overexpressed different Trxs and evaluated their ability 229 to compensate for TrxA function. 230 TrxA, TrxB, TrxQ and TrxC were expressed under the control of the arsenite-inducible 231 _arsB_ promoter, generating the parental overexpression strains STXA3P, STXBOEP, STXQOEP 232 and STXCOE1P, respectively (Fig. 1a-b). Unlike the previously characterized STXA2 strain 233 (Mallén-Ponce et al., 2021), these constructs retain the original ribosome-binding site upstream 234 of the _arsB_ promoter, while the endogenous _trxA_ gene remains intact (Fig. S1). In addition to the 235 wild type (WT) background, the same constructs were introduced into the SARSR (SR) 236 background, which lacks the _arsR_ gene encoding the repressor of the _arsBHC_ operon (López 237 Maury et al., 2003). This generated the SRTXA3P, SRTXBOEP, SRTXQOEP and SRTXCOE1P 238 parental strains (Fig. S2a). In contrast to WT-derived overexpression strains, SR-derived strains 239 exhibit constitutive expression of the _arsBHC_ operon and do not respond to arsenite (hereafter 240 inducer). 241 Western blot analysis confirmed that TrxA protein levels in the parental STXA3P strain 242 were approximately 2-fold higher than in the WT strain (Fig. 1c). In the parental STXBOEP and 243 STXQOEP strains, TrxB and TrxQ protein levels were approximately 7- and 35-fold higher than 244 in the WT strain, respectively. These increases resulted in TrxB and TrxQ protein levels 245 comparable to those of endogenous TrxA in WT cells, consistent with the expression levels 246 reported for these Trxs in previous studies (Florencio et al., 2006; Baers et al., 2019; Wang et al., 247 2022; Spät et al., 2023). Comparable protein levels were observed in the SRTXA3P, SRTXBOEP 248 and SRTXQOEP strains, where TrxA expression occurs constitutively (Fig. S2b). Despite Trx 249 overexpression, no detectable phenotypic alterations were observed (Fig. 1d, Fig. S2c). 250 The endogenous _trxA_ gene was then disrupted in these parental backgrounds by insertion 251 of a spectinomycin/streptomycin resistance cassette. PCR analysis showed that the disrupted _trxA_ 252 gene could not be fully segregated in the STXBOE, STXQOE and STXCOE1 strains or in their 253 corresponding SR-derived strains (Fig. 1e, Fig. S2d), indicating that TrxB, TrxQ and TrxC cannot 254 functionally replace TrxA even when expressed at protein levels comparable to those of TrxA. 255 Several attempts to achieve complete chromosome segregation were unsuccessful, even at 1 mM 256 inducer, a concentration sufficient to achieve maximal expression from the _arsB_ promoter. In 257 contrast, complete segregation of the disrupted _trxA_ locus was achieved in the STXA3 and [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 258 SRTXA3 strains (Figs. 1e, Fig. S2d), which accumulated TrxA at levels comparable to those in 259 the WT and SR, respectively (Fig. 1c, Fig. S2b). 260 These TrxA levels were sufficient to prevent any detectable phenotype in the STXA3 261 strain upon inducer removal (Fig. S2e-f), unlike the previously characterized STXA2 strain 262 (Mallén-Ponce et al., 2022). Given that the inability of TrxB, TrxQ and TrxC to replace TrxA 263 appears to reflect specific functions of TrxA rather than differences in protein abundance, we 264 asked whether an alternative, structurally distinct thiol-reducing system could be bypass this 265 requirement. Although NTRC is absent in _Synechocystis_, this enzyme is present in other 266 cyanobacteria and chloroplasts, where it contributes to thiol-based redox regulation using 267 NADPH as the source of reducing power. Its physiological role in cyanobacteria and chloroplasts 268 has been the subject of intense debate in recent years (Cejudo et al., 2021; Mallén-Ponce et al., 269 2022). Because NTRC contains both Trx reductase and Trx domains within a single polypeptide, 270 we investigated whether its expression in _Synechocystis_ could compensate for reduced TrxA 271 levels or functionally replace the FTR/TrxA pathway. To this end, strains expressing NTRC from 272 _Anabaena_ sp. PCC7120 were generated in the _Synechocystis_ WT and STXA2 backgrounds, 273 yielding the SNTRC and STXA2C strains, respectively. For overexpression, the _ntrC_ gene was 274 placed under the control of the constitutive _cpcB_ promoter (Fig. S3a), and complete segregation 275 of NTRC and STXA2C strains was confirmed by PCR analysis (Fig. S3b). Subsequently, 276 expression of NTRC protein was confirmed by western blot analysis (Fig. S3c). Growth analysis 277 revealed that the NTRC expression did not increase the growth rate, and no phenotypic changes 278 were observed (Fig. S3d-e). Likewise, no differences in growth were observed between the 279 STXA2C and STXA2 strains, either in the presence of the inducer or after its removal. These data 280 indicate that heterologous NTRC does not compensate for reduced TrxA levels or functionally 281 replace FTR/TrxA system under these conditions. 282 **Surface properties distinguish the different types of thioredoxin** 283 To better understand the structural basis underlying functional specificity of the different 284 Trxs, we conducted a comparative analysis of their sequence and surface-related properties. All 285 native Trxs contain a fairly similar number of residues at the N- and C-terminal sides relative to 286 the active site (Table 1, Fig. S4). In contrast, their net charges vary among different Trxs (Table 287 1). TrxQ has a net charge of 0, while TrxA, TrxB, and TrxC have net charges of -5.1, -7.1, and 288 5.2, respectively. Accordingly, TrxA and TrxB have similar theoretical pI values, whereas TrxC 289 and particularly TrxQ differ from this pattern (Table 1). The unusually high pI of TrxQ likely 290 reflects the presence of multiple basic residues. Pairwise sequence identities among the four Trxs 291 are relatively low, ranging from 21.5% to 39.1%, with TrxC showing the lowest overall identity 292 to the other _Synechocystis_ Trxs (Table S1, Fig. S4). 293 To further explore structural determinants of specificity, we analyzed the surface 294 properties of individual Trxs. The structures of _Synechocystis_ Trxs were predicted using [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 295 AlphaFold (Jumper et al., 2021; Evans et al., 2022). Consistent with previous studies (Eklund et 296 al., 1991), all Trxs adopt the conserved Trx fold consisting of a central β-sheet flanked by α 297 helices (Fig. 2), with the catalytic motif exposed on their surface. Moreover, all Trxs display a 298 mixed polarity around the catalytic cysteines, including TrxC and its unconventional active site 299 motif (WCGLC). In TrxA, the acidic residues Asp61 and Glu62 form a mildly electronegative 300 patch on α-helix 3 that is highly conserved among photosynthetic organisms (Mallén-Ponce et al., 301 2022). In addition, a conserved loop comprising residues 70-76 contains a conserved Arg73 302 whose guanidine group protrudes from the molecular surface (Fig. 2, Fig. S4). These structural 303 features suggest that surface-exposed residues may therefore play an important role in target 304 recognition. 305 The surface of TrxB is more similar to that of TrxA, although TrxB exhibits a larger 306 electronegative potential on the α-helix 3 and β-strand 3 side of the active site owing to the 307 presence of several acidic residues, including Glu58, Glu72 and Glu90 (Fig. 2, Fig. S4). 