Interfacial water on collagen nanoribbons by 3D AFM

Collagen is the most abundant structural protein in mammals. *

Type I collagen in its fibril form has a characteristic pattern structure that alternates two regions called gap and overlap. The structure and properties of collagens are highly dependent on the water and mineral content of the environment. *

In the article “Interfacial water on collagen nanoribbons by 3D AFM” Diana M. Arvelo, Clara Garcia-Sacristan, Enrique Chacón, Pedro Tarazona and Ricardo Garcia describe how they apply three dimensional atomic force microscopy (3D AFM) to characterize at angstrom-scale resolution the interfacial water structure of collagen nanoribbons.*

Three-dimensional AFM (3D AFM) is an AFM method developed for imaging at high-spatial resolution solid–liquid interfaces in the three spatial coordinates.*

This method has provided atomic-scale images of hydration and solvation layers on a variety of rigid and atomically flat surfaces, such as mica, gibbsite, boehmite, graphite, or 2D materials.*

However, imaging hydration layers on soft materials such as collagen is more challenging than on atomically flat crystalline surfaces.*

On the one hand, the force applied by the AFM tip might deform the protein. On the other hand, the height variations across gap and overlap regions might complicate the imaging of interfacial water.*

In recent years, 3D AFM has expanded its capabilities to image interfacial water on soft materials such as proteins, biopolymers, DNA, lipids, membrane proteins, and cells.*

Those experiments were performed with hydrophilic SiOx AFM tips which are negatively charged under neutral pH conditions.*

The imaging contrast mechanisms and the role of the AFM tip’s composition on the observed solvation structure are under discussion.*

More generally, both theory and experiments performed with very high salt concentrations indicated that the contrast observed in 3D AFM reflects an interplay between water particle and surface charge density distributions.*

For their article the authors apply 3D AFM to study at molecular-scale spatial resolution the structure of interfacial water on collagen nanoribbons.*

Diana M. Arvelo et al. study the influence of the AFM tip’s charge and the salt concentration on the interfacial solvent structure. They report that the interfacial structure depends on the water particle and ion charge density distributions. A non-charged AFM tip reveals the formation of hydration layers on both gap and overlap regions. A negatively charged AFM tip shows that on a gap region, the solvation structure might depart from that of the hydration layers. This effect is attributed to the adsorption of ions from the solution. Those ions occupy the voids existing between collagen molecules in the gap region. *

A home-made three-dimensional AFM was implemented on a commercially available AFM. 3D AFM was performed in the amplitude modulation mode by exciting the AFM cantilever at its first eigenmode.
At the same time when the AFM cantilever oscillates with respect to its equilibrium position, a sinusoidal signal is applied to the z-piezo to modify the relative z-distance between the sample and the AFM tip. *

Diana M. Arvelo et al. have used z-piezo displacements with amplitudes of 2.0 nm and a period (frequency) of 10 ms (100 Hz). The z-piezo signal is synchronized with the xy-displacements in such a way that for each xy-position on the surface of the material, the AFM tip performs a single and complete z-cycle. The z-data are read out every 10.24 µs and stored in 512 pixels (256 pixels per half cycle). Each xy-plane of the 3D map contains 80 × 64 pixels. Hence, the total time to acquire such a 3D-AFM image is 52 s.*

The 3D AFM experiments were performed with two types of AFM probes with different surface chemistries which have different chemical properties in aqueous solutions.*

The high-density carbon/diamond-like AFM tips grown on quartz-like AFM cantilevers that Diana M. Arvelo et al. et al used ( NanoWorld Ultra-Short Cantilevers USC-F1.2-k7.3 for high-speed AFM) remain uncharged at pH 7.4 and are called “neutral” AFM tips in the article.

The silicon AFM cantilevers with silicon AFM tips (NanoWorld Arrow-UHFAuD ultra-high frequency AFM probes) are negatively charged at neutral pH (silicon tips for short in the text) and were used to observe the formation of collagen nanoribbons.

All silicon AFM tips are readily oxidized and are usually covered by a thin native oxide layer which is hydrophilic.

The hydroxyl groups on the surface of the silicon AFM tip become negatively charged while the carbon AFM tips remain neutral (unchanged).

To image at angstrom-scale resolution, the interfacial water structure on the collagen requires reducing the lateral and vertical imaging sizes, respectively, to 5 and 1.5 nm.*

First, the authors introduce the results obtained with carbon-based tips (uncharged, NanoWorld Ultra-Short Cantilevers USC-F1.2-k7.3 for high-speed AFM). Figure 2 (of the cited article) shows some representative 2D force xz panels obtained on gap and overlap regions of a collagen nanoribbon in a concentration of 300 mM KCl. The panels are extracted from a 3D AFM image. The interlayer distances in a gap region are d1 = 0.28 nm and d2 = 0.33 nm (average values) [Fig. 2(a)], while those in an overlap region are d1 = 0.29 and d2 = 0.32 nm (average values) [Fig. 2(b)]. Those values coincide within the experimental error with the values expected for hydration layers on hydrophilic surfaces.

Next, the authors repeated the experiment using other salt concentrations. Figure 2(c) shows that the interlayers distances (within the experimental error) do not depend on the salt concentration or the collagen region. Diana M. Arvelo et al.  remark that entropic effects make the second layer more disordered than the first; therefore, d2 ≥ d1.

The structure and properties of collagens are highly dependent on the water and mineral content of the environment.

