Skip to main content
Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

The Nβ motif of NaTrxh directs secretion as an endoplasmic reticulum transit peptide and variations might result in different cellular targeting

  • Andre Zaragoza-Gómez,

    Roles Data curation, Formal analysis, Investigation, Validation, Visualization, Writing – review & editing

    Affiliations Departamento de Biología Comparada, Facultad de Ciencias, Universidad Nacional Autónoma de México, Ciudad Universitaria, UNAM, Ciudad de Mexico, México, Posgrado en Ciencias Biológicas, Universidad Nacional Autónoma de México, Ciudad de Mexico, México

  • Emilio García-Caffarel,

    Roles Data curation, Investigation, Validation, Writing – review & editing

    Affiliation Departamento de Biología Comparada, Facultad de Ciencias, Universidad Nacional Autónoma de México, Ciudad Universitaria, UNAM, Ciudad de Mexico, México

  • Yuridia Cruz-Zamora,

    Roles Conceptualization, Methodology, Resources

    Affiliation Departamento de Bioquímica, Facultad de Química, Universidad Nacional Autónoma de México, Ciudad Universitaria, UNAM, Ciudad de Mexico, México

  • James González,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Writing – review & editing

    Affiliation Departamento de Biología Celular, Facultad de Ciencias, Universidad Nacional Autónoma de México, Ciudad Universitaria, UNAM, Ciudad de Mexico, México

  • Víctor Hugo Anaya-Muñoz,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Writing – review & editing

    Affiliation Escuela Nacional Estudios Superiores unidad Morelia, Universidad Nacional Autónoma de México, Campus Morelia, Morelia, Michoacán, México

  • Felipe Cruz-García,

    Roles Conceptualization, Investigation, Resources, Supervision, Writing – review & editing

    Affiliation Departamento de Bioquímica, Facultad de Química, Universidad Nacional Autónoma de México, Ciudad Universitaria, UNAM, Ciudad de Mexico, México

  • Javier Andrés Juárez-Díaz

    Roles Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Visualization, Writing – original draft, Writing – review & editing

    Affiliation Departamento de Biología Comparada, Facultad de Ciencias, Universidad Nacional Autónoma de México, Ciudad Universitaria, UNAM, Ciudad de Mexico, México


Soluble secretory proteins with a signal peptide reach the extracellular space through the endoplasmic reticulum-Golgi conventional pathway. During translation, the signal peptide is recognised by the signal recognition particle and results in a co-translational translocation to the endoplasmic reticulum to continue the secretory pathway. However, soluble secretory proteins lacking a signal peptide are also abundant, and several unconventional (endoplasmic reticulum/Golgi independent) pathways have been proposed and some demonstrated. This work describes new features of the secretion signal called Nβ, originally identified in NaTrxh, a plant extracellular thioredoxin, that does not possess an orthodox signal peptide. We provide evidence that other proteins, including thioredoxins type h, with similar sequences are also signal peptide-lacking secretory proteins. To be a secretion signal, positions 5, 8 and 9 must contain neutral residues in plant proteins–a negative residue in position 8 is suggested in animal proteins–to maintain the Nβ motif negatively charged and a hydrophilic profile. Moreover, our results suggest that the NaTrxh translocation to the endoplasmic reticulum occurs as a post-translational event. Finally, the Nβ motif sequence at the N- or C-terminus could be a feature that may help to predict protein localisation, mainly in plant and animal proteins.


Protein secretion is an important cellular mechanism essential for many processes in all organisms, such as cell division and proper response to biotic and abiotic stresses [1, 2]. In plant cells, for instance, the cell wall contains components, such as glycoproteins and enzymes, some of which are integrated into the plasma membrane (PM) and others are secreted towards the extracellular space to form part of the cell wall itself [35]. These components predominantly reach the PM via the vesicular trafficking endomembrane system.

It is thought that both membrane-integral components and secreted ones are localised via the well-characterised conventional protein secretion route. This pathway involves proteins that contain a signal peptide (SP), mainly characterised by its N-terminal position and its hydrophobic profile [1, 6, 7]. This is recognised by the signal recognition particle (SRP)–a conserved ribonucleoprotein–[1, 8, 9]. When an SP-containing protein is being translated, the SRP recognizes the SP and transports the mRNA-ribosome-nascent protein complex towards the endoplasmic reticulum (ER), where the SP is cleaved, and translation fully occurs [8, 10]. Then, the protein is transported within vesicles to the Golgi apparatus, from which vesicles are released towards the PM. Once the vesicles fuse to the PM, the secretory proteins are released to the extracellular space whilst membrane-integral components remain as part of new PM material [1, 5, 11].

The conventional secretory pathway allows for important post-translational modifications required for specific protein functions and/or solubilisation once in the extracellular space. It is in the ER and mainly in the Golgi apparatus where protein glycosylation and oligosaccharide formation occur, which in plant cells is essential for cell wall deposition [3, 5, 10].

Reports regarding SP-lacking or leader-less secreted proteins have been increasing for a long time. This means the occurrence of secretory pathways that are ER/Golgi independent, referred to as the unconventional protein secretory routes (either ER, Golgi, or ER-Golgi independent are considered) [12, 13]. All these ER/Golgi independent pathways–some hypothesised and others experimentally verified–direct soluble leader-less proteins towards the extracellular space and have been mainly studied in mammalian and yeast cells, although in plants, they have gained increasing relevance since the list of SP-lacking secreted plant proteins has been growing [2, 1214].

One such SP-lacking secretory protein is the thioredoxin (Trx) NaTrxh, identified in Nicotiana alata, that localises to the extracellular matrix of stylar transmitting tissue [15]; it is clustered within subgroup 2 of type h Trxs (Trxh-S2), which share an N-terminal extension–some also possess a C-terminal one–with sequences which are not fully conserved [15, 16]. In vitro NaTrxh interacts with and reduces the S-RNase protein [15, 17, 18], which is a secretory protein of the stylar transmitting tissue and is the female S-determinant of the S-RNase-based self-incompatibility (SI) system (rev. within [1921]). S-RNase reduction by NaTrxh results in a seven-fold increase in ribonuclease activity that is essential for self-pollen rejection in Nicotiana [18].

Different algorithms used to predict cellular localisation produce contradictory results regarding NaTrxh cellular localisation [15]. The first 16 amino acid residues of NaTrxh (Met-1 to Ala-16, referred to as the Nα motif; Fig 1A) do not participate in its secretion [17], as predicted by hidden Markov models [15, 22]. Instead, the sequence between Ala-17 and Pro-27 –denominated as the Nβ motif (Fig 1A)–directs NaTrxh secretion, despite not being at the very N-terminus and possessing a predominantly hydrophilic profile [17]. This suggested that SRP should not be able to recognise the Nβ motif and therefore, NaTrxh secretion would follow an unconventional secretion pathway. However, when immunolocalised in N. alata styles, NaTrxh was associated with membrane systems, including the ER, the Golgi apparatus and presumably secretion vesicles [17].

Fig 1. The Nβ motif sequence is found in proteins from all taxa.

(A) Schematic representation of the primary structure of NaTrxh, where the Nα motif (orange) comprises from Met-1 (M1) to Ala-16 (A16); the Nβ motif (red), which sequence is shown (Ala-17 to Pro-27) [colour code refers to the charge of each amino acid (green: neutral; black: hydrophobic; red: negative)]; the active site (grey) contains the typical WCGPC motif; the C-terminal extension is shown in cyan and comprises from Glu-136 (E136) to Gln-152 (Q152). (B) Distribution of the eukaryotic proteins (304 sequences) that contain the Nβ motif or a similar one. (C) Consensus sequence (featured as logos) of all the Nβ motifs found in eukaryotic proteins. Colour code as in (A).

The fact that NaTrxh secretion is due to the Nβ motif was proven in past literature [17] by fusing its short sequence to the N-terminal of the green fluorescent protein (GFP)–creating an Nβ-GFP fusion protein–and tested by transient expression assays in onion epidermal cells, which as a model system, have been useful in studying cell processes including secretion [2325]. NaTrxh-GFP and Nβ-GFP fusion proteins were secreted using the conventional secretion elements: both passed through the ER and secretion was brefeldin-A sensitive–i.e., proceeded via the Golgi apparatus–[17].

While soluble secretory SP-containing proteins are considered to follow the conventional secretory pathway, those that do not contain an orthodox SP are suggested to be secreted by an ER/Golgi independent one. However, through an analysis of the Nβ motif from NaTrxh, we hypothesised that NaTrxh follows a “semi-conventional” secretion pathway in which the Nβ motif functions as a transit peptide towards the ER to continue its secretory pathway to reach the extracellular space. This suggests a post-translational translocation instead of a co-translational one, as occurs with a typical SP-containing protein recognised by the SRP when translated. In addition, our data suggest that the Nβ motif might help to predict whether a protein lacking an SP might be secreted, specifically in plant proteins and presumably in animal ones. Mainly, among Trxh-S2 with similar N-extensions, two major groups were found, one with a similar Nβ motif that could be predicted to act as secretory proteins and another that may be a mitochondrial one, suggesting that few changes within the Nβ motif might direct proteins towards different cellular compartments, contributing to the vast physiological roles in which Trxs are being involved.