308 According to the sequence alignment, these residues correspond to Asn59, Arg73 and Val91, 309 respectively, in TrxA, which contribute neutral or positive character to the equivalent surface 310 region. Interestingly, Glu58 and Glu90 are present in all analyzed cyanobacterial TrxB proteins 311 and represent distinctive features of this Trx type, whereas Glu72 is found in 75% of the analyzed 312 cyanobacterial TrxB proteins (Mallén-Ponce et al., 2022). In TrxQ, a detailed comparison reveals 313 an electropositive patch near the N-terminal region, mainly associated with the α-helix 1 and β 314 strand 2 (Fig. 2). This region contains several conserved basic residues, including Lys4, Lys5, 315 Lys19, Arg48, Lys55, Lys60 and Lys80 (Fig. S4). Finally, TrxC is unique to cyanobacteria and 316 contains a non-canonical active site (Fig. 2). Notably, it harbors a leucine residue instead of the 317 conserved proline typically found between the two catalytic cysteines of canonical Trxs (López 318 Maury et al., 2018). This substitution likely alters the geometry of the active-site motif and may 319 affect both structural stability and redox properties (Collet and Messens, 2010). In addition, 320 structural modeling reveals a shorter and more distant β-strand 4 relative to the active site, 321 associated with the absence of a conserved glycine residue corresponding to Gly92 in TrxA (Fig. 322 2). In TrxC, this position is occupied by a negatively charged Asp105, which interacts with the 323 positively charged Arg107. Together with an additional Arg102-Glu104 interaction, these 324 charged residues form a stabilizing network that may contribute to the distinctive folding of TrxC. 325 These structural features may underlie the inability of TrxC to reduce substrates such as insulin 326 or FBPase (López-Maury et al., 2018). 327 Three glycine residues are conserved among Trxs (Fig. S4). The first, located within the 328 canonical CGPC active site motif, helps maintain the conformation of the catalytic site and 329 contributes to determining the redox potential. The other two glycines flank b-strand 4 and define 330 its length. These residues are structurally important because they occupy equivalent positions in 331 turns found in other Trx structures (Collet and Messens, 2010). [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 332 Overall, this analysis indicates that _Synechocystis_ Trxs share a conserved Trx fold but 333 differ in surface properties that may contribute to functional diversification. TrxA and TrxB are 334 the most structurally and electrostatically similar, although TrxB harbors specific acidic residues 335 (Glu58, Glu72 and Glu90) that enhance the negative electrostatic potential surrounding the active 336 site. These observations suggest that TrxB is the most suitable scaffold for dissecting the surface 337 features that confer TrxA-specific function. 338 **Functional complementation of TrxA by chimeric thioredoxins** 339 To investigate the determinants of TrxA-specific function, we constructed two 340 _Synechocystis_ mutant strains expressing TrxA/TrxB chimeric proteins. These hybrids were 341 generated by interchanging the N- and C-terminal regions of TrxA and TrxB with respect to their 342 active sites (Fig. 3a-b). Using the same strategy described above, the chimeric genes were placed 343 under the control of the arsenite-inducible _arsB_ promoter in a WT background, generating the 344 STXABP and STXBAP strains (Fig. 3c). These resulting strains were verified by PCR analysis 345 and shown to be fully segregated (Fig. 3d). Next, the endogenous _trxA_ gene was disrupted in these 346 strains by inserting a spectinomycin/streptomycin resistance cassette, generating the STXAB and 347 STXBA strains (Fig. 3c). PCR analysis revealed incomplete segregation of the disrupted _trxA_ 348 locus in the STXAB strain, whereas the STXBA strain showed complete segregation (Fig. 3d). 349 Similar segregation patterns were obtained when the same constructs were introduced into the SR 350 background strain, further supporting the robustness of these findings (Fig. S6). These results 351 indicate that the C-terminal region of TrxA contains determinants required for functional 352 replacement of endogenous TrxA. Consistent with this finding, the C-terminal region contains 353 several conserved motifs among cyanobacterial TrxA orthologs and chloroplast _m_ -type Trxs 354 (Mallén-Ponce et al., 2022). 355 To further explore the essential domains within the C-terminal region of TrxA, we 356 generated additional mutant strains expressing TrxB/TrxA chimeric proteins. In these constructs, 357 the N-terminal region of TrxB up to the active site was maintained, whereas increasing portions 358 of the C-terminal region of TrxA were progressively replaced by the corresponding region of 359 TrxB, generating STXB48AP, STXB63AP and STXB86AP strains (Figs. 3a-c). PCR analysis 360 confirmed that these strains were fully segregated (Fig. 3d). Subsequently, the endogenous _trxA_ 361 gene was disrupted in these strains. PCR analysis revealed that, among these new chimeric strains, 362 only STXB48A achieved complete segregation (Fig. 3d), suggesting that the residues essential 363 for cell viability are not restricted to those located adjacent to the active site. A similar pattern 364 was obtained when the SR strain was used as the genetic background (Fig. S6). The essential 365 region retained in STXB48A (residues 49-106) corresponds to α-helices 3 and 4, as well as β 366 strands 3 and 4, and overlaps with highly conserved regions of TrxA (Mallén-Ponce et al., 2022). 367 This contrasts with other Trxs, including TrxB, whose fully conserved region is restricted to a 368 few amino acids near the active site (Mallén-Ponce et al., 2022). [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 369 To assess the functional impact of these structural differences, we examined the growth 370 and phenotypic characteristics of the STXBA and STXB48A strains under induced and non 371 induced conditions. The WT and STXA3 strains were used as controls. Under induced conditions, 372 there were no significant differences in growth between the STXBA strain and the control strains 373 (Fig. 4a-b), whereas the STXB48A strain showed a slight growth delay. Interestingly, removal of 374 the inducer caused a progressive growth retardation mainly in the STXBA strain (Fig. 4a-b), an 375 effect not observed in the STXA3 strain. This difference may be related to the high initial 376 intracellular levels of TrxA protein accumulated in the STXA3 strain under induced conditions 377 (Fig. 1c). In addition, both STXBA and STXB48A strains displayed a distinct phenotypic change 378 upon inducer removal, characterized by reduced levels of chlorophyll _a_ and phycobilins (Fig. 4c, 379 Fig. S7), whereas carotenoid levels were only slightly affected (Fig. 4c). These findings suggest 380 that TrxA functionality is influenced by specific structural determinants located within the 381 conserved C-terminal regions. 382 **Characterization of the chimeric thioredoxins** 383 The active site of Trxs is located in a surface-exposed loop at the beginning of α-helix 2, 384 with the catalytic cysteine accessible to the solvent (Fig. 2). Residues near the active site 385 contribute to Trx target recognition, including photosynthetic Trxs (Collet and Messens, 2010; 386 Zaffagnini et al., 2018). To investigate the structural determinants underlying TrxA specificity, 387 we analyzed the interaction of TrxA and the TrxA/TrxB chimeric proteins with the bifunctional 388 enzyme FBP/SBPase from _Synechocystis_, a CB cycle enzyme essential under photoautotrophic 389 conditions and previously identified as a TrxA target (Mallén-Ponce et al., 2021). 390 As a first step, we examined the molecular surfaces of the different chimeric proteins. _In_ 391 _silico_ analysis predicted that TrxAB has a net charge of -8.1 and a theoretical pI of 4.7 (Table 1). 