Biological context

Spider dragline silk is well recognized for its biocompatibility and unsurpassed toughness, which is a combination of superior strength and extensibility. The average toughness of dragline silk from Araneus diadematus is 180 MJ/m3, which is approximately three times tougher than that of man-made synthetic fibers, such as Kevlar 49 (Heim et al. 2009). Due to its fascinating properties, dragline silk has become a promising biomaterial for many industrial applications, such as artificial skin, drug delivery, and surgical suturing (Lammel et al. 2011; Wendt et al. 2011; Hennecke et al. 2013).

Dragline silk is a hierarchically structured fiber that is composed of several proteins, known as major ampullate spidroins (MaSps). Among spiders, these proteins are different in terms of molecular weight, amino acid composition and functional (i.e., mechanical) impact on mechanical properties. The archetypal dragline spidroins are MaSp1 and MaSp2, which mainly exhibit differences in their proline content (MaSp 1 < 0.4%, MaSp2 > 10%) (Guerette et al. 1996). Recently, several studies have demonstrated that the composition of dragline silks is more complex than previously known and includes multiple components, such as MaSp1, MaSp2, MaSp3, cysteine-rich rich protein (CRPs) and low molecular weight nonspidroin components (Pham et al. 2014; Chaw et al. 2015; Larracas et al. 2016; Kono et al. 2019, 2021). However, the transformation of soluble spidroin into insoluble fiber is still not completely understood (Malay et al. 2022).

Spidroins consist of a large core of repetitive domains flanked by small nonrepetitive N-terminal and C-terminal domains (molecular weights of NTD and CTD are 14 kDa and 12.5 kDa, respectively) (Gatesy et al. 2001; Garb et al. 2010). Each repeat contains 30–60 amino acid residues with specific amino acid motifs. The consensus motifs in the repetitive domain mainly consist of polyalanine (number of alanine residues = 4–12) and glycine-rich sequences (GGX- or GPGXX-). The β-sheet structure originating from the polyalanine region is the key motif responsible for the tensile strength of dragline silk (Keten et al. 2010). The glycine-rich region contains the GGX motif in MaSp1 and the GPGXX motif in MaSp2, which adopts a PPII helix and type II β-turn structure, respectively (Simmons et al. 1994; Kümmerlen et al. 1996; van Beek et al. 2002; Jenkins et al. 2010; Jenkins et al. 2012), and is believed to confer elasticity to silk fibers. The higher proline content in MaSp2 suggests that this protein is more elastic than MaSp1 (Rising et al. 2005). In the storage sac, spidroin is stored at high concentrations (up to 50%) (Hijirida et al. 1996) and converted into insoluble spider silks after experiencing pH and ion gradients across the major ampullate gland (Knight and Vollrath 2001; Andersson et al. 2014). Despite the predominant size of the repetitive domain, the self-assembly of spider silks is regulated by the small and highly conserved terminal domains (Askarieh et al. 2010; Hagn et al. 2010). At neutral pH and high chaotropic salt concentrations, similar to the conditions found in the storage sac, the CTD forms a folded dimer. In contrast, at acidic pH and low chaotropic salt concentration, which resembles the condition at the spinning duct, the CTD becomes unstructured and triggers spider silk self-assembly (Hagn et al. 2010; Andersson et al. 2014). Previous studies indicated that folded dimer CTD is required to maintain the solubility of spidroin, while the unfolding of CTD was believed to act as a precursor for β-sheet formation, suggesting the essential role of CTD as a molecular switch for storage and spider silk self-assembly (Hagn et al. 2010; Kronqvist et al. 2014).

Compared to the MaSp NTD from different species, which exhibits the most conserved region in silk proteins (Motriuk-Smith et al. 2005), the amino acid sequences of the CTD from different species are more varied (Fig. 1a). Nevertheless, the cysteine residue in the CTD is conserved across spider silk types and species and facilitates dimer formation (Fig. 1a). Furthermore, the CTD also contains a conserved salt bridge that is formed by the arginine side chain in helix 2 and the glutamic acid side chain in helix 4 (Fig. 1a). This salt bridge is pH sensitive and functions to stabilize the folded CTD by forming intramolecular interactions between helix 2 and helix 4 (Strickland et al. 2018). At acidic pH, the salt bridge will be disrupted, which leads to unfolding of the CTD (Hagn et al. 2010).