The eleven-residue long motif, Nβ, is present in transport and cell traffic proteins

The Nβ motif sequence (Fig 1A), identified as essential for NaTrxh secretion in plant cells [17], was used for BLASTP analysis. The output raised 383 protein sequences containing identical or similar sequences, 304 of them being eukaryotic proteins (S1 Table), all registered at the National Center for Biotechnology Information database (NCBI; Within the eukaryotic protein sequences containing an Nβ motif (or similar), 266 are from animals, 22 from fungi and 16 from plants (Fig 1B and S1 Table). All 304 sequences were used to generate the Nβ motif-like eukaryotic consensus sequence shown in Fig 1C.

The small size of the Nβ motif itself, just eleven amino acid residues long, may lead to random identification of similar but functionally unrelated sequences that would not represent any relationship to a secretory signal as hypothesised. However, 191 of the animal proteins detected are associated with membrane traffic, from which 98 are involved in endocytosis and exocytosis (most of the sequences correspond to orthologs of Unc119 or C-Jun); among the plant proteins found, four are identified as chaperones belonging to the Nicotiana genus and annotated as Trxh-S2, where NaTrxh is grouped (S1 Table) [15]. The chance that these results were due to random factors is diminished because most proteins containing the Nβ motif of NaTrxh–or a similar one–are related to membrane-traffic mobility in cellular processes, suggesting an Nβ presence-function relationship.

Amino acid positions in the Nβ motif required for secretion in plant cells

The consensus sequence of the Nβ motif comprising all sequences found in the different eukaryotic proteins (Fig 1C) showed that positions 1 and 3 are the least conserved and include two hydrophobic amino acid residues (Fig 1A and 1C). To determine if these amino acids were functionally relevant and to define the minimum size of the functional Nβ motif, we generated two types of protein variants (Fig 2A). The first one used the NaTrxh-GFP fusion protein, to which deletions of different sizes at its N-terminus were made (single residues at either 3 or 6 positions from the Nα motif): NaTrxhΔNα(+3)-GFP and NaTrxhΔNα(+6)-GFP; the second type was generated by deleting 3 or 6 amino acid residues from start of the Nβ-GFP sequence: Nβ(-3)-GFP and Nβ(-6)-GFP. All these protein variants were transiently expressed in onion epidermal cells.

Fig 2. The first three Nβ motif residues do not contribute to its secretion signal function.

(A) Schematic representation of the two types of constructs generated to assess the size of the functional Nβ motif. In one, deletions were made from NaTrxh fused to GFP (left) and the other from the Nβ motif directly fused to GFP (right). The Nα motif is displayed in orange, the Nβ in red, the core of NaTrxh in blue, the C-motif in cyan and GFP in green. Light grey represents the deleted residues in both types of constructs; for the Nβ motif directly fused to GFP (right), in dark grey are indicated the initial methionine and the linker sequence between the insert and the GFP sequence. (B) Transient expression assays in onion epidermal cells and observed under confocal microscopy for GFP fluorescence to determine the localisation of NaTrxh-GFP, Nβ-GFP, NaTrxhΔNα(+3)-GFP, Nβ(-3)-GFP, NaTrxhΔNα(+6)-GFP, Nβ(-6)-GFP. Left panels: GFP fluorescence; right panels: merged image of bright field plus GFP fluorescence. Cells were pre-incubated with 1 M NaCl for plasmolysis. Cyt: cytoplasm; CW: cell wall. Scale bar: 50 μm.

When the Nβ deletions were transiently expressed in onion epidermal cells, we found that the first three positions are not required for the Nβ motif to direct protein secretion because while both NaTrxhΔNα(+3)-GFP and Nβ(-3)-GFP were localised in the extracellular space, NaTrxhΔNα(+6)-GFP and Nβ(-6)-GFP localised to the cytoplasm (Fig 2B). NaTrxh-GFP and Nβ-GFP, used as controls, both localised to the extracellular space (Fig 2B) as previously reported by [17]. These results confirm that the Nβ motif-guided secretion is independent of the core of NaTrxh, as previously reported [17].

As indicated above, the consensus sequence of the Nβ motif shown in Fig 1C is formed from all the eukaryotic sequences found in the set of predicted secretory and cytoplasmic proteins obtained from the BLASTP search (S1 Table). To discriminate between these groups, we used the UniProtKB database [26], where the information regarding cellular localisation is based on a score that varies from low information (score 1) to complete and experimental evidence (score 5). We clustered these proteins into four categories (from A to D) based on UniProtKB scores and their cellular localisation (Fig 3A): A, secreted proteins with score = 5 (8 sequences); B, secreted proteins with score over 1 and under 3 (78 sequences); C, cytoplasmic proteins with score = 5 (8 sequences); D, proteins with no information regarding their cellular localization (210 sequences).

Fig 3. Positions 5, 8 and 9 within the Nβ motif differ between secretory and cytosolic proteins.

(A) Distribution of the protein sequences among the categories generated based on the UniProtKB database regarding cellular localisation: A category (2.63%); B category (25.66%); C category (2.63%); D category (69.08%). (B) Logos of the sequences found in A category (upper) and C category proteins (down). (C) Transient expression assays in onion epidermal cells (as in Fig 2) of the different Nβ variants fused to GFP. Cyt: cytoplasm; CW: cell wall. Scale bar: 50 μm.

Although there is solid evidence regarding NaTrxh secretion [15, 17], NaTrxh was grouped within the D category with other Trxh-S2 that have the same N-terminal extension sequence as NaTrxh. Based on the UniProtKB available data, this finding means the A category is likely underestimated. The D category appears to be overestimated–it is indeed the largest category (Fig 3A)–. The same may occur with the B category, for which finding an Nβ motif might predict protein secretion, increasing its score (as with the A category) as discussed below. The results shown in Fig 3A and further analysis consider NaTrxh to reside within the A category.

To assess which amino acid residues might be essential for the Nβ motif to work as an SP-like sequence, we compared the two distinctive consensus sequences generated by contrasting localization-category proteins (featured as logos; Fig 3B): (1) generated from A category proteins; and (2) from C category proteins. Comparing these two consensus sequences revealed that positions 5, 8 and 9 vary from one group to another. While the A category proteins contain a serine residue in positions 5 and 9, some C category ones contain Gly-5 and Ala-9. Thus, to test whether these changes are relevant for the Nβ motif to direct secretion, we generated the Nβ(S5G) and the Nβ(S9A) variants–both replacing an A category residue with a C category residue–using the Nβ-GFP sequence as a template, which contains the NaTrxh Nβ sequence (Fig 1A). The results from transient expression in onion epidermal cells showed that while Nβ(S5G)-GFP was secreted, Nβ(S9A)-GFP was localised within the cytoplasm (Fig 3C). This provides evidence that, while variations in position 5 of the Nβ motif appear not to be relevant, the motif cannot be predicted to act as a secretion signal sequence if an Ala-9 –or probably any hydrophobic residue at this position–is present. This suggests that position 9 requires a serine residue (or a neutral one) to maintain its role as a secretion signal.

Regarding position 8, while most of the category A proteins–Unc-119 orthologs–contain a glutamate residue–NaTrxh being an exception since it contains a serine, which is considered in the consensus sequence of Fig 3B–, the C category proteins mainly contain an aspartate residue here and only a few have glutamate at this position (Fig 3B). Therefore, we generated the following variants fused to GFP: (1) Nβ(S8D) to mimic the sequence found in proteins within C category; (2) Nβ(S8E), which is the predominant form in the A category and could be expected to work as a signal peptide-like sequence; and (3) Nβ(S8A) to assess the relevance of a negatively charged residue at this position.

Results (Fig 3C) indicated that the protein variant Nβ(S8A)-GFP is localised at the extracellular space (Fig 3C), indicating that the change from a neutral to a hydrophobic residue did not affect the role of the Nβ motif as a secretion signal at this position. However, when Ser-8 was replaced by a negative residue as found in the Asp and Glu residues, both Nβ(S8D)- and Nβ(S8E)-GFP were localised within the cell (Fig 3C). These data provide different possible scenarios. First, Ser-8 appears essential for the Nβ motif to direct protein secretion. However, the amino acid predominantly found at this position is Glu in secretory proteins (A category) (Fig 3B). Notably, as Glu-8 was found in Unc-119 orthologs, which has a major role in protein traffic–of acylated proteins–to subcellular destinations [27], a glutamate at this position is suggested to be crucial for the Nβ motif to work as a secretory signal or as a traffic peptide within this taxon and, probably, if there is Asp-8, the protein would not be secreted nor related to a traffic role since this residue is common among the cytoplasmic proteins (C category) (Fig 3B). This hypothesis needs to be further assessed. Both amino acids are negative, but maybe in this case, at this position, the small size difference between them is enough to be recognized or not as a secretory/traffic signal especially in animal cells. In contrast, plant proteins appear to require a serine at position 8. This hypothesis is reinforced by the fact that NaTrxh was grouped within the D category (proteins with no information regarding its cellular localization) according to the available UniProtKB data despite the clear evidence supporting its secretion [15, 17]. Some proteins within the D category are annotated as Trxh-S2, meaning that they have N-terminal extensions and, notably, they all possess the Ser-8 residue (Fig 4), reinforcing the hypothesis that the D category–and probably also B–might be overestimated, and A category underestimated.

Fig 4. Analysis of different N-extensions of Trxh-S2.

Multiple alignment analysis of different Trxh-S2 considering E. coli Trx1 (EcTrx1; AAA24693.1; blue sequence) as an outgroup. All Trxh-S2 are longer than E. coli Trx1, predominantly due to the N-terminal extensions present in all Trxh-S2 sequences and the C-terminal extensions (cyan) present in some of them. Notably, NaTrxh (AAY42864.1; bolded) possesses the longest C-terminal extension. The extensions of those Trxs that have been associated to the plasma membrane are labelled in turquoise [AtTrxh9 (AAS49091.1) and AtTrxh8 (AAG52561.1) from A. thaliana] and that to mitochondrial in purple [PtTrxh2 (AAL90749.1) from Populus hybrid]. Dark grey areas: Nα motifs or equivalent. Yellow: Nβ sequences; purple: Nβ variations that might result in a mitochondrial transit peptide; black: shortened Nβ sequences; green highlighted: the WCGPC active site of Trxs. The relevant identified differences are highlighted in blue. The other sequences used and their respective GenBank accession codes are: A. thaliana: AtTrxh7 (AAD39316.1) and AtTrxh2 (Q38879.2; in bold because of its proven mitochondrial localisation); Eutrema salsugineum: EsTrxh2 (XP_024014606.1); Raphanus sativus: RsTrxh2 (XP018446242.1); Brassica napus: BnTrxh2 (XP_013647760.2); B. rapa: BrTrxh2 (009125122.1); Capsella rubella: CrTrxh2 (XP006284753.1); Camelina sativa: CsTrxh2 (XP_019089899.1); Arachis ipaensis: AiTrxh2 (016194214.1); Ipomea batatas: IbTrxh2 (AAQ23133.1); Actinidia rufa: ArTrx2 (GFY87055.1); Petunia hybrida: PhTrxh (UYX45859.1); Nicotiana attenuata: NatTrxh2L (XP_019242743.1); N. tomentosiformis: NtoTrxh2 (XP_009602293.1); N. sylvestris: NsTrxh2 (XP_009795987.1); N. tabacum: NtTrxh2L (XP016508831.1); Datura stramonium: DsTrx (MCD9559198.1); Solanum verrucosum: SvTrxh2 (XP049346357.1); S. pennellii: SpTrxh2 (XP_015063846.1); Senna tora: StTrxh2 (KAF7819668.1); Prosopis alba: PaTrxh2 (XP_028806431.1); Rosa chinensis: RcTrxh2 (XP_024180631.1); Juglans hybrid: JhTrxh2L (XP_041003776.1); Carya illinoinensis: CiTrxh2L (XP_042969725.1). Upper panel: first part of the multiple alignment, where N-extensions are present; lower panel: the rest of the aligned primary structures.

The Nβ motif might be crucial for targeting of Trxs to different cellular locations

In plants, at least eight types of Trxs have been identified, and there is a relationship between cellular localisation and Trx function (reviewed in [28]). For example, Trx types f, m, x, and y are chloroplast proteins, type o Trxs localise to the mitochondria, and those of type s are associated with the ER [2934]. Type h Trxs (Trxh) form the largest and most heterogenous group and have been considered as cytosolic proteins [16, 35]. However, there is evidence of some exceptions, like PtTrxh2 from Populus tremula and Trxh-2 from A. thaliana (AtTrxh2), which localise to the mitochondria [36, 37] or NaTrxh, which is located at the extracellular matrix of the stylar transmitting tissue in N. alata [15]. In addition, some other Trxh can move from cell to cell via plasmodesmata, suggesting an intercellular messenger role as found for some animal Trxs [3740]. Coincidentally, all the Trxh with a non-cytosolic localisation are clustered within subgroup 2, in which these oxidoreductases are characterised by possessing an N-terminal extension, some also a C-terminal one, whose sequences are not conserved [16].

From the BLASTP results (see above), we identified four Trxh-S2 from Nicotiana (S1 Table), namely: NaTrxh (from N. alata), NtTrxh2L (N. tabacum), NsTrxh2 (N. sylvestris) and NtoTrxh2 (N. tomentosiformis). All four of them possess identical Nβ sequences (Fig 4).

To include more Trxh-S2 sequences in the analysis, a second BLASTP search was performed using the whole NaTrxh sequence, from which the AtTrxh2 sequence was retrieved (Fig 4), which corresponds to the mitochondrial Trxh-2 from A. thaliana. This sequence was used for an additional BLASTP. Only Trxh-S2 were considered for a multiple alignment analysis using the E. coli Trx1 (EcTrx1) as an outgroup (Fig 4). We also included the mitochondrial PtTrxh2, as well as AtTrxh9 and AtTrxh8 because they are known to be membrane-associated proteins [41].

As shown in Fig 4, all sequences contained the conserved redox-active site WCGPC, some contain a C-terminal extension–the longest one belonging to NaTrxh–, and all contain an evident N-terminal extension of similar length when compared to the E. coli Trx. Most of the sequences retrieved from the BLASTP search contain similar Nβ sequences but also present differences. All those containing Ser-5 and Ser-9 (positions refer to the original Nβ sequence from Fig 1A) are clustered together (Fig 4, yellow cluster) and, according to our data, might be predicted as secretory proteins. The AiTrxh2 (a Trxh-S2 from Arachis ipaensis) was not included in this same cluster because it contains Ala-9. Due to this factor, it can be predicted that AiTrxh2 has a cytoplasmic localisation in concordance with the location of Nβ(S9A)-GFP (Fig 3C). The light-green cluster of Fig 4 contains Trxh-S2 proteins with shorter Nβ sequences. This group exhibits significant residue variations, namely: S5A and S7A, E10D and P11D (positions refer to the original Nβ sequence). This suggests a cellular localisation other than the extracellular space.

The Nβ(S8E)-GFP variant is accumulated in the cytosol when expressed in onion epidermal cells (Fig 3C). A group of sequences was found to contain Glu at this position and, except for one, they have Ala-5 instead of Ser-5; these all were grouped with the mitochondrial localised AtTrxh2 (Fig 4, purple cluster), raising the possibility that these differences result in formation of a transit peptide to direct the proteins to the mitochondria.

Our data suggest that this short sequence might be an evolutionary source for different cellular localisations. Different combinations of amino acid residues in the N-terminal extension could have evolved to target Trxs to different cellular compartments and organelles, including the apoplast, where they would reduce different target proteins, which amplify the role of these oxidoreductases in several physiological plant processes. This, in turn, would provide an additional explanation of the great diversity of these proteins in plants, particularly among type h Trxs.

The Nβ motif of secretory proteins is predominantly located towards the N-terminus and a structural trait appears to make it function as a secretion signal

The hallmark feature of Trxh-S2 proteins is that they possess N-terminal extensions whose sequences are quite variable [16]. The Nβ motif is located between positions 17 and 27 in NaTrxh (Fig 1A), and when it is fused to the GFP N-terminus (Nβ-GFP fusion protein), it leads to GFP secretion from plant cells (Fig 2B; [17]). These data suggest that the Nβ motif must be located towards the N-terminal to exert an SP-like role, similar to a hydrophobic typical SP. Therefore, we classified the output sequences considered in Fig 1B–and S1 Table–according to the region in which the Nβ motif is located within each protein. For this aim, each sequence was divided into four equal parts (P1: first 25% of the primary structure towards the N-terminus; P2 and P3: second and third quarter, respectively; P4: the final quarter towards the C-terminus) and was classified according to which quarter the Nβ motif is located in. As shown in Fig 5A, 101 proteins contained the Nβ motif in P1, 65 in P2, 45 in P3 and 88 in P4, resulting in a non-random distribution (χ2 = 27.974, df = 3, p<0.0001), indicating that indeed, most of the proteins contain this motif towards the N-terminus.

Fig 5. Nβ motif is mostly located towards the N-terminus of extracellular proteins.

(A) Distribution of the proteins according to the localisation of the Nβ motif. The sequences were divided into four quarters: P1: 25% of the protein towards the N-terminus; P2 and P3: second and third quarters, respectively; P4: 25% of the protein towards the C-terminus. (B) Annotated functions of the proteins that contain the Nβ motif in each position. PI: peptidases and inhibitors; RibB: ribosomal biogenesis; Chap: chaperones; Aut: autophagy; ER: endoplasmic reticulum; MR: membrane recycling; TF: transcription factors; LM: lipid metabolism; End: endocytosis; Ub: ubiquitination; Exo: exocytosis; Lys: lysosome; MT: membrane transport; Exo(synth): exocytosis (synthesis); DR: DNA repair; RB: RNA biogenesis; SO: sarcomere organization; Unk: unkown. (C) Distribution of the cellular localisation based on the categories generated using the UniProtKB database in each position (colour code is the same as in Fig 3A). P1, P2, P3 and P4 labels between (B) and (C) refer to upper (B) and lower (C) pie-charts.