392 Electrostatic surface modeling revealed an extended electronegative region surrounding the active 393 site and α-helix 3, associated with the presence of acidic residues including Asp9, Glu59, Glu73 394 and Glu91, compared to TrxA and even TrxB (Fig. 5a). This distribution generates a continuous 395 negative electrostatic potential around the active site. In addition, negatively charged residues in 396 the N-terminal region of TrxAB, derived from the N-terminal region of TrxB, generate 397 electronegative patches on the opposite side of the active site (Fig. 5a). In contrast to TrxAB, 398 TrxBA has a predicted net charge and theoretical pI closer to those of TrxA (Table 1). Structural 399 modeling revealed that the region surrounding the active site has a mixed polarity highly similar 400 to TrxA (Fig. 5a). This similarity likely reflects the contribution of the TrxA derived C-terminal 401 region, whereas TrxB N-terminal region is positioned away from the active site, where it has a 402 more neutral character than the corresponding region in TrxA. 403 We next analyzed TrxB48A, TrxB63A and TrxB86A, which exhibit less negative net 404 charges and higher theoretical pI values than TrxA (Table 1). Structural modeling revealed that [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 405 these chimeric Trxs retain the Trx fold but display marked differences in the electrostatic potential 406 of their solvent-accessible surfaces compared with native TrxA (Fig. 2, Fig. 5a). 407 Sequence and structural analysis uncovered differences among the chimeric proteins that may 408 contribute to TrxA function. In TrxB48A, α-helix 2 lacks the two negatively charged residues 409 present in TrxA, Asp42 and Glu43 (Fig. 3b). Consistently, this region displays reduced negative 410 electrostatic potential compared with TrxBA, which maintains a surface polarity similar to that 411 of TrxA (Fig. 5a). However, these electrostatic differences do not appear to be essential for 412 viability, as fully segregated strains lacking TrxA were obtained from cells expressing either 413 TrxBA or TrxB48A. 414 Modelling TrxB63A revealed a more extended electronegative region around the active 415 site (Fig. 5a). Because the region downstream of residue 63 is derived from TrxA, this 416 electrostatic redistribution may reflect the combined contribution of acidic residues retained 417 within the TrxB-derived N-terminal region, together with changes in their orientation and solvent 418 exposure in the chimeric structure (Fig. 3b). Previous studies showed that substitutions of acidic 419 residues within this region increased the efficiency of chloroplast FBPase activation, supporting 420 a role for local electrostatic properties in target recognition (Eklund et al., 1991; de Lamotte 421 Guery et al., 1991; Wangensteen et al., 2001). The negative electrostatic character is further 422 enhanced in TrxB86A, which retains the TrxB-derived Glu72 at the position equivalent to the 423 positively charged Arg73 of TrxA (Fig. 3b, Fig. S4). This position is located within a region 424 highly conserved among cyanobacterial TrxA orthologs and chloroplast _m_ -type Trxs (Mallén 425 Ponce et al., 2022), and may contribute to the structural organization of the target-recognition 426 surface. Overall, although all chimeric proteins retain the canonical Trx fold and active-site motif, 427 notable differences are observed in the electrostatic potential of their solvent-accessible surfaces. 428 Several studies have suggested that protein-protein interactions between Trxs and their 429 targets proteins depend on complementary electrostatic surfaces (Soulié et al., 1985; Rodriguez 430 Suarez et al., 1997; Mora-Garcı́a et al., 1998; Collin et al., 2003; Balmer et al., 2004). We 431 therefore examined the potential impact of changes in the electrostatic potential around the active 432 site using the enzyme FBP/SBPase, a previously identified TrxA target in _Synechocystis_ (Mallén 433 Ponce et al., 2021). Crystal structures of FBP/SBPase from _Synechocystis_ and _Synechococcus_ 434 revealed the structural basis for its redox regulation, in which Cys75 and Cys84 form a regulatory 435 disulfide bond (Feng et al., 2014; Cotton et al., 2015). However, biochemical studies in 436 _Synechocystis_ suggested that Cys75 and Cys99 may rearrange to form a disulfide bond in the 437 inactive enzyme state (Feng et al., 2014). Consistently, _in vivo_ redox analysis of FBP/SBPase 438 showed higher oxidation levels in a TrxA knockdown mutant than in WT, resulting in reduced 439 enzyme activity (Mallén-Ponce et al., 2021). Given the differences observed in the electrostatic 440 properties of the chimeric Trxs (Fig. 5a), we used AlphaFold 3 to generate a computational model 441 of the TrxA-FBP/SBPase interaction. The predicted models showed robust quality metrics (ipTM [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 442 and pTM scores >0.7 and a favorable predicted aligned error (PAE) plot), supporting the 443 reliability of the predicted interaction. As previously described, _Synechocystis_ FBP/SBPase 444 adopts a tetrameric structure formed by C1-C4 and C2-C3 dimer pairs, each composed of two 445 monomers (Feng et al., 2014 ). The redox-regulated Cys75-Cys84 disulfide bond is located in a 446 partially flexible loop between two β-sheets. In our model, the surface electrostatic potential 447 calculated using the APBS algorithm revealed an extended electronegative surface around the 448 regulatory Cys residues of FBP/SBPase (Fig. 5c). 449 We next performed molecular docking analysis between TrxA and dimeric FBP/SBPase 450 to investigate their potential interactions. Alphafold 3 predicted several docking models of TrxA 451 on the FBP/SBPase surface. In the four highest-scoring models, the nucleophilic Cys of the TrxA 452 active site was positioned close to the regulatory Cys84 of FBP/SBPase (Fig. 5d). To further 453 investigate the TrxA-FBP/SBPase interaction, we performed clustering analysis using 454 PICKCLUSTER ( Genz et al., 2023) . This analysis identified potential interaction interfaces 455 between TrxA and FBP/SBPase, highlighting residues involved in coordinating TrxA binding to 456 the target protein. Across the predicted models, the interaction network was consistently anchored 457 by residues surrounding the TrxA active site (Fig. S8). Recognition of FBP/SBPase was primarily 458 mediated by Cys32 of TrxA together with adjacent active-site residues (Fig. S8). This interaction 459 was further supported by a conserved surface region encompassing residues Gly71-Thr77, which 460 includes Arg73 (Fig. S9). Several models also suggested additional contact regions involving 461 charged residues such as Asp61 (Fig. S8). In all cases, these regions are highly conserved among 462 different TrxA ( _m_ -type) proteins and are likely directly involved in the interaction with their target 463 proteins. In FBP/SBPase, residues within the flexible recognition loop adjacent to Cys84 (Asp79 464 Pro86) appear to stabilize the loop conformation and contribute to the interaction with TrxA (Fig. 465 S9). 466 Molecular docking analysis of TrxB and dimeric FBP/SBPase provided similar results, 467 revealing an interaction network involving three regions that include the active site in the different 468 predicted models (Fig. S8). Unlike TrxA, these regions of TrxB contain several negatively 469 charged residues, including Glu72 and Glu90, which are predicted to interact with positively 470 charged residues of FBP/SBPase, such as Arg50 and Lys81. 471 We next analyzed the potential interactions between the chimeric Trxs and FBP/SBPase. 