Fig. 1
figure 1

The dimeric form of the MaSps CTD from different species is facilitated by disulfide bonds. a Sequence alignment of the MaSps CTDs from different species. LH-M2 Latrodectus hesperus MaSp2 CTD; LH-M1 Latrodectus hesperus MaSp1 CTD; ADF-3 Araneus diadematus 3 CTD; TC-M2 Triconephila clavipes MaSp2 CTD; AV-M1 Argiope ventricosus MaSp1 CTD. The cysteine residue is conserved on those sequences (indicated by black arrow), and the arginine residue (indicated by blue arrow) and glutamic acid residue (indicated by red arrow) are also conserved and hypothesized to form salt bridges, as mentioned previously (Strickland et al. 2018). b SDS page profile of pure recombinant (13C, 15N) L. hesperus MaSp2 CTD in monomeric form (M) and dimeric form (D). The monomeric form of L. hesperus MaSp2 CTD was obtained by adding β-mercaptoethanol to the protein solution since it can reduce disulfide bonds. The molecular weights of the monomeric form and dimeric form of L. hesperus CTD are 12.2 and 24.4 kDa, respectively

Full size image

In this study, we present 1H, 13C, and 15N backbone and side chain chemical shift assignments and the dynamics of soluble CTD MaSp2 from L. hesperus at pH 7 in the presence of 300 mM NaCl in dimeric form (Fig. 1b) at 25 °C. The backbone chemical shift assignment of CTD MaSp2 was also translated into a secondary structure, demonstrating that the 5 helix regions were connected by a more flexible linker.

Method and experiments

Recombinant protein expression and purification

The spidroin CTD MaSp2 used in the study was based on sequences from Latrodectus hesperus (UniProt accession code A6YIY0). Constructs encoding the WT sequences were purchased from GenScript as NdeI/XhoI inserts in the pET15b vector (Novagen). The amino acid sequence used for the L. hesperus MaSp2 CTD is shown in Fig. 2a.

Fig. 2
figure 2

NMR assignment of L. hesperus MaSp2 CTD at pH 7 in the presence 300 mM NaCl (a). Amino acid sequences of L. hespereus CTD MaSp2 used in this construct. b 2D 1H-15N HSQC of backbone L. hesperus MaSp2 CTD signals at pH 7 in the presence 300 mM NaCl. The signals originating from unfolded populations are indicated in dashed box and glycine signal which is potentially originating from proline cis–trans isomerization is also indicated (G74*)

Full size image

Doubly labeled (13C, 15N) L. hesperus CTD MaSp2 was expressed in E. coli BL21 (DE3). Cells were grown initially in 5 mL Luria Bertani (LB) medium supplemented with 100 µg/mL ampicillin sodium at 37 °C with overnight shaking at 160 rpm. The LB culture was then transferred into 100 mL of M9 minimal medium containing 2 g/L 13C-glucose and 1 g/L 15N-ammonium chloride supplemented with ampicillin. Cells were grown overnight at 37 °C with shaking at 160 rpm. This preculture was transferred into the main culture of M9 minimal medium containing 100 μg/mL ampicillin. The cells were grown until they reached OD600 ∼ 1. Then, protein expression was induced by adding 0.4 mM isopropyl β-d-1-thiogalactopyranoside (IPTG). The cell cultures were continuously incubated overnight at 20 °C with shaking at 160 rpm.