When each position category was related to the function of each protein, as shown in Fig 5B, P1 and P2 were more homogeneous regarding their annotated roles [81.13% are involved in membrane recycling (P1) and 80% in endocytosis (P2)], in contrast with P3 and P4, which lacked any predominant reported role.

While all the A category proteins shown in Fig 3A contain the Nβ motif in P1, none from the category C proteins did. Instead, for category C proteins, the Nβ sequences were distributed in P2, P3 and P4 (Fig 5C). These data together led us to hypothesise that the Nβ must be located towards the N-terminus of the protein to function as an SP-like sequence.

To test this hypothesis, we generated two different constructs (Fig 6A): (1) the chimeric protein NaTrxhΔNαβ-Nβ-GFP, formed from the core of NaTrxh without its N-terminal extension (NaTrxhΔNαβ) followed by the Nβ-GFP fusion sequence to assess a central position of the Nβ motif; and (2) GFP-Nβ, which served to evaluate the role of the Nβ motif when located in the C-terminal position.

Fig 6. Nβ motif recognition is probably due to a structural feature rather than a positional one.

(A) Schematic representation of the constructs generated to assess the position of the Nβ motif as a secretion signal and with the Nβ sequence inverted towards the GFP N-terminus. (B) Transient expression assays in onion epidermal cells of the NaTrxh(ΔNαβ)-Nβ-GFP chimera, the GFP-Nβ and the Nβ(inv)-GFP fusion proteins as in Fig 2. Cyt: cytoplasm; CW: cell wall. Scale bar: 50 μm. (C) Tertiary structure of GFP (PDB 6YLQ) indicating its N- (pink) and C-termini (grey). The figure was generated using PyMOL.

When the NaTrxhΔNαβ-Nβ-GFP chimeric protein was transiently expressed in onion epidermal cells, GFP fluorescence, as predicted, was observed in the cytoplasm (Fig 6B), indicating that the Nβ motif was not identified as a secretion signal within the middle core of the protein, which corresponds to P2 and P3 positions of our previous analysis (Fig 5). Unexpectedly, when the Nβ motif was localised at the C-terminus (GFP-Nβ construct), the GFP signal was detected in the apoplast of the onion epidermal cells (Fig 6B), indicating that the Nβ motif is able to direct secretion whether localised at the N- or the C-terminus. It is noteworthy that, although none of the proteins that contain the Nβ motif towards its C-terminus (P4) are annotated as secretory proteins (A category; Fig 5C), in none of the cases is it present at the very C-terminus; the closest Nβ sequence to the C-terminus end of the cytoplasmic proteins (C category) is located 122 residues far from it (NP_694967.3 accession code, which is 648 amino acids long; S1 Table).

In the case of the GFP-Nβ protein, the Nβ motif is not only at the very C-terminal end of the protein, but it is also likely to be a mobile element that would be free and exposed to the solvent in a similar manner to the N-terminal extension of NaTrxh [42]. As long as the Nβ motif is free and solvent-exposed, no matter if it localises at the N- or C-terminus, factors–yet to identify–interact with it, directing translocation of GFP into the ER. This latter assumption is based on the tertiary structure of GFP (PDB 6YLQ), in which both N- and C-termini are outside the β-barrel and are oriented towards the same direction (Fig 6C). From the BLASTP analysis, XP_002018569.1–510 amino acids long–and PLN81894.1–562 amino acids long–(both GenBank accession codes) contain the Nβ motif at P4 (12 and 15 residues from its C-terminal, respectively) were included in the B category (S1 Table). According to the models predicted by AlphaFold in UniProtKB, the Nβ motif might be mobile and solvent exposed in both cases, suggesting that these proteins might be secreted. As expected, PLN81894.1 is associated with PM and transport activity (S1 Table). These data reinforce the hypothesis that the B category is possibly overestimated.

To assess the relevance of the Nβ motif itself, its charge and/or the free and solvent-exposed structure it needs to function as a signal peptide, an inverted sequence version of the Nβ was fused to the GFP N-terminus (Nβinv-GFP; Fig 6A). When the Nβinv-GFP fusion protein was transiently expressed in onion epidermal cells it localised to the extracellular space (Fig 6B) as the Nβ-GFP, GFP-Nβ and NaTrxh-GFP do. This outcome indicates that the Nβ motif’s overall charge–negative at physiological pH–is essential to lead protein secretion rather than the amino acid position. The negative charge is mainly provided by Glu-4 and Glu-10, both present in A and C category Nβ sequences (Fig 3B). Still, position 9 must contain a serine residue, as shown in Fig 3C, or at least a neutral polar residue to maintain the hydrophilic profile of the Nβ sequence.

The Nβ-directed secretion involves a post-translational translocation to the endoplasmic reticulum

NaTrxh utilises the ER-Golgi elements of the secretion pathway, as directed by the Nβ motif and demonstrated by the Nβ-GFP fusion protein, which also passes through the ER and the Golgi apparatus [17]. This might suggest a co-translational translocation of the protein towards the ER for its secretion following the conventional pathway. When SRP recognizes an SP at the N-terminus of nascent proteins, it translocates the mRNA-ribosome-nascent protein complex to the ER membrane to further continue translation towards the ER lumen [6, 8]. Although the SP sequences are extremely variable in both size and composition, they all share a hydrophobic core region [43] that interacts with the hydrophobic residues of the M domain of SRP54, which is one of the six proteins that form eukaryotic SRP [9, 44]. However, since the Nβ motif is hydrophilic with an overall negative charge, an SRP-dependent ER translocation can be discarded.

The fact that the GFP-Nβ protein was found in the extracellular space (Fig 6B) also discards the possibility of GFP-Nβ translation coupled to the ER translocation by SRP. Therefore, we hypothesised that the GFP-Nβ protein is fully translated in the cytosol and then translocated to the ER to continue via the Golgi apparatus to ultimately reach the apoplast [17]. To test this, the ER retention signal KDEL [4548] was added to the GFP-Nβ [GFP-Nβ(KDEL); Fig 7A] and transiently expressed in onion epidermal cells. As shown in Fig 7B, the GFP-Nβ(KDEL) protein was detected within the cell exhibiting a typical ER-retained protein pattern [23, 46, 49], and similar to that found for NaTrxh-GFP(KDEL) and Nβ-GFP(KDEL) fusion proteins using the same type of expression assays [17]. Thus, our data provide evidence that once the Nβ motif is free and exposed to the solvent in any protein, either at its N- or very C-terminal end, and is at least 8-residue long, is recognised as a transit peptide to the ER to be followed by protein secretion (Fig 8). It is possible that the functional sequence of this signal is X-X-X-E-S/G-G-S-S/E-S-E-P, where the fifth position must be a neutral amino acid residue (at least a serine or a glycine) and position 9 a serine residue. The eighth position needs a serine in the case of plant proteins–especially for Trxh-S2– for a secretory function, but if there is Glu-8 –together with Ala-5– it is possible that the sequence instead functions as a mitochondrial transit peptide, as might be the case for AtTrxh2 [37]. Moreover, in the case of animal systems, there is another possibility: that of a requirement for a glutamate (Glu-8) in order to provide a secretory or a traffic function. Both possibilities are worthy of further investigation.

Fig 7. The GFP-Nβ fusion protein uses the ER for its secretion.

(A) Schematic representation of the GFP-Nβ construct with the ER retention signal KDEL [GFP-Nβ(KDEL)]. (B) Transient expression assay in onion epidermal cells of the GFP-Nβ(KDEL) protein. Nuclei were stained with propidium iodide before observation (magenta fluorescence). Left panels: GFP fluorescence; middle panels: propidium iodide fluorescence; right panels: GFP and nucleus-labelled merged image (upper) and bright field, GFP and nucleus-labelled merged image (lower). Cyt: cytoplasm; CW: cell wall; Nu: nucleus. Scale bar: 50 μm.

Fig 8. The Nβ “semi-conventional” secretion pathway differs from the conventional and unconventional pathways.

In the conventional secretion pathway (CSP), during translation of a protein with a signal peptide (SP), the SP is recognized by SRP, resulting in the translocation of the mRNA-ribosome-nascent protein to the endoplasmic reticulum lumen, where the SP is cleaved, and translation is completed. The unconventional secretion pathways (USP) comprise: Golgi apparatus bypass (1), where the protein is directly secreted from the ER; ER-independent routes, where proteins may follow secretion after Golgi incorporation (2), through incorporation in multivesicular bodies (3), or through alternative direct secretion pathways (4). In the case of Nβ motif-containing proteins, such as NaTrxh, these are translated in the cytosol and, in an SRP-independent manner, are translocated to the ER to then progress to secretion via the CSP elements (ER, Golgi and secretory vesicles). Figure created using Biorender.