472 Molecular docking analysis predicted that TrxBA interacted with FBP/SBPase through regions 473 similar to those identified for TrxA across the different models. In contrast, only one of the four 474 highest-scoring TrxAB models positioned its active site close to Cys84 of FBP/SBPase, consistent 475 with a potential redox-regulatory interaction (Fig. S8). The remaining chimeric Trxs also showed 476 similar predicted interaction regions, with some specific differences among them. The main 477 differences between TrxB48A and TrxB63A involved the turn connecting β-strand 2 and α-helix 478 3. Changes in this region were associated with a rearrangement of the predicted contacts of [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 479 TrxB63A and also TrxB86A with FBP/SBPase. Specifically, in the models where TrxB63A and 480 TrxB86A interacted through the loop containing Cys85, Glu58 was positioned close to Lys81 of 481 FBP/SBpase, among other predicted contacts. Interestingly, previous studies in _E. coli_ and plants 482 have shown that substitutions of residues in this region affect the activity of TrxA ( _m_ -type) (de 483 Lamotte-Guery et al., 1991; Wangensteen et al., 2001; Yokochi et al., 2019). 484 Given the essential role of FBP/SBPase under phototrophic conditions and its redox 485 regulation by TrxA, we next examined the effects of different Trxs on FBPase activity (Fig. 5e). 486 In the absence of both DTT and Trxs, FBP/SBPase displayed only basal activity, whereas addition 487 of TrxA or TrxB increased FBPase activity, with TrxA showing a markedly stronger effect. The 488 TrxBA chimera promoted activation to a level similar to that observed with TrxA, whereas TrxAB 489 failed to activate FBP/SBPase. These results are consistent with the idea that preservation of key 490 TrxA-like interface regions favors efficient functional regulation of FBP/SBPase. Interestingly, 491 subsequent analysis of the TrxB48A, TrxB63A and TrxB86A chimeras showed that TrxB48A 492 retained an activation capacity comparable to that of TrxA, whereas T rxB63A and TrxB86A 493 failed to activate the enzyme. Together, these results indicate that specific structural determinants 494 within TrxA are required for efficient activation of FBP/SBPase. Moreover, the predicted 495 structural rearrangements between TrxB48A and TrxB63A may help explain t he loss of 496 FBP/SBPase activation observed in TrxB63A and TrxB86A. 497 TrxA ( _m_ -type) also interacts with and regulates multiple target proteins in photosynthetic 498 eukaryotes, including several CB cycle enzymes. Therefore, we analyzed the activation of pea 499 chloroplast fructose-1,6-bisphosphatase by TrxA, TrxB and the different chimeric Trxs. Pea 500 FBPase is a homotetramer composed of approximately 40 kDa subunits, each of which contains 501 a regulatory disulfide bond between the conserved Cys153 and Cys173 residues on its surface 502 (Chiadmi et al., 1999) . Reduction of this disulfide enables activation by chloroplast _f-_ and _m-_ type 503 Trxs under light conditions ( López-Jaramillo et al., 1997; Neira et al., 2024) . Previous studies 504 have shown that Pea FBPase activation is primarily driven by electrostatic interactions, with 505 hydrophobic interactions playing a secondary role ( Pla and Lopez-Gorge, 1981; Soulié et al., 506 1985; Crawford et al., 1986; Mora-Garcı́a et al., 1998; Wangensteen et al., 2001 ). Consistent with 507 this, Pea FBPase displays a large electronegative surface surrounding the two redox-sensitive 508 cysteines, a feature conserved among plant orthologs ( Marcus and Harrsch, 1990). 509 Molecular docking analysis identified a TrxA- Pea FBPase model in which the nucleophilic 510 Cys of TrxA was positioned close to Cys153 of Pea FBPase (Fig. S10b). The predicted interface 511 involved residues adjacent to the TrxA active site, as well as residues located between α-helix 3 512 and β-strand 3 (Arg73-Thr77). A second flexible region, located between β-strand 4 and α-helix 513 4 (Gly92-Pro95), was also positioned near the interface. Together, these regions positioned TrxA 514 close to Cys153 of Pea FBPase and to adjacent surface regions that may contribute to the 515 recognition of Pea FBPase (Fig. S10b). [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 516 Docking simulations performed with TrxB and the chimeric Trxs did not identify clearly 517 productive interaction near Cys153. Only one TrxB48A model approached the regulatory site, 518 although the predicted distance was not compatible with an efficient redox interaction. To 519 examine whether the activity of the chimeras was target dependent, we also tested their ability to 520 activate pea FBPase _in vitro_ . The activation pattern of Pea FBPase was broadly consistent with 521 that observed for the cyanobacterial FBP/SBPase. TrxA, TrxBA, and TrxB48A displayed similar 522 activation levels, whereas TrxB showed somewhat lower activity. Progressive replacement of the 523 C-terminal region of TrxA with the corresponding region of TrxB resulted in a progressive loss 524 of activation capacity, which was markedly reduced with TrxB63A and no longer detectable with 525 TrxB86A (Fig. S10c). 526 527 **DISCUSSION** 528 The inability to eliminate the single _trxA_ gene in _Synechocystis_, as well as all Trx- _m_ 529 isoforms present in plant chloroplasts strongly supports the idea that these proteins perform non 530 redundant functions. However, TrxA (or Trx- _m_ ) is also the most abundant Trx both in 531 cyanobacteria and chloroplasts, respectively (Florencio et al., 2006; Okegawa and Motohashi, 532 2015), raising the possibility that its essentiality could be linked to its high intracellular 533 abundance. Given that _Synechocystis_ encodes several additional Trxs, these proteins might 534 compensate for the loss of TrxA if expressed at comparable levels. To address this possibility, 535 strains overexpressing different Trxs were generated and evaluated for their ability to functionally 536 substitute for TrxA. 537 Our results demonstrate that overexpression of TrxB, TrxQ or TrxC failed to compensate 538 for the loss of TrxA (Fig. 1a, Fig. 1b), indicating that TrxA essentiality cannot be explained solely 539 by protein abundance. Overexpression restored TrxA protein levels to values comparable to those 540 in the WT strain, whereas TrxB and TrxQ accumulated to approximately 7- and 35-fold higher 541 levels, respectively (Fig. 1c). These increases are consistent with the lower endogenous 542 abundance previously reported for these Trxs (Florencio et al., 2006; Baers et al., 2019; Wang et 543 al., 2022; Spät et al., 2023), bringing their levels into a range comparable to that of TrxA in the 544 WT strain. Nevertheless, neither TrxB nor TrxQ could functionally replace TrxA under these 545 conditions. Together, these findings strongly support the existence of structural and/or functional 546 properties specific to TrxA that cannot be compensated for by other Trxs, even when the latter 547 accumulate to substantially higher levels. 548 Heterologous expression of NTRC, an alternative thiol-reducing system, did not produce 549 any detectable phenotype in either the WT strain or STXA2, which has reduced TrxA levels (Fig. 550 S3d-e). This finding is consistent with the idea that TrxA performs non-redundant functions that 551 cannot be compensated for simply by introducing an alternative source of reducing power. [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 552 Nevertheless, the involvement of other redox systems, including glutaredoxins (Sánchez-Riego 553 et al., 2013), cannot be excluded. 