Cells were harvested by centrifugation, resuspended in 50 mM Tris–Cl, 10% glycerol and 0.1% Triton X-100 (pH 7.5) and stored at − 80 °C until further use. Frozen cell slurry was thawed at 37 °C and enzymatically lysed by adding hen egg lysozyme at 50 mg per 1 L culture (Wako) supplemented with 250 U of TurboNuclease (Accelagen) and complete EDTA-free protease inhibitor cocktail (Roche). After incubation at 4 °C with stirring for 4 h, the lysates were centrifuged at 8000 rpm at 4 °C for 30 min. The clear supernatant fraction was loaded onto a 5-mL HisTrap column and washed extensively with a solution containing 20 mM Tris–HCl, 20 mM imidazole, and 500 mM NaCl (pH 7.5). The His-tagged CTD MaSp2 was eluted in solution containing 20 mM Tris–HCl, 500 mM imidazole, and 500 mM NaCl (pH 7.5). The His-tag was removed by overnight digestion with thrombin at 4 °C. For NMR measurements, the buffer was exchanged into 10 mM phosphate buffer pH 7 in the presence of 300 mM NaCl. Protein concentration was determined by measuring absorbance at 280 nm using a Nanodrop instrument (Thermo Fischer Scientific). The purity of the protein was evaluated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) (Fig. 1b). The NMR sample contained ~ 0.8–1 mM of dimeric form of (13C, 15N) L. hesperus MaSp2 CTD in 10 mM phosphate buffer pH 7 in the presence of 300 mM NaCl supplemented with 10% D2O and 0.1 mM DSS.

NMR experiments

Sequential backbone chemical shifts of L. hesperus MaSp2 CTD at pH 7 were assigned based on several 2D and 3D NMR experiments, including 2D 1H–15N HSQC, 2D 1H–13C HSQC aliphatic, 3D HNCO/3D HN(CA)CO (Ikura et al. 1990; Yamazaki et al. 1994), 3D CBCA(CO)NH/3D HNCACB (Wittekind and Mueller 1993), 3D HNCA/3D HN(CO)CA (Bax and Ikura 1991), and 3D HBHA(CO)NH (Grzesiek and Bax 1993). Assignment of proton and carbon aliphatic and side chain chemical shifts was accomplished by assigning 3D H(CCO)NH and 3D (H)C(CO)NH TOCSY spectra (Logan et al. 1993). The side chain carboxyl group of aspartic and glutamic acids as well as the side chain carbonyl group of asparagine and glutamine chemical shift of L. hesperus MaSp2 CTD at pH 7 were assigned based on the 2D H2(C)CO spectrum (Oda et al. 1994; Oktaviani et al. 2011). The 2D correlation between the Nε and Hδ of the arginine chemical shift was assigned based on the 2D H2(C)N spectrum (André et al. 2007). All NMR data were recorded at 25 °C (298 K) using a triple resonance TCI cryogenic probe and a z-axis gradient coil with a 700 MHz Bruker spectrometer. All spectra were processed using NMRPipe (Delaglio et al. 1995) and analyzed using NMRFAM-SPARKY (Lee et al. 2015). Sodium trimethylsilylpropanesulfonate (DSS) was used as a reference standard for all NMR signals based on IUPAC recommendations (Markley et al. 1998). The structural propensities of the L. hesperus MaSp2 CTD at pH 7 and 25 °C were determined using the neighbor-corrected structural propensity calculator (ncSPC) (Tamiola and Mulder 2012).

To characterize the protein backbone dynamics on the picosecond-nanosecond time scale, {1H}-15N heteronuclear NOE experiments were measured using 700 MHz NMR machines (Bruker) by recording the following spectra: an initial spectrum recorded without the initial proton saturation and a second spectrum recorded with initial proton saturation (3 s). The steady-state NOE values were determined based on the ratios of the average intensities of the peaks with and without proton saturation. The standard deviations of the NOE values were calculated from the background noise level using the following formula:

$$\frac{\sigma NOE}{NOE}=\sqrt{{\left(\frac{\sigma {I}_{sat}}{{I}_{sat}}\right)}^{2}+{\left(\frac{\sigma {I}_{unsat}}{{I}_{unsat}}\right)}^{2}}$$
(1)

where Isat and Iunsat are the measured intensities of the peaks in the presence and absence of proton saturation, respectively. The noise in the background regions of the spectra, which were measured with initial proton saturation and without initial proton saturation, are indicated by σIsat and σIunsat, respectively.