The conventional secretion route, in addition to passage through the ER, the Golgi apparatus, and the exit of the proteins from it within vesicles, includes the assistance of co-translational incorporation of the protein–while translated in the ribosome–into the ER by SRP (Fig 8). This translocation is the result of the SRP-SP interaction, in which the SP at the N-terminal end of the protein must be a hydrophobic element [810] with eight residues of minimal length [44]. The Nβ motif is quite the opposite since it is hydrophilic–negatively charged–as previously described [17] and further analysed in this work. This eliminates the possibility of Nβ being recognised by the hydrophobic groove of the M domain of SRP54 –a conserved component of eukaryotic SRP–that is responsible for SP recognition [9, 44]. However, the NaTrxh-GFP, Nβ-GFP ([17] and Fig 2B) and GFP-Nβ fusion proteins reach the extracellular space by passing through the ER (Figs 6B and 7B). This, and particularly the latter, in which the Nβ is at GFP C-terminus, indicates that the protein was fully translated in the cytosol and then incorporated into the ER (Fig 8), indicating that the ER localisation was not due to an SRP-dependent translocation of the nascent protein during its translation, as occurs in the conventional secretion pathway. The Nβ-led secretion, as reported for NaTrxh-GFP and Nβ-GFP, also involves the Golgi apparatus since it is brefeldin-A sensitive [17]. Therefore, the secretory pathway for Nβ-containing proteins differs from both conventional and unconventional secretion pathways. While it is not ER nor Golgi independent–varying from the unconventional proposed pathways–, the entry to the ER-Golgi route also differs from the conventional one, suggesting an additional pathway that would be “semi-conventional” in nature (Fig 8). Likewise, the secretion of GFP-Nβ protein suggests that the Nβ motif is not processed as occurs with most classical SPs [810]. The presence of this motif in mature NaTrxh is crucial for interaction with the S-RNase in order to regulate its ribonuclease activity [18].

The results reported here suggest that the Nβ motif of NaTrxh acts as a transit peptide towards the ER (Figs 6B and 7B). This post-translational ER translocation for secretory proteins has been reported in the case of small-size proteins (peptide hormones), proteins with an SP with a low hydrophobicity and those with an SP located at the C-terminus [10]. An SRP-independent ER translocation has been reported to be achieved by the assistance of chaperones, like Hsc70, to prevent protein misfolding and by calmodulin, which is the responsible for the ER post-translational translocation, where another chaperone of the Hsp70 family–present in the ER lumen–appears to be the assistant for the movement of the targeted protein through the translocation channel towards the ER lumen [10].

Regarding the prevalence of the N-terminal extension of NaTrxh, some proteins localised to different cell compartments are first translated into the cytosol and then translocated towards the target organelle due to the presence of a transit peptide. However, some remain in the cytosol until a signal directs them to the corresponding organelle. A good example of this latter group are nuclear proteins, such as transcription factors, that remain in a latent form in the cytosol and only in response to certain stimuli are they re-located to the nucleus. An example of this is the NF-κB complex, which contains a nuclear transit peptide but remains in the cytosol because its inhibitor, I-κB, is tightly bound and obscures the transit peptide [50, 51]. During an immune or inflammatory response, for instance, I-κB is phosphorylated and degraded, releasing NF-κB and exposing the nuclear localisation sequence, resulting in its translocation towards the nucleus, where it will activate transcription of specific target genes [50, 51]. A similar scenario could occur with NaTrxh or any Nβ-containing protein with the features described here but translocated towards the ER rather than nucleus. This would represent an efficient mechanism to regulate a portion of soluble protein secretion, which would occur when the Nβ is exposed.

NaTrxh is directly involved in Nicotiana’s gametophytic SI system, known as S-RNase-dependent SI [18]. SI is a genetically controlled system that acts as a barrier to self-pollination and involves complex pollen-pistil interactions in which secreted stylar proteins play key roles [1921, 52, 53]. The S-RNase–the female S determinant–and NaTrxh colocalise to the extracellular matrix of the stylar transmitting tissue in N. alata [15]. S-RNases and other proteins involved in this gametophytic SI system contain a conventional SP and are secreted by the stylar cells [54, 55], the exception being NaTrxh, where secretion is due to the presence of its Nβ motif [17].

In either compatible (non-self-pollen) or incompatible (self-pollen) crosses, S-RNase enters the pollen tube cytoplasm [56, 57]. For self-pollen rejection, reduction of a specific disulphide bond in S-RNase by NaTrxh is required in order to increase its ribonuclease activity; transgenic Nicotiana plants expressing a non-functional NaTrxh are unable to reject wild-type self-pollen [18]. These data indicate that secreted stylar NaTrxh must enter the pollen tube cytoplasm, where the electron donor system is available, in order to reduce S-RNase.

Once in the pollen tube cytoplasm, how does NaTrxh, with its Nβ motif, avoid secretion from the pollen tube? How does it stay in the cytosol and not become targeted to the ER? The answer could rely on the stable interaction between S-RNase and NaTrxh, in which both N- and C-termini are involved [15, 17, 18]. Although this interaction has only been observed in vitro, the N-terminal extension of NaTrxh is essential to reduce S-RNase [18]. The NaTrxh-S-RNase interaction results in a protein-protein complex where the Nβ motif is stabilised, as modelled using the crystal structures of both NaTrxh (PDB 6X0B) and SF11-RNase (PDB 1IOO), where stabilising interactions are detected between the NaTrxh N-terminal extension and SF11-RNase [42].

If not completely hidden, the NaTrxh N-terminal extension might become a stable and ordered element [42]. This evidence also explains why proteins, including the NaTrxhΔNαβ-Nβ-GFP chimera, that contain the Nβ in an inner position, although potentially exposed to the solvent, are not secreted, because this motif might form stable secondary structural elements. Additionally, in the case of the GFP-Nβ protein, the GFP C-terminus is likely to be accessible, away from the GFP β-barrel. The Nβ motif would be mobile, disordered and fully exposed to the solvent, which might be the reason it works as an ER translocation and further secretion signal in the GFP-Nβ fusion protein (Figs 6 and 7). This reinforces the idea that the Nβ motif structural features crucial to be a secretory/traffic signal are opposite to those of a classical SP that, besides the hydrophobicity, preferentially adopts an α-helix structure [44, 58].

Trxs are widely distributed proteins, from prokaryotes to eukaryotes, that reduce disulphide bonds of target proteins [59]. The polymorphism of these proteins reflects the different roles they are involved in, which are as wide as the target proteins they reduce [28, 60]. Plant Trxs represent a large and complex system in which 8 classes of them have been identified. The type h thioredoxins form the largest and most heterogenous group and are subdivided into three subgroups; the subgroup 2 Trxh contains extensions towards the N-terminus some also at the C-terminus [16, 36]. Although the Trxh-S2 N- and C-terminal extensions are not fully conserved, some share common features. Those Trxh-S2 proteins that have the most similar sequence to the Nβ of NaTrxh were clustered together (Fig 4), and possibly they all are secreted in a similar manner. However, another group of Trxh-S2 contained some differences. They grouped with a known mitochondrial Trx, AtTrxh2 [37], as shown in Fig 4. This reinforces the idea that this short sequence functions as a transit peptide, but interestingly, select changes to the amino acid sequence–at least two are detected in this work–might influence its target cellular localisation. Therefore, having initially acquired an Nβ sequence, further evolutionary changes to the sequence could lead to different Trx homologs being differentially localised in the cell, where they could reduce new target proteins. For example, whilst having serine residues at positions 5 and 8 results in a secretory motif, Ala-5 and Glu-8 might result in a mitochondrial localisation sequence.

Dissecting both NaTrxh extensions has provided useful information on how they are involved in NaTrxh cellular localisation and its specificity towards its identified target protein in N. alata styles–the S-RNase–. Both E. coli Trx and NaTrxh reduce insulin disulphide bonds as expected from any Trx [15, 61], but E. coli Trx is not able to reduce S-RNase [18]; the major difference between these two Trxs is the occurrence of the N- and C-terminal sequences in NaTrxh. In addition, according to the structural model of the NaTrxh-S-RNase protein complex, the NaTrxh N-terminus assists the correct orientation of the interaction to reduce only one of the four disulphide bonds that the S-RNases typically contain [42]. The differences detected in the Nβ sequences of the Trxh-S2 (Fig 4), apart from those in positions 5, 8 and 9, might contribute to different specificities regarding their respective target proteins. However, there is also the Nα motif, which in the case of NaTrxh, contributes to the interaction with S-RNase and its reduction [17, 18]. This motif, with some variation, is also present in the other analysed Trxh-S2 (Fig 4) and could contribute to their target specificity too.

Finally, finding other leader-less proteins containing a similar Nβ motif towards the N-terminus–like NaTrxh–with annotated functions related to protein traffic, membrane mobility and secretion (some experimentally confirmed [15, 17, 37, 62, 63]) raises the possibility that their cellular localisation is due to this short sequence. Therefore, the Nβ motif sequence might help to predict the protein localisation of those proteins that contain it.

The cellular localisation of some proteins containing a similar Nβ motif is not yet known (B and D categories of our analysis; Figs 3 and 5). Some, particularly the Trxh-S2 proteins, might move from the B or D category to the A category, as was the case of NaTrxh itself. Notably, the Nβ motif of NaTrxh and other plant proteins containing this motif, must contain a serine at position 8. However, a glutamate residue was usually found at this position in animal-secretory proteins–or traffic-related ones, as is the case of Unc-119– (Fig 3). This opens the possibility of distinctive functional secretion motifs between plant and animal proteins that are worthy of being investigated.


The Nβ motif leads to a post-translational translocation to the ER to further continue secretion, which means that it acts as a transit peptide to the ER. Thus, the NaTrxh secretion–or other Nβ-containing protein with the features analysed here and mainly in plant proteins and cells–, is SRP-independent but utilises both ER and Golgi, referred here as a “semi-conventional” secretion pathway (Fig 8).

The role of the Nβ motif as a secretion or transit peptide depends in its hydrophilic profile with a negative charge and being solvent-exposed as a mobile element rather than its position, so it can be either at the N- or C-terminus but not in a stable conformation as would happen if located at the core of any protein.

Based on what it is known about NaTrxh, its N-terminal extension is not cleaved. While NaTrxh is secreted by the transmitting stylar cells [15], it should remain in the pollen tube cytoplasm after its incorporation during pollination [18] probably because of its interaction with S-RNase which might cause the Nβ motif to be hidden. This suggests the occurrence of a secretion/localisation regulated mechanism, especially for those SP-lacking soluble proteins.

The Ser-8 and Ser-9 appear to be essential for the Nβ motif to work as a secretion signal in plant proteins, particularly in Trxh-S2. In the case of animal proteins, it is possible that the Nβ-like motif plays a traffic role more than a secretory one, although this needs to be further analysed. This latter assumption relies on the fact that animal “secreted” proteins contain Glu-8; however, when this residue was found in plant Trxs–with other sequence variations–it apparently means a mitochondrial localisation, which would also have biological relevance and needs further assessment.

Materials and methods

In silico analysis, image processing and statistics

A BLASTP query with default settings (NCBI; was executed using the Nβ motif sequence as a query. The output sequences were retrieved and classified according to the type of organism they belonged to (animals, fungi, plants). Sequence logos were constructed using the WebLogo server ( to show the consensus sequences [64, 65].

Confocal images were processed with ImageJ Fiji. Statistical analysis and graphs were developed using GraphPad Prism.

Deletion of 3 and 6 positions on the Nβ motif and generation of GFP fusion DNA constructs

The NaTrxh sequence was previously cloned into pENTR/D-TOPO (Invitrogen) [17] to generate pENTR:NaTrxh. This was used as a polymerase chain reaction (PCR) template to amplify and produce [Platinum SuperFi PCR master mix (Invitrogen)] the NaTrxhΔNα(+3) and the NaTrxhΔNα(+6) sequences using the forward primers Nβ(-3)-F or Nβ(-6)-F, respectively (S2 Table), with the NaT-R reverse primer for both (S2 Table).

Using the pEG103:Nβ construct [17], in which the sequence is upstream of the GFP gene in the pEarleyGateway103 vector [66] as template, the Nβ(-3)-GFP and the Nβ(-6)-GFP sequences were obtained by PCR with the same forward primers as above and the GFP-R as reverse (S2 Table). Additionally, we also amplified the full Nβ-GFP sequence from the pEG103:Nβ construct using the Nβ-F and GFP-R primers (S2 Table).

PCR products were cloned into pENTR/D-TOPO (Invitrogen) following the manufacturer’s instructions. Then the sequences were transferred by recombination using LR recombinase enzymatic mixture (Invitrogen) to the pEG103 vector [NaTrxhΔNα(+3) and NaTrxhΔNα(+6), generating the pEG103:NaTrxhΔNα(+3) and the pEG103:NaTrxhΔNα(+6) constructs, respectively], or to pK2GW7 [67] in the case of Nβ-GFP, Nβ(-3)-GFP and Nβ(-6)-GFP, resulting in the pK2:Nβ-GFP, pK2:Nβ(-3)-GFP and the pK2:Nβ(-6)-GFP constructs, respectively. All constructs were transformed into Escherichia coli XL10-GOLD cells (Agilent), pEG103 construct transformants were selected using kanamycin (50 μg/mL in solid and 100 μg/mL in liquid medium) and pK2GW7 construct transformants with spectinomycin (50 μg/mL). All constructs were confirmed by DNA sequencing.

Site-directed mutagenesis

To obtain the different Nβ variants (S5G, S8E, S8D, S8A and S9A; positions corresponding to the Nβ motif sequence), the pK2:Nβ-GFP construct was used as template for site-directed mutagenesis by PCR using the QuickChange Lightning Site-Directed Mutagenesis Kit (Agilent) following the manufacturer’s instructions with the concomitant mutagenic primers for each target base substitution (S2 Table). Presence of the desired mutations was confirmed for each construct by DNA sequencing and confirmed no other nucleotide changes were detected. All resulting sequences were transformed into E. coli XL10-GOLD cells and selected using spectinomycin (50 μg/mL) resistance.

Generation of DNA constructs with the Nβ motif at different positions, with the Nβ sequence inverted, or with the ER retention signal

To assess the relevance of the Nβ position within the protein to its role as a secretion signal sequence, we generated two different constructs: (1) NaTrxh(ΔNαβ)-Nβ-GFP; and (2) GFP-Nβ. The first one, encodes a chimeric protein that contains the Nβ motif between the core of NaTrxh (without its N-terminal extension) and GFP, and the second one contains this motif at the GFP C-terminus.

The NaTrxh(ΔNαβ)-Nβ sequence was generated by two sequential PCR amplifications. In the first one, the pair of primers NaTΔαβ-F and NaT-Nβ-R1 (S2 Table) were used with the pENTR:NaTrxh construct [17] as a template. The PCR product was analysed by electrophoresis and purified to use it as the template for the second amplification, in which the same forward primer was used and the reverse one was NaT-Nβ-R2 (S2 Table).

The GFP-Nβ sequence was also generated by two sequential PCR amplifications. For the first one, with the pEG103 vector as the template, the primers GFP-F and GFP-Nβ-R1 (S2 Table) were used. The PCR product was analysed by electrophoresis and purified to use it as template for the second amplification, in which the same forward primer was used and the reverse one was GFP-Nβ-R2 (S2 Table).

The Nβinv-GFP sequence was obtained after two sequential PCR amplifications as described for the NaTrxh(ΔNαβ)-Nβ and GFP-Nβ sequences but using pK2:Nβ-GFP as the template and the primers Nβinv-F1 and GFP-R2 (S2 Table) in the first reaction and for the second amplification the primers Nβinv-F2 and GFP-R2 were used (S2 Table).

The GFP-Nβ(KDEL) protein contains the ER retention signal at its C-terminus [23, 4549]. Its coding sequence was generated by PCR using the pENTR4:GFP-Nβ construct (described below) as the template and the primers GFP-F and NβKDEL-R (S2 Table).

The four sequences were subcloned into pJET using the CloneJET PCR cloning Kit (ThermoScientific) and confirmed by DNA sequencing. The inserts were released by Bam-HI and Eco-RI digestion and ligated to pENTR4 (Invitrogen; previously digested with the same restriction enzymes), separated by electrophoresis and purified (eliminating the dual selection cassette between the attL1 and attL2 sites). The resulting constructs [pENTR4:NaTrxh(ΔNαβ)-Nβ, pENTR4:GFP-Nβ, pENTR4:Nβinv-GFP and pENTR4:GFP-Nβ(KDEL)] were transformed into E. coli XL10-GOLD cells, selecting transformant cells with kanamycin (50 μg/mL in solid and 100 μg/mL in liquid medium).

To generate the NaTrxh(ΔNαβ)-Nβ-GFP fusion, we transferred the sequence from pENTR4:NaTrxh(ΔNαβ)-Nβ to pEG103 by LR recombination. Since both entry and destination vectors provide kanamycin resistance, we followed a previously reported strategy [68] of amplifying by PCR [using pENTR4:NaTrxh(ΔNαβ)-Nβ as the template] a fragment from upstream of the attL1 site (pENTR4-F primer; S2 Table) to downstream of the attL2 site (pENTR4-R primer; S2 Table), purifying the resulting PCR product and using it for the recombination step. The pENTR4:GFP-Nβ, pENTR4:Nβinv-GFP and pENTR4:GFP-Nβ(KDEL) were used directly for the LR recombination reaction into pK2GW7, which confers spectinomycin resistance (selected using 50 μg/mL spectinomycin). All these constructs [pEG103:NaTrxh(ΔNαβ)-Nβ, pK2:GFP-Nβ, pK2:Nβinv-GFP and pK2:GFP-Nβ(KDEL)] were transformed into E. coli XL10-GOLD cells.

Transient expression assays in onion epidermal cells

The size, the relative transparency of onion (Allium cepa) epidermal cells that usually occur as a monolayer represent a clear advantage for light microscopy assays [23] and are useful in analysing cell processes such as secretion [2325]. Each construct, either in pEG103 or pK2GW7, was individually bombarded into onion epidermal cells as previously reported [17, 23] using 2 cm2 slices, which were placed on 0.5x Murashige-Skoog medium, pH 5.5–57, with 1% (w/v) agar and 4.5 g/L TC-gel (Caisson). Transformation protocol (biolistic) was performed with 1.0 μm tungsten particles (BioRad M17) at a 6 cm distance with a pressure of 1100 psi in te PDS1000/He system (BioRad). After 24–48 h of particle bombardment, onion epidermal cells were incubated in 1.0 M NaCl for 5 to 10 min to separate the plasma membrane from the cell wall (plasmolysis). Fluorescence was visualized using an Olympus FV 1000 confocal microscope or Zeiss LSM 800 confocal microscope [From Fig 3C: Nβ(S5G), Nβ(S8A), Nβ(S8D) and Nβ(S9A)]; in both cases with 485/670 nm excitation/emission light for GFP. Where indicated, the nuclei were stained with propidium iodine (Sigma-Aldrich) prior to plasmolysis and visualised with 570/670 nm excitation/emission light. The localization of GFP fluorescence was confirmed at least three times (independent transformation assays) for each construct, from where each construct exhibited the same localization and having at least five cells showing it was considered as consistent.

Supporting information

S1 Table. Sequences raised from the BLASTP analysis.

Retrieved sequences from the BLASTP analysis with an identical or similar Nβ motif sequence. Protein sequences are arranged in rows with the information of each one of them in the columns. Column A: assigned number of the proteins, organised according to the taxon each one corresponds to. Column B: GenBank accession code; Column C: name of the protein and the species where it is from; Columns D and E: query coverage and identity percentages with Nβ motif, respectively; Column F: annotated functions; Column G and H: species and taxon where each protein is from, respectively; Column I: amino acid position of the Nβ motif in the primary structure; Column J: position where the Nβ motif locates within the primary structure (P1, P2, P3 or P4, according to the analysis described in the main text); Column K: size of the primary structure; Column L: cellular localisation. EC: extracellular, Cyt: cytoplasmic, NA: data not available; Column M: score regarding the cellular localisation according to the UniProtKB database, where 5 is the highest and 1 the lowest; Column N: score resulted from SignalP 6.0 to predict the presence of a signal peptide, where 1 is a high probability of the presence of a signal peptide; Column O: organelle or cellular region to which the protein is associated (NA: data not available). The colour code represents the different categories generated in this analysis (described in the main text): green (A category), yellow (B category), red (C category) and gray (D category).


S2 Table. Primers used for the different constructs generated for transient expression assays in onion epidermal cells.



We thank K. Jiménez-Durán (USAII-FQ-UNAM), R. Rincón-Heredia and A. Rosas-Arellano (IFC-UNAM) for their technical assistance with confocal microscopy. To M. Olvera-Flores (FQ-UNAM) for her technical assistance with bombardment transformation. To A. Beacham (Harper Adams University) for the editing work. We also thank Posgrado en Ciencias Biológicas, UNAM, for the support provided to A.Z-G. and J.A.J-D.


  1. 1. Shikano S, Colley KJ. Secretory pathway. In: William J, Lennarz M, Lane MD, editors. Ecnyclopedia of Biological Chemistry. Academic Press. Elsevier, Inc.; 2013. pp. 203–209.
  2. 2. Wang X, Chung KP, Lin W, Jiang L. Protein secretion in plants: conventional and unconventional pathways and new techniques. J Exp Bot. 2018; 69: 21–37.
  3. 3. Driouich A, Follet-Gueye ML, Bernard S, Kousar S, Chevalier L, Vicre-Gibouin M, et al. Golgi-mediated synthesis and secretion of matrix polysaccharides of primary cell wall of higher plants. Front Plant Sci. 2012; 3: 79.
  4. 4. Worden N, Park E, Drakakaki G. Trans-Golgi network: an intersection of trafficking cell wall components. J Integr Plant Biol. 2012; 54: 875–886. pmid:23088668
  5. 5. Drakakaki G, Dandekar A. Protein secretion: how many secretory routes does a plant cell have? Plant Sci. 2013; 203–204: 74–78. pmid:23415330
  6. 6. Blobel G, Dobberstein B. Transfer of proteins across membranes. J Cell Biol. 1975; 67: 835–851.
  7. 7. Gierasch L. Signal sequences. Biochemistry. 1989; 28: 923–930. pmid:2653440
  8. 8. Keenan RJ, Freymann DM, Stroud RM, Walter P. The signal recognition particle. Annu Rev Biochem. 2001; 70: 755–775. pmid:11395422
  9. 9. Grudnik P, Bange G, Sinning I. Protein targeting by the signal recognition particle. Biol. Chem. 2009; 390: 775–782. pmid:19558326
  10. 10. Chang Z, Fu X. Biogenesis of secretory proteins in eukaryotic and prokaryotic cells. In: Bradshaw RA, Hart GW, Stahl PD, editors. Encyclopedia of cell biology (second edition). Academic Press; 2023. pp. 689–702.
  11. 11. Maas SLN, Breakefield XO, Weaver AM. Extracellular vesicles: unique intercellular delivery vehicles. Trends Cell Biol. 2017; 27: 172–188. pmid:27979573
  12. 12. Ding Y, Wang J, Wang J, Stierhof Y-D, Robinson DG, Jiang L. Unconventional protein secretion. Trends Plant Sci. 2012; 17: 606–615. pmid:22784825
  13. 13. Rabouille C. Pathways of unconventional protein secretion. Trends Cell Biol. 2017; 27: 230–240. pmid:27989656
  14. 14. Nombela C, Gil C, Chaffin WL. Non-conventional protein secretion in yeast. Trends Microbiol. 2006; 14: 15–21. pmid:16356720
  15. 15. Juárez-Díaz JA, McClure B, Vázquez-Sanana S, Guevara-García A, León-Mejía P, Márquez-Guzmán J, et al. A novel thioredoxin h is secreted in Nicotiana alata and reduces S-RNase in vitro. J Biol Chem. 2006; 281: 3418–3424.
  16. 16. Gelhaye E, Rouhier N, Jacquot JP. The thioredoxin h system of higher plants. Plant Physiol Biochem. 2004; 42: 265–271.
  17. 17. Ávila-Castañeda A, Juárez-Díaz JA, Rodríguez-Sotres R, Bravo-Alberto CE, Ibarra-Sánchez CP, Zavala-Castillo A, et al. A novel motif in the NaTrxh N-terminus promotes its secretion, whereas the C-terminus participates in its interaction with S-RNase in vitro. BMC Plant Biol. 2014; 14: 147.
  18. 18. Torres-Rodríguez MD, Cruz-Zamora Y, Juárez-Díaz JA, Mooney B, McClure BA, Cruz-García F. NaTrxh is an essential protein for pollen rejection in Nicotiana by increasing S-RNase activity. Plant J. 2020; 103: 1304–1317.
  19. 19. McClure BA, Franklin-Tong V. Gametophytic self-incompatibility: understanding the cellular mechanisms involved in “self” pollen tube inhibition. Planta. 2006; 224: 233–245. pmid:16794841
  20. 20. McClure B, Cruz-García F, Romero C. Compatibility and incompatibility in S-RNase-based systems. Ann Bot. 2011; 108: 647–658. pmid:21803740
  21. 21. Goring D, Cruz-Garcia F, Franklin-Tong V. Self-incompatibility. eLS (Plant Science). 2022; 3: 1–12.
  22. 22. Bendtsen JD, Jensen LJ, Blom N, von Heijne G, Brunak S. Feature-based prediction of non-classical and leaderless protein secretion. Protein Eng Des Sel. 2004; 17: 349–356. pmid:15115854
  23. 23. Scott A, Wyatt S, Tsou P-L, Robertson D, Strömgren A. Model system for plant cell biology: GFP imaging in living onion epidermal cells. BioTechniques. 1999; 26: 1125–1132. pmid:10376152
  24. 24. Sun W, Cao Z, Zhao Y, Zhang H. A simple and effective method for protein subcellular localization using Agrobacterium-mediated transformation of onion epidermal cells. Biologia. 2007; 62: 529–532.
  25. 25. Kitajima A, Satsuma S, Okada H, Hamada Y, Kaneko K, Nanjo Y, et al. The rice α-amylase glycoprotein is targeted from the Golgi apparatus through the secretory pathway to the plastids. Plant Cell. 2009; 21: 2844–2858.
  26. 26. The UniProt Consortium. UniProt: the universal protein knowledgebase in 2023. Nucl Acids Res. 2023; 51: D523–D531. pmid:36408920
  27. 27. Constantine R, Zhang H, Gerstner CD, Frederick JM, Baehr W. Uncoordinated (UNC)119: coordinating the trafficking of myristoylated proteins. Vision Res. 2012; 75: 26–32. pmid:23000199
  28. 28. Geigenberger P, Thormählen I, Daloso DM, Fernie AR. The unprecedented versatility of plant thioredoxin system. Trends Plant Sci. 2017; 22: 249–262.
  29. 29. Buchanan BB. Regulation of CO2 assimilation in oxygenic photosynthesis: the ferredoxin/thioredoxin system. Perspectives on its discovery, present status, and future development. Arch Biochem Biophys. 1991; 288: 1–9.
  30. 30. Mestres-Ortega D, Meyer Y. The Arabidopsis thaliana genome encodes at least four thioredoxins m and a new prokaryotic-like thioredoxin. Gene. 1999; 240: 307–316.
  31. 31. Laloi C, Rayapuram N, Chartier Y, Grienenberger JM, Bonnard G, Meyer Y. Identification and characterization of a mitochondrial thioredoxin system in plants. Proc Natl Acad Sci USA. 2001; 98: 14144–14149. pmid:11717467
  32. 32. Collin V, Issakidis-Bourget E, Marchand C, Hirasawa M, Lancelin J-M, Knaff DB, et al. The Arabidopsis plastidial thioredoxins. New functions and new insights into specificity. J Biol Chem. 2003; 278: 23747–23752.
  33. 33. Lemaire SD, Collin V, Keryer E, Quesada M, Miginiac-Maslow M. Characterization of thioredoxin y, a new type of thioredoxin identified in the genome of Chlamydomonas reinhardtii. FEBS Lett. 2003; 543: 87–92.
  34. 34. Alkhalfioui F, Renard M, Frendo P, Keichinger C, Meyer Y, Gelhaya , et al. A novel type of thioredoxin dedicated to symbiosis in legumes. Plant Physiol. 2008; 148: 424–435. pmid:18614707
  35. 35. Rivera-Madrid R, Mestres D, Marinho P, Jacquot J-P, Decottignies P, Miginiac-Maslow M, et al. Evidence for five divergent thioredoxin h sequences in Arabidopsis thaliana. Proc Natl Acad Sci USA. 1995; 92: 5620–5624.
  36. 36. Gelhaye E, Rouhier N, Gerard J, Jolivet Y, Gualberto J, Navrot N, et al. A specific form of thioredoxin h occurs in plant mitochondria and regulates the alternative oxidase. Proc Natl Acad Sci USA. 2004; 101: 14545–14550.
  37. 37. Meng L, Wong JH, Feldman LJ, Lemaux PG, Buchanan BB. A membrane-associated thioredoxin required for plant growth moves from cell to cell, suggestive of a role in intercellular communication. Proc Natl Acad Sci USA. 2010; 107: 3900–3905. pmid:20133584
  38. 38. Ishiwatari Y, Honda C, Kawashima I, Nakamura SI, Hirano H, Mori S, et al. Thioredoxin h is one of the major proteins in rice phloem sap. Planta. 1995; 195: 456–463. pmid:7766047
  39. 39. Ishiwatari Y, Fujiwara T, McFarland KC, Nemoto K, Hayashi H, Chino M, et al. Rice phloem thioredoxin h has the capacity to mediate its own cell-to-cell transport through plasmodesmata. Planta. 1998; 205: 12–22. pmid:9599802
  40. 40. Pekkari K, Avila-Carino J, Bengtsson Å, Gurunath R, Scheynius A, Holmgren A. Truncated thioredoxin (Trx80) induces production of interleukin-12 and enhances CD14 expression in human monocytes. Blood. 2001; 97: 3184–3190. pmid:11342447
  41. 41. Jones AM, Xuan Y, Xu M, Wang R-S, Ho C-H, Lalonde S, et al. Border control–a membrane-linked interactome of Arabidopsis. Science. 2014; 344: 711–716.
  42. 42. Torres-Rodríguez MD, González-Segura L, Rodríguez-Sotres R, Juárez-Díaz JA, Cruz-Zamora Y, Cruz-García F. High resolution crystal structure of NaTrxh from Nicotiana alata and its interaction with the S-RNase. J Struct Biol. 2020; 212: 107578.
  43. 43. von Heijne G. Protein targeting signals. Curr Opin Cell Biol. 1990; 2: 604–608. pmid:2252586
  44. 44. Janda CY, Li J, Oubridge C, Hernández H, Robinson CV, Nagai K. Recognition of a signal peptide by the signal recognition particle. Nature. 2010; 465: 507–511. pmid:20364120
  45. 45. Munro S, Pelham HRB. An Hsp70-like protein in the ER: identity with the 78 kd glucose-regulated protein and immunoglobulin heavy chain binding protein. Cell. 1986; 46: 291–300. pmid:3087629
  46. 46. Munro S, Pelham HRB. A C-terminal signal prevents secretion of lumninal ER proteins. Cell. 1987; 48: 899–907.
  47. 47. Andres DA, Dickerson IM, Dixon JE. Variants of the carboxyl-terminal KDEL sequence direct intracellular retention. J Biol Chem. 1990; 265: 5952–5955. pmid:2318841
  48. 48. Andres DA, Rhodes JD, Meisel RL, Dixon JE. Characterization of the carboxyl-terminal sequences responsible for protein retention in the endoplasmic reticulum. J Biol Chem. 1991; 266: 14277–14282. pmid:1650354
  49. 49. Arakawa R, Chong DKX, Merritt L, Langridge WHR. Expression of cholera toxin B subunit oligomers in transgenic potato plants. Transgenic Res. 1997; 6: 403–413. pmid:9423288
  50. 50. Gilmore TD. Introduction to NF-κB: players, pathways, perspectives. Oncogene. 2006; 25: 6680–6684.
  51. 51. Wan F, Lenardo MJ. The nuclear signaling of NF-κB–current knowledge, new insights, and future perspectives. Cell Res. 2010; 20: 24–33.
  52. 52. de Nettancourt D. The genetics of self-incompatibility. In: De Nettancourt , editor. Incompatibility and incongruity in wild and cultivated plants. Berlin: Springer Verlag; 2001. pp. 25–72.
  53. 53. Takayama S, Isogai A. Self-incompatibility in plants. Annu Rev Plant Biol. 2005; 56: 467–489. pmid:15862104
  54. 54. Anderson MA, Cornish EC, Mau S-L, Williams EG, Hoggart R, Atkinson A, et al. Cloning of cDNA for stylar glycoprotein associated with expression of self-incompatibility in Nicotiana alata. Nature. 1986; 321: 38–44.
  55. 55. Anderson MA, McFadden GI, Bernatzky R, Atkinson A, Orpin T, Dedman H, et al. Sequence variability of three alleles of the self-incompatibility gene of Nicotiana alata. Plant Cell. 1989; 1: 483–491.
  56. 56. Luu D-T, Qin X, Morse D, Cappadocia M. S-RNase uptake by compatible pollen tubes in gametophytic self-incompatibility. Nature. 2000; 407: 649–651. pmid:11034216
  57. 57. Goldraij A, Kondo K, Lee CB, Hancock CN, Sivaguru M, Vazquez-Santana S, et al. Compartmentalization of S-RNase and HT-B degradation in self-incompatible Nicotiana. Nature. 2006; 439: 805–810.
  58. 58. Akopian D, Shen K, Zhang X, Shan S. Signal recognition particle: an essential protein-targeting machine. Annu Rev Biochem. 2013; 82: 693–721. pmid:23414305
  59. 59. Holmgren A. Thioredoxin. Ann Rev Biochem. 1985; 54: 237–271. pmid:3896121
  60. 60. Marchand C, Maréchal PL, Meyer Y, Miginiac-Maslow M, Issakidis-Bourget E, Decottignies P. New targets of Arabidopsis thioredoxins revealed by proteomic analysis. Proteomics. 2004; 4: 2696–2706. pmid:15352244
  61. 61. Holmgren A. Thioredoxin catalyzes the reduction of insulin disulfides by dithiothreitol and dihydrolipoamide. J Biol Chem. 1979; 254: 9627–9632. pmid:385588
  62. 62. Vidulescu C, Clejan S, O’Connor KC. Vesicle traffic through intercellular bridges in DU 145 human prostate cancer cells. J Cell Mol Med. 2004; 8: 388–396. pmid:15491514
  63. 63. Lee Y, Chung S, Baek IK, Lee TH, Paik SY, Lee J. UNC119a bridges the transmission of Fyn signals to Rab11, leading to the completion of cytokinesis. Cell cycle. 2013; 12: 1303–1315. pmid:23535298
  64. 64. Schneider TD, Stephens RM. Sequence logos: a new way to display consensus sequences. Nucl Acids Res. 1990; 18: 6097–6100. pmid:2172928
  65. 65. Crooks GE, Hon G, Chandonia JM, Brenner SE. WebLogo: a sequence logo generator. Genome Res. 2004; 14: 1188–1190. pmid:15173120
  66. 66. Earley KW, Haag JR, Pontes O, Opper K, Juehne T, Song K, et al. Gateway-compatible vectors for plant functional genomics and proteomics. Plant J. 2006; 45: 616–629. pmid:16441352
  67. 67. Karimi M, Inzé D, Depicker A. GATEWAY vectors for Agrobacterium-mediated plant transformation. Trends Plant Sci. 2002; 7: 193–195.
  68. 68. Kumar K, Yadav S, Purayanur S, Verna PK. An alternative approach in Gateway cloning when the bacterial antibiotic selection cassettes of the entry clone and destination vector are the same. Mol Biotechnol. 2013; 54: 133–140.