554 The physicochemical basis of Trx specificity in photosynthetic organisms has been 555 extensively investigated, particularly in cyanobacterial TrxA, its chloroplast _m_ -type ortholog and 556 chloroplast Trx- _f_ . Comparative biochemical analyses in plants showed that trx- _m_ and trx- _f_ exhibit 557 distinct target preferences, with Trx- _m_ more efficiently activating NADP-malate dehydrogenase 558 and Trx- _f_ preferentially activating FBPase (Wolosiuk et al., 1979). Crystallographic analyses of 559 spinach Trx- _f_ and Trx- _m_ revealed differences in surface charge distribution and loop 560 conformations surrounding the WCGPC active site motif (Capitani et al., 2000), while three 561 dimensional modeling suggested that subtle variations in electrostatic surface potential contribute 562 to preferential target recognition (Collin et al., 2003). Similarly, crystallographic analyses of 563 chloroplast Trx- _f_ isoforms from _Chlamydomonas reinhardtii_ revealed a distinctive electropositive 564 surface surrounding the active site that complements the electronegative surface of their target 565 proteins (Lemaire et al., 2018). 566 Our structural analyses revealed that the cyanobacterial Trxs TrxA, TrxB and TrxQ share 567 a high degree of structural similarity with chloroplast Trxs, both in their canonical fold and in the 568 electrostatic properties of the region surrounding the active site (Collin et al., 2003; Lemaire et 569 al., 2018; Table 1; Fig. 2). Nevertheless, these proteins displayed clear differences in surface 570 charge distribution and loop conformation (Fig. 2), particularly in regions flanking the catalytic 571 motif, suggesting distinct capacities to interact with specific redox partners. The coexistence of 572 overall structural conservation and local divergence is consistent with the idea that the molecular 573 determinants underlying Trx-substrate specificity were already established in the prokaryotic 574 ancestors of plastids and subsequently retained during chloroplast evolution. 575 Although the structures of many Trx target proteins have been resolved, relatively few Trx– 576 target complexes have been characterized structurally, limiting our understanding of the 577 interaction interfaces and molecular contacts that determine target specificity. Previous modeling 578 and comparative analyses in _Chlamydomonas reinhardtii_ showed that Trx- _f_ has an electropositive 579 crown surrounding the active site and a flexible hinge region downstream of α-helix 1, both of 580 which may contribute to target recognition (Collin et al., 2003; Lemaire et al., 2018). By contrast, 581 Trx- _m_ displays a mixed electrostatic surface around its catalytic cysteines, whereas Trx- _x_ and 582 Trx- _y_ are characterized by predominantly neutral or electronegative surfaces. These observations 583 are consistent with the structural models generated in this study and support the idea that both the 584 distribution of surface charges and its spatial arrangement relative to the active site contribute to 585 Trx target specificity. 586 Consistent with these observations, the structural models generated in this study indicate 587 that TrxA has a more balanced electrostatic surface, whereas TrxB (Trx- _x_ type) and TrxQ (Trx- _y_ 588 type) display more localized charge distributions. The spatial arrangement of these charges [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 589 appears to influence target recognition: TrxB displays a conserved expanded electronegative 590 surface surrounding the active site, which may be less favorable for productive interactions with 591 CB cycle enzymes. By contrast, TrxQ exhibits a localized basic patch near its N-terminal region 592 that could affect its recognition of complementary target surfaces (Collin et al., 2003; Fig. 2). 593 Consistent with their established roles, TrxB (Trx- _x_ type) and TrxQ (Trx- _y_ type) have been 594 primarily associated with antioxidant functions and have shown lower efficiency in regulating CB 595 cycle enzymes. These differences support the idea that subtle structural and electrostatic 596 variations contribute to the functional diversification of Trxs in photosynthetic organisms (Collin 597 et al., 2003; Collin et al., 2004; Pérez-Pérez et al., 2009). 598 In our study, the generation of chimeric Trxs provided direct evidence that the C-terminal 599 region of TrxA plays a central role in defining its essential function. The ability of TrxBA, but 600 not TrxAB, to complement the loss of TrxA indicates that structural determinants required for 601 productive interactions with target proteins are associated with the C-terminal region. This 602 asymmetry suggests that the region downstream of the active site contributes to positioning the 603 catalytic cysteines in a configuration compatible with efficient electron transfer. Importantly, this 604 region encompasses α-helices 3 and 4 and adjacent β-strands, which were further delimited as 605 essential in our progressive chimeric constructs. Alterations within this segment led to loss of 606 function despite preservation of the overall Trx fold, indicating that global structural integrity 607 alone is not sufficient to sustain activity. This interpretation is consistent with previous 608 mutagenesis and biochemical studies showing that substitutions of residues surrounding the active 609 site can affect Trx-dependent activation of target enzymes, such as FBPase and NADP-malate 610 dehydrogenase, without disrupting the canonical fold. Studies of plant Trx-m proteins have 611 likewise shown that sequence variations in conserved regions can affect FBPase activation 612 (López-Jaramillo et al., 1997). In particular, several _in vitro_ studies have introduced substitutions 613 that neutralize or remove negatively charged residues in regions flanking the active site, resulting 614 in significant changes in target enzyme activation efficiency. These modifications alter the 615 electrostatic environment and the orientation of the interacting surfaces, potentially affecting 616 target recognition and positioning rather than the catalytic chemistry of the active-site cysteines 617 (de Lamotte-Guery et al., 1991; López-Jaramillo et al., 1997; Wangensteen et al., 2001; Yokochi 618 et al., 2019). Indeed, removal of acidic residues or introduction of basic residues has even been 619 reported to enhanced target enzyme activation, highlighting the role of local charge distribution 620 in modulating Trx specificity and catalytic efficiency. 621 The _in vitro_ activation assays support the idea that Trx specificity depends on localized 622 structural determinants rather than on conservation of the canonical Trx fold alone. The activation 623 patterns observed with both _Synechocystis_ FBP/SBPase and Pea FBPase were broadly consistent 624 with the _in vivo_ complementation results (Fig. 4, Fig. 5, Fig. S8), suggesting that structural 625 features retained in functional chimeras contribute to productive interactions with physiologically [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 626 relevant targets. Notably, TrxA, TrxBA, and TrxB48A displayed similar activation levels towards 627 Pea FBPase, whereas TrxB showed somewhat lower activity (Fig. S10c). Further replacement of 628 the C-terminal region of TrxA progressively impaired activation, which was markedly reduced 629 with TrxB63A and no longer detectable with TrxB86A (Fig. S10c). This pattern supports a 630 contribution of the C-terminal region of TrxA to Pea FBPase activation, while indicating that the 631 extent of this contribution depends on the structural context of the chimera. However, the 632 differences between the in vivo complementation and in vitro activation assays indicate that the 633 structural requirements for TrxA-dependent essential functions cannot be fully inferred from the 634 activation of a single target enzyme. 635 Overall, our results extend previous observations from single-residue substitutions to larger 636 structural modules, demonstrating that the C-terminal region of TrxA acts as a functional unit 637 integrating key structural features required for target recognition and efficient electron transfer. 638 This modular behavior is further supported by the ability of the TrxBA chimera to restore 639 function, indicating that transplantation of this region into a TrxB scaffold is sufficient to confer 640 substantial TrxA-like functionality. 641 **Author contributions** 642 M.J.M.-P., M.J.H. and F.J.F. plan and designed the research; M.J.M.-P., J.M-L. and M.J.H. 643 performed the research; M.J.M.-P. and M.J.H. analyze the data; M.J.M.-P. wrote the article and 644 M.J.H. and F.J.F. contributed the final article version; M.J.M.-P. and M.J.H. supervised the work; 645 M.J.M.-P. and F.J.F. funding acquisition. 646 **Acknowledgements** 647 We thank Marian Sahrawy for providing the pea FBPase expression construct and Luis 648 López-Maury for providing the _Synechocystis_ FBP/SBPase expression construct. 649 650 **Supplementary Material** 651 **Figure S1** . Schematic representation of the _arsB_ promoter region used for thioredoxin expression. 652 **Figure S2** . Construction and characterization of thioredoxin-overexpressing strains in the SR 653 background. 654 **Figure S3** . Construction and characterization of NTRC-expressing strains. 655 **Figure S4** . Multiple sequence alignment of _Synechocystis_ thioredoxins. 656 **Figure S5** . Multiple sequence alignment of TrxA and _m_ -type thioredoxins from photosynthetic 657 organisms. 658 **Figure S6** . PCR Verification of strains expressing TrxA/TrxB chimeric thioredoxins in the SR 659 background. 660 **Figure S7** . Chlorophyll concentration in strains expressing TrxA/TrxB chimeric thioredoxins. 661 **Figure S8** . AlphaFold analysis of residue interactions between TrxA, TrxB and TrxA/TrxB 662 chimeric thioredoxins with _Synechocystis_ FBP/SBPase. [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 663 **Figure S9** . AlphaFold model of the TrxA-SyFBP/SBPase interface. 664 **Figure S10** . Structural and functional analysis of Pea FBPase activation by TrxA, TrxB and 665 chimeric thioredoxins. 666 **Table S1.** Sequence identity among _Synechocystis_ thioredoxins. 667 **Table S2.** Synechocystis strains used in this study. 668 **Table S3.** Oligonucleotides used in this study. 669 **Table S4.** Antibodies used in this study. 670 **Table S5.** Plasmids used in this study. 671 672 **Funding** 673 This work was supported by grant PID2022-138317NB-I00 to F.J.F., financed by 674 MCIU/AEI/10.13039/501100011033 and by "FEDER, Una manera de hacer Europa", and by 675 Junta de Andalucía (group BIO-0284); and by a Ramón y Cajal fellowship to M.J.M.-P. 676 (RYC2024-051248-I), funded by the Spanish Ministry of Science, Innovation and Universities. 677 678 **Conflicts of Interest** 679 The authors declare no conflict of interest. 680 681 **Data availability statement** 682 All data are incorporated into the article or its online supplementary material. 683 684 **REFERENCES** 685 **Baers LL, Breckels LM, Mills LA, Gatto L, Deery MJ, Stevens TJ, Howe CJ, Lilley KS,** 686 **Lea-Smith DJ** (2019) Proteome mapping of a cyanobacterium reveals distinct 687 compartment organization and cell-dispersed metabolism. Plant Physiol **181** : 1721–1738 688 **Balmer Y, Koller A, Val G del, Schürmann P, Buchanan BB** (2004) Proteomics uncovers 689 proteins interacting electrostatically with thioredoxin in chloroplasts. Photosynth Res **79** : 690 275–280 691 **Balsera M, Buchanan BB** (2019) Evolution of the thioredoxin system as a step enabling 692 adaptation to oxidative stress. Free Radic Biol Med **140** : 28–35 693 **Balsera M, Uberegui E, Susanti D, Schmitz RA, Mukhopadhyay B, Schürmann P,** 694 **Buchanan BB** (2013) Ferredoxin:thioredoxin reductase (FTR) links the regulation of 695 oxygenic photosynthesis to deeply rooted bacteria. Planta **237** : 619–635 696 **Buey RM, Arellano JB, López-Maury L, Galindo-Trigo S, Velázquez-Campoy A, Revuelta** 697 **JL, de Pereda JM, Florencio FJ, Schürmann P, Buchanan BB, et al** (2017a) 698 Unprecedented pathway of reducing equivalents in a diflavin-linked disulfide 699 oxidoreductase. Proc Natl Acad Sci U S A **114** : 12725–12730 700 **Buey RM, Galindo-Trigo S, López-Maury L, Velázquez-Campoy A, Revuelta JL, Florencio** 701 **FJ, de Pereda JM, Schürmann P, Buchanan BB, Balsera M** (2017b) A new member [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 702 of the thioredoxin reductase family from early oxygenic photosynthetic organisms. Mol 703 Plant **10** : 212–215 704 **Capitani G, Marković-Housley Z, DelVal G, Morris M, Jansonius JN, ürmann P** (2000) 705 Crystal structures of two functionally different thioredoxins in spinach chloroplasts. J 706 Mol Biol **302** : 135–154 707 **Cejudo FJ, González M-C, Pérez-Ruiz JM** (2021) Redox regulation of chloroplast metabolism. 708 Plant Physiol **186** : 9–21 709 **Chiadmi M, Navaza A, Miginiac‐Maslow M, Jacquot J, Cherfils J** (1999) Redox signalling 710 in the chloroplast: structure of oxidized pea fructose‐1,6‐bisphosphate phosphatase. 711 EMBO J **18** : 6809–6815 712 **Collet J-F, Messens J** (2010) Structure, function, and mechanism of thioredoxin proteins. 713 Antioxid Redox Signal **13** : 1205–1216 714 **Collin V, Issakidis-Bourguet E, Marchand C, Hirasawa M, Lancelin J-M, Knaff DB,** 715 **Miginiac-Maslow M** (2003) The _Arabidopsis_ plastidial thioredoxins New functions and 716 new insights into specificity. J Biol Chem **278** : 23747–23752 717 **Collin V, Lamkemeyer P, Miginiac-Maslow M, Hirasawa M, Knaff DB, Dietz K-J, Issakidis-** 718 **Bourguet E** (2004) Characterization of Plastidial Thioredoxins from Arabidopsis 719 Belonging to the New y-Type. Plant Physiol **136** : 4088–4095 720 **Cotton CAR, Kabasakal BV, Miah NA, Murray JW** (2015) Structure of the dual-function 721 fructose-1, 6/sedoheptulose-1, 7-bisphosphatase from _Thermosynechococcus elongatus_ 722 bound with sedoheptulose-7-phosphate. Acta Crystallogr Sect F Struct Biol Commun **71** : 723 1341–1345 724 **Crawford NA, Yee BC, Hutcheson SW, Wolosiuk RA, Buchanan BB** (1986) Enzyme 725 regulation in C4 photosynthesis: Purification, properties, and activities of thioredoxins 726 from C4 and C3 plants. Arch Biochem Biophys **244** : 1–15 727 **Doello S, Shvarev D, Theune M, Sauerwein J, Klon A, Keskin E, Boehm M, Gutekunst K,** 728 **Forchhammer K** (2024) Structural basis of the allosteric regulation of cyanobacterial 729 glucose-6-phosphate dehydrogenase by the redox sensor OpcA. Proc Natl Acad Sci **121** : 730 e2411604121 731 **Eklund H, Gleason FK, Holmgren A** (1991) Structural and functional relations among 732 thioredoxins of different species. Proteins Struct Funct Bioinforma **11** : 13–28 733 **Evans R, O’Neill M, Pritzel A, Antropova N, Senior A, Green T, Žídek A, Bates R, Blackwell** 734 **S, Yim J, et al** (2022) Protein complex prediction with AlphaFold-Multimer. bioRxiv 735 2021.10.04.463034 736 **Feng L, Sun Y, Deng H, Li D, Wan J, Wang X, Wang W, Liao X, Ren Y, Hu X** (2014) 737 Structural and biochemical characterization of fructose 1,6/sedoheptulose 1,7 738 bisphosphatase from the cyanobacterium _Synechocystis_ strain 6803. FEBS J **281** : 916– 739 926 740 **Florencio FJ, Pérez-Pérez ME, López-Maury L, Mata-Cabana A, Lindahl M** (2006) The 741 diversity and complexity of the cyanobacterial thioredoxin systems. Photosynth Res **89** : 742 157–171 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 743 **García-Cañas R, Florencio FJ, López-Maury L** (2022) Back to the future: Transplanting the 744 chloroplast TrxF–FBPase–SBPase redox system to cyanobacteria. Front Plant Sci **13** : 745 1052019 746 **Geigenberger P, Thormählen I, Daloso DM, Fernie AR** (2017) The unprecedented versatility 747 of the plant thioredoxin system. Trends Plant Sci **22** : 249–262 748 **Genz LR, Mulvaney T, Nair S, Topf M** (2023) PICKLUSTER: a protein-interface clustering 749 and analysis plug-in for UCSF ChimeraX. Bioinformatics **39** : btad629 750 **Genz LR, Nair S, Nagar N, Topf M** (2025) Assessing scoring metrics for A LPHA F OLD 2 and 751 A LPHA F OLD 3 protein complex predictions. Protein Sci **34** : e70327 752 **Gerbling K-P, Steup M, Latzko E** (1986) Fructose 1,6-bisphosphatase form B from 753 _Synechococcus leopoliensis_ hydrolyzes both fructose and sedoheptulose bisphosphate. 754 Plant Physiol **80** : 716–720 755 **Gurrieri L, Fermani S, Zaffagnini M, Sparla F, Trost P** (2021) Calvin–Benson cycle 756 regulation is getting complex. Trends Plant Sci **26** : 898–912 757 **Holmgren A, Johansson C, Berndt C, Lonn ME, Hudemann C, Lillig CH** (2005) Thiol redox 758 control via thioredoxin and glutaredoxin systems. Biochem Soc Trans **33** : 1375–1377 759 **Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, Tunyasuvunakool K,** 760 **Bates R, Žídek A, Potapenko A, et al** (2021) Highly accurate protein structure 761 prediction with AlphaFold. Nature **596** : 583–589 762 **Jurrus E, Engel D, Star K, Monson K, Brandi J, Felberg LE, Brookes DH, Wilson L, Chen** 763 **J, Liles K, et al** (2018) Improvements to the APBS biomolecular solvation software suite. 764 Protein Sci **27** : 112–128 765 **de Lamotte-Guery F, Miginiac-Maslow M, Decottignies P, Stein M, Minard P, Jacquot J-P** 766 (1991) Mutation of a negatively charged amino acid in thioredoxin modifies its reactivity 767 with chloroplastic enzymes. Eur J Biochem **196** : 287–294 768 **Lemaire SD, Lombardi G, Mancini A, Carbone A, Henri J** (2025) Functional recoding of 769 Chlamydomonas reinhardtii thioredoxin type-h into photosynthetic type-f by switching 770 selectivity determinants. Front Plant Sci. doi: 10.3389/fpls.2025.1554272 771 **Lemaire SD, Tedesco D, Crozet P, Michelet L, Fermani S, Zaffagnini M, Henri J** (2018) 772 Crystal structure of chloroplastic thioredoxin f2 from _Chlamydomonas reinhardtii_ reveals 773 distinct surface properties. Antioxidants. doi: 10.3390/antiox7120171 774 **López-Jaramillo J, Chueca A, Jacquot JP, Hermoso R, Lazaro JJ, Sahrawy M, Gorge JL** 775 (1997) High-yield expression of pea thioredoxin m and assessment of its efficiency in 776 chloroplast fructose-1,6-bisphosphatase activation. Plant Physiol **114** : 1169–1175 777 **López-Maury L, Florencio FJ, Reyes JC** (2003) Arsenic sensing and resistance system in the 778 cyanobacterium _Synechocystis_ sp. strain PCC 6803. J Bacteriol **185** : 5363–5371 779 **López-Maury L, Heredia-Martínez LG, Florencio FJ** (2018) Characterization of TrxC, an 780 atypical thioredoxin exclusively present in cyanobacteria. Antioxid Basel Switz **7** : 164 781 **Mallén-Ponce MJ, Florencio FJ, Huertas MJ** (2024) Thioredoxin A regulates protein synthesis 782 to maintain carbon and nitrogen partitioning in cyanobacteria. Plant Physiol **195** : 2921– 783 2936 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 784 **Mallén-Ponce MJ, Huertas MJ, Florencio FJ** (2022) Exploring the diversity of the thioredoxin 785 systems in cyanobacteria. Antioxidants. doi: 10.3390/antiox11040654 786 **Mallén-Ponce MJ, Huertas MJ, Sánchez-Riego AM, Florencio FJ** (2021) Depletion of m-type 787 thioredoxin impairs photosynthesis, carbon fixation, and oxidative stress in 788 cyanobacteria. Plant Physiol **187** : 1325–1340 789 **Marcus F, Harrsch PB** (1990) Amino acid sequence of spinach chloroplast fructose-1,6790 bisphosphatase. Arch Biochem Biophys **279** : 151–157 791 **Markwell MAK, Haas SM, Bieber LL, Tolbert Ne** (1978) A modification of the Lowry 792 procedure to simplify protein determination in membrane and lipoprotein samples. Anal 793 Biochem **87** : 206–210 794 **Michelet L, Zaffagnini M, Morisse S, Sparla F, Pérez-Pérez ME, Francia F, Danon A,** 795 **Marchand C, Fermani S, Trost P** (2013) Redox regulation of the Calvin–Benson cycle: 796 something old, something new. Front Plant Sci **4** : 470 797 **Mihara S, Sugiura K, Yoshida K, Hisabori T** (2020) Thioredoxin targets are regulated in 798 heterocysts of cyanobacterium _Anabaena_ sp. PCC 7120 in a light-independent manner. J 799 Exp Bot **71** : 2018–2027 800 **Mora-Garcı́a S, Rodrı́guez-Suárez R, Wolosiuk RA** (1998) Role of Electrostatic Interactions 801 on the Affinity of Thioredoxin for Target Proteins: Recognition of chloroplast fructose802 1,6-bisphosphatase by mutant _Escherichia coli_ thioredoxins. J Biol Chem **273** : 16273– 803 16280 804 **Muller EG, Buchanan BB** (1989) Thioredoxin is essential for photosynthetic growth. The 805 thioredoxin m gene of _Anacystis nidulans_ . J Biol Chem **264** : 4008–4014 806 **Navarro F, Florencio FJ** (1996) The cyanobacterial thioredoxin gene is required for both 807 photoautotrophic and heterotrophic growth. Plant Physiol **111** : 1067–1075 808 **Neira JL, Palomino-Schätzlein M, Rejas V, Traverso JA, Rico M, López-Gorgé J, Chueca** 809 **A, Cámara-Artigas A** (2024) Three-dimensional solution structure, dynamics and 810 binding of thioredoxin m from _Pisum sativum_ . Int J Biol Macromol **262** : 129781 811 **Okegawa Y, Motohashi K** (2015) Chloroplastic thioredoxin m functions as a major regulator of 812 Calvin cycle enzymes during photosynthesis _in vivo_ . Plant J **84** : 900–913 813 **Pascual MB, Mata-Cabana A, Florencio FJ, Lindahl M, Cejudo FJ** (2011) A comparative 814 analysis of the NADPH thioredoxin reductase C-2-Cys peroxiredoxin system from plants 815 and cyanobacteria. Plant Physiol **155** : 1806–1816 816 **Pérez-Pérez ME, Martín-Figueroa E, Florencio FJ** (2009) Photosynthetic regulation of the 817 cyanobacterium _Synechocystis_ sp. PCC 6803 thioredoxin system and functional analysis 818 of TrxB (Trx x) and TrxQ (Trx y) thioredoxins. Mol Plant **2** : 270–283 819 **Pérez-Ruiz JM, Spínola MC, Kirchsteiger K, Moreno J, Sahrawy M, Cejudo FJ** (2006) Rice 820 NTRC is a high-efficiency redox system for chloroplast protection against oxidative 821 damage. Plant Cell **18** : 2356–2368 822 **Pla A, Lopez-Gorge J** (1981) Thioredoxin/fructose-1,6-bisphosphatase affinity in the enzyme 823 activation by the ferredoxin-thioredoxin system. Biochim Biophys Acta BBA - Bioenerg 824 **636** : 113–118 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 825 **Reyes JC, Crespo JL, Garcia-Dominguez M, Florencio FJ** (1995) Electron transport controls 826 glutamine synthetase activity in the facultative heterotrophic cyanobacterium 827 _Synechocystis_ sp. PCC 6803. Plant Physiol **109** : 899–905 828 **Rippka R** (1988) Isolation and purification of cyanobacteria. Methods Enzymol **167** : 3–27 829 **Rodriguez-Suarez RJ, Mora-Garcı́a S, Wolosiuk RA** (1997) Characterization of cysteine 830 residues involved in the reductive activation and the structural stability of rapeseed 831 ( _Brassica napus_ ) chloroplast Fructose-1,6-bisphosphatase. Biochem Biophys Res 832 Commun **232** : 388–393 833 **Sánchez-Riego AM, López-Maury L, Florencio FJ** (2013) Glutaredoxins are essential for stress 834 adaptation in the cyanobacterium _Synechocystis_ sp. PCC 6803. Front Plant Sci **4** : 428 835 **Schürmann P, Buchanan BB** (2008) The ferredoxin/thioredoxin system of oxygenic 836 photosynthesis. Antioxid Redox Signal **10** : 1235–1274 837 **Serrato AJ, Pérez-Ruiz JM, Spínola MC, Cejudo FJ** (2004) A novel NADPH thioredoxin 838 reductase, localized in the chloroplast, which deficiency causes hypersensitivity to abiotic 839 stress in _Arabidopsis thaliana_ . J Biol Chem **279** : 43821–43827 840 **Soulié J-M, Buc J, Riviére M, Ricard J** (1985) Equilibrium binding of thioredoxin FB to 841 chloroplastic fructose bisphosphatase. Eur J Biochem **152** : 565–568 842 **Spät P, Krauspe V, Hess WR, Maček B, Nalpas N** (2023) Deep proteogenomics of a 843 photosynthetic cyanobacterium. J Proteome Res **22** : 1969–1983 844 **Tanaka K, Kondo A, Hasunuma T** (2026) Systematic determination of disulfide bond reduction 845 potentials reveals a nonequilibrium redox hierarchy in cyanobacteria. Proc Natl Acad Sci. 846 doi: 10.1073/pnas.2600150123 847 **Vieira Dos Santos C, Laugier E, Tarrago L, Massot V, Issakidis-Bourguet E, Rouhier N,** 848 **Rey P** (2007) Specificity of thioredoxins and glutaredoxins as electron donors to two 849 distinct classes of Arabidopsis plastidial methionine sulfoxide reductases B. FEBS Lett 850 **581** : 4371–4376 851 **Wang J, Huang X, Ge H, Wang Y, Chen W, Zheng L, Huang C, Yang H, Li L, Sui N, et al** 852 (2022) The quantitative proteome atlas of a model cyanobacterium. J Genet Genomics 853 **49** : 96–108 854 **Wang P, Liu J, Liu B, Feng D, Da Q, Shu S, Su J, Zhang Y, Wang J, Wang H-B** (2013) 855 Evidence for a role of chloroplastic m-type thioredoxins in the biogenesis of photosystem 856 II in _Arabidopsis_ . Plant Physiol **163** : 1710–1728 857 **Wangensteen OS, Chueca A, Hirasawa M, Sahrawy M, Knaff DB, López Gorgé J** (2001) 858 Binding features of chloroplast fructose-1,6-bisphosphatase-thioredoxin interaction. 859 Biochim Biophys Acta BBA - Protein Struct Mol Enzymol **1547** : 156–166 860 **Wedel N, Soll J, Paap BK** (1997) CP12 provides a new mode of light regulation of Calvin cycle 861 activity in higher plants. Proc Natl Acad Sci U S A **94** : 10479–10484 862 **Wolosiuk RA, Crawford NA, Yee BC, Buchanan BB** (1979) Isolation of three thioredoxins 863 from spinach leaves. J Biol Chem **254** : 1627–1632 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 864 **Yokochi Y, Sugiura K, Takemura K, Yoshida K, Hara S, Wakabayashi K, Kitao A, Hisabori** 865 **T** (2019) Impact of key residues within chloroplast thioredoxin-f on recognition for 866 reduction and oxidation of target proteins. J Biol Chem **294** : 17437–17450 867 **Yoshida K, Hara S, Hisabori T** (2015) Thioredoxin Selectivity for Thiol-based Redox 868 Regulation of Target Proteins in Chloroplasts. J Biol Chem **290** : 14278–14288 869 **Zaffagnini M, Fermani S, Marchand CH, Costa A, Sparla F, Rouhier N, Geigenberger P,** 870 **Lemaire SD, Trost P** (2018) Redox homeostasis in photosynthetic organisms: Novel and 871 established thiol-based molecular mechanisms. Antioxid Redox Signal **31:3** : 155–210 872 873 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 874 881 strains at 0h and 48 h in the presence of inducer. **(e)** PCR analysis of _trxA_ segregation in the 882 indicated strains. M, molecular size marker. Strains carrying the disrupted _trxA_ gene are named 883 without the “P” suffix. 884 Alt text: Multi-panel figure showing schematic diagrams of the _Synechocystis_ strains 885 overexpressing thioredoxins TrxA, TrxB, TrxQ and TrxC, agarose gels of PCR products verifying 886 the strains and _trx_ A segregation, a western blot detecting TrxA, TrxB and TrxQ, and photographs 887 of liquid cultures of the different strains at 0 and 48 h after induction. 888 889 890 891 892 893 894 895 896 897 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 898 899 908 Alt text: Molecular surface models of _Synechocystis_ thioredoxins (TrxA, TrxB, TrxQ and TrxC) 909 colored by electrostatic potential from blue (positive) to red (negative), each shown from two 910 opposite faces, with all four sharing the same fold but differing in surface charge distribution. A 911 ribbon diagram on the right labels the helices α1–α4, strands β1–β4 and the N- and C-terminus. 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 927 930 active site and the regions derived from TrxA and TrxB are indicated. **(b)** Sequence alignment of 931 TrxA, TrxB and the corresponding chimeric Trxs, showing the positions at which the TrxA and 932 TrxB sequences were exchanged. Residue numbering and the active-site region are indicated. **(c)** 933 Genetic strategy for the expression of the chimeric proteins under the control of the arsenite 934 inducible arsB promoter in the WT and SR genetic backgrounds. **(d)** PCR analysis of the parental 935 strain and the corresponding derivatives carrying a disrupted _trxA_ gene. Strains retaining the 936 endogenous _trxA_ gene are designated with the “P” suffix. M, molecular size marker. 937 Alt text: A figure comprising several panel schematic diagrams of the _Synechocystis_ chimeric 938 strains, indicating the regions derived from each protein and the location of the active site, an 939 amino acid sequence alignment highlighting the exchange sites, a diagram of the genetic strategy 940 for expressing the chimeras using the arsB promoter and agarose gels showing the PCR products 941 confirming the segregation of the strains. 942 943 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 944 945 955 absorption spectra after 48 h with asterisks marking the carotenoid, phycobiliprotein and 956 chlorophyll peaks. 957 958 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 959 960 961 **Figure 5** . **Structural and functional characterization of TrxA/TrxB chimeric thioredoxins.** 962 **(a)** Electrostatic surface representation of TrxA/TrxB chimeric Trxs predicted with AlphaFold, 963 shown alongside a ribbon model highlighting secondary structural elements (α-helices and β 964 strands), as well as N- and C-terminus. Electrostatic potentials were calculated using the APBS 965 algorithm and are shown from electropositive (blue) to electronegative (red). The smaller 966 representations show the opposite face of each Trx. **(b)** Structural representation of the tetrameric 967 FBP/SBPase, showing the four subunits and the flexible loop containing the regulatory cysteines 968 Cys75 and Cys84. **(c)** Electrostatic surface representation of FBP/SBPase showing the region 969 surrounding the regulatory cysteines Cys75 and Cys84, shown in two orientations (top view and 970 45º rotation). Electrostatic potentials were calculated using the APBS algorithm and are shown 971 from electropositive (blue) to electronegative (red). **(d)** AlphaFold3-predicted model of the 972 interaction between TrxA and dimeric FBP/SBPase, showing the proximity of the nucleophilic 973 cysteine of TrxA to Cys84 of FBP/SBPase. **(e)** Trx-dependent activation of recombinant 974 FBP/SBPase. Fructose-1,6-bisphosphatase activity was measured using untreated FBP/SBPase 975 (−DTT) or after incubation with DTT alone (No Trx) or with DTT in the presence of TrxA, TrxB 976 and the indicated TrxA/TrxB chimeric Trxs. Data points represent individual measurements, and 977 error bars show the mean ± SD ( _n_ = 6). Asterisks represent significant differences relative to the 978 No Trx control, as determined by one-way ANOVA followed by Dunnett’s multiple-comparisons 979 test: ***P < 0.001; **P < 0.01; *P < 0.05; not significant, ≥0.05. [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764636; this version posted November 9, 2026. The copyright holder for this](https://doi.org/10.64898/2026.11.05.764636) preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in [perpetuity. It is made available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 980 Alt text: Molecular surface models of the TrxA/TrxB chimeric thioredoxins colored by 981 electrostatic potential (blue positive, red negative) with ribbon diagrams, a structural model of the 982 tetrameric FBP/SBPase highlighting the flexible loop with regulatory Cys75 and Cys84, its 983 surface electrostatic potential around these cysteines in two orientations, and a predicted model 984 of TrxA bound to dimeric FBP/SBPase. A dot plot that shows FBPase activity without DTT, with 985 DTT alone, or with DTT plus TrxA, TrxB or each chimera, with asterisks marking significant 986 differences from the DTT-only control. 987 988 989 992