Extent of assignment and data deposition

The 2D 1H-15N HSQC spectrum (Fig. 2b) demonstrates a wide dispersion of L. hesperus CTD MaSp2 at pH 7 on the amide proton domain, suggesting that this protein is well folded. The amide proton and nitrogen chemical shifts from the first 3 amino acids (GSH), which originated from the pET 15b vector, as well as the amide proton and nitrogen resonances of G7, S17, S18, and V19, which are part of the flexible N-terminus linker of the CTD, could not be assigned, possibly due to rapid exchange of amide protons with water. Furthermore, the amide proton and nitrogen resonances of T52, S76, S77 and G107 could not be assigned. As the structure of CTD responses to the change of pH and becomes fully unfolded at acidic pH (Hagn et al. 2010; Andersson et al. 2014), interestingly, a few visible signals originating from unfolded population of CTD (shown in dashed box) are observable, even at neutral pH (Fig. 2b). In addition, the low intensity of glycine signal which is potentially originating from proline cis–trans isomerization (G74*) is also noticeable (Fig. 2b). In total, the backbone 1HN, 15NH, 13C’, 13Cα, 13Cβ, and 1Hα chemical shift assignments were approximately 90% complete, and the completeness of the backbone chemical shift assignments of the structured region (residues 20–125) was approximately 93%. The assignment of aliphatic carbon and proton chemical shifts in the side chain based on H(CCO)NH and (H)C(CO)NH experiments was approximately 79% complete. The Cβ chemical shift of C78 was found at 44.148 ppm, suggesting that the cysteine was oxidized to form a dimer via a disulfide bond. Carboxyl and carbonyl side chain chemical shifts of Asp, Glu, Gln and Asn were achieved via the assignment of the 2D H2(C)CO spectrum, which gives the correlation between Hγ-Cδ chemical shifts for Glu and Gln and Hβ-Cγ chemical shifts for Asp and Asn. Interestingly, compared to previously reported structures on the CTD (PDB accession code 2KHM) (Hagn et al. 2010), which demonstrated two intramolecular salt bridges (R43–D93 and R52–E101), the L. hesperus MaSp2 CTD has only 1 arginine (R38), which is located in helix 2. The side chain of Nε of R38, which is located at helix 2, was assigned at 87.5 ppm, suggesting that this arginine is protonated at this pH (Platzer et al. 2014). Potentially, the R38 side chain from helix 2 forms an intramolecular salt bridge with the E87 side chain from helix 4, since these two amino acids are conserved throughout many different species and silk types and hypothesized to form a salt bridge by a previous study (Strickland et al. 2018). In agreement, the carboxyl side chain Cδ of E87 was assigned in the upfield region (177.138 ppm), suggesting that this glutamic acid forms a salt bridge. These data are similar to our recent finding on Triconephila clavipes MaSps NTD, where the carboxyl side chain chemical shift of D40 was shifted to the upfield region (approximately 175 ppm) upon salt bridge formation with its lysine counterparts (Oktaviani et al. 2023).

Translation of backbone chemical shifts (15NH, 13C’, 13Cα, 13Cβ, 1Hα) into secondary structure using neighbor-corrected structural propensity (ncSPC) (Tamiola and Mulder 2012) demonstrated that the L. hesperus MaSp2 CTD at pH 7 contains 5 helix regions (Fig. 3a). Our data also showed that helix 1 and helix 2 are kinked (Fig. 3a) due to the presence of proline, similar to the previously reported structure of the CTD of Araneus diadematus fibroin 3 (ADF-3) (PDB ID: 2KHM) (Hagn et al. 2010). Based on {1H}-15N heteronuclear data, those five helix regions were rigid, and they were connected by a more flexible linker (Fig. 3b). Here, our NMR data suggests that the CTD structures are conserved across the silk type and species, despite the amino acid sequences being relatively divergent (less than 50% sequence similarities). This study is relevant for understanding the conservation of CTD structure and mechanism that leads to the formation of stable spider dragline silks.

Fig. 3
figure 3

Secondary structure and dynamics of L. hesperus MaSp2 CTD in 10 mM phosphate buffer pH 7.0 and 300 mM NaCl. a The secondary structure of the L. hesperus MaSp2 CTD at pH 7 demonstrated that it has five helical regions connected by a linker. b {1 H}-15N heteronuclear NOE measurements for CTD MaSp2. High NOE values (> 0.5) indicate a relatively rigid structure, while low NOE values indicate a more flexible conformation

Full size image