For a neutral AFM tip (USC-F1.2-k7.3), the interfacial water structure is characterized by the oscillation of the water particle density distribution with a value of 0.3 nm (hydration layers). The interfacial structure does not depend on the collagen region.

For a negatively charged AFM tip (NanoWorld Arrow-UHFAuD ultra-high frequency AFM probes) the interfacial structure might depend on the collagen region.

Hydration layers are observed in overlap regions, while in gap regions, the interfacial solvent structure is dominated by electrostatic interactions. These interactions generate interlayer distances of 0.2 nm.

The achieved results still need to be explained by the theory of 3D AFM. More detailed theoretical simulations, which are beyond the scope of the cited study, will be required to quantitatively explain the interlayer distances observed over gap regions.

However, the results presented by the authors highlight the potential of 3D AFM to identify the solvent structures on proteins and the complexity of those interfaces.*

Figure 2 from Diana M. Arvelo et al. 2024 “Interfacial water on collagen nanoribbons by 3D AFM”Interfacial liquid water structure on collagen provided by an uncharged tip. (a) 2D force maps (x, y) of the interfacial water structure in the gap region. The map is obtained in a 300 mM KCl solution. The force–distance curves in the bottom of the image are obtained from the top panel. (b) 2D force maps (x, y) of the interfacial water structure in the overlap region. The force–distance curves in the bottom of the image are obtained from the top panel. (c) Statistics of d1 and d2 distances measured from several collagen–water interfaces. The individual force–distance curves from the bottom panels of (a) and (b) are plotted in gray. The average force–distance curve is highlighted by a thick continuous line. The experiments are performed with USC-F1.2-k7.3 cantilevers. Experimental parameters: f = 745 kHz; k = 6.7 N m−1; Q = 8.3; A0 = 150 pm; Asp = 100 pm. The neutral AFM tips used for the research in this article were NanoWorld Ultra-Short Cantilevers USC-F1.2-k7.3 for high-speed AFM (quartz-like AFM cantilevers with a high-density carbon AFM tip grown on them)
Figure 2 from Diana M. Arvelo et al. 2024 “Interfacial water on collagen nanoribbons by 3D AFM”
Interfacial liquid water structure on collagen provided by an uncharged tip. (a) 2D force maps (x, y) of the interfacial water structure in the gap region. The map is obtained in a 300 mM KCl solution. The force–distance curves in the bottom of the image are obtained from the top panel. (b) 2D force maps (x, y) of the interfacial water structure in the overlap region. The force–distance curves in the bottom of the image are obtained from the top panel. (c) Statistics of d1 and d2 distances measured from several collagen–water interfaces. The individual force–distance curves from the bottom panels of (a) and (b) are plotted in gray. The average force–distance curve is highlighted by a thick continuous line. The experiments are performed with USC-F1.2-k7.3 cantilevers. Experimental parameters: f = 745 kHz; k = 6.7 N m−1; Q = 8.3; A0 = 150 pm; Asp = 100 pm.

 

 

Figure 3 from Diana M. Arvelo et al. 2024 “Interfacial water on collagen nanoribbons by 3D AFM”Interfacial liquid water structure on collagen provided by a negatively charged tip. (a) 2D force maps (x, y) of the interfacial water structure in the gap region. The map is obtained in a 300 mM KCl solution. The force–distance curves in the bottom of the image are obtained from the top panel. (b) 2D force maps (x, y) of the interfacial water structure in the overlap region. The force–distance curves in the bottom of the image are obtained from the top panel. (c) Statistics of d1 and d2 distances measured from several collagen–water interfaces. In the bottom panels of (a) and (b), the individual force–distance curves from the bottom panels of (a) and (b) are plotted in gray. The average force–distance curve is highlighted by a thick continuous line. The images were captured using ArrowUHF AuD cantilevers. Experimental parameters: f = 745 kHz; k = 8.3 N m−1; Q = 4.5; A0 = 170 pm; Asp = 100 pm. The negatively charged AFM tips used for the research in this article were NanoWorld Arrow-UHFAuD ultra-high frequency AFM probes.
Figure 3 from Diana M. Arvelo et al. 2024 “Interfacial water on collagen nanoribbons by 3D AFM”
Interfacial liquid water structure on collagen provided by a negatively charged tip. (a) 2D force maps (x, y) of the interfacial water structure in the gap region. The map is obtained in a 300 mM KCl solution. The force–distance curves in the bottom of the image are obtained from the top panel. (b) 2D force maps (x, y) of the interfacial water structure in the overlap region. The force–distance curves in the bottom of the image are obtained from the top panel. (c) Statistics of d1 and d2 distances measured from several collagen–water interfaces. In the bottom panels of (a) and (b), the individual force–distance curves from the bottom panels of (a) and (b) are plotted in gray. The average force–distance curve is highlighted by a thick continuous line. The images were captured using ArrowUHF AuD cantilevers. Experimental parameters: f = 745 kHz; k = 8.3 N m−1; Q = 4.5; A0 = 170 pm; Asp = 100 pm.

*Diana M. Arvelo, Clara Garcia-Sacristan, Enrique Chacón, Pedro Tarazona and Ricardo Garcia
Interfacial water on collagen nanoribbons by 3D AFM
Journal of Chemical Physics 160, 164714 (2024)
DOI: https://doi.org/10.1063/5.0205611

The article “Interfacial water on collagen nanoribbons by 3D AFM” by Diana M. Arvelo, Clara Garcia-Sacristan, Enrique Chacón, Pedro Tarazona and Ricardo Garcia is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third-party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit https://creativecommons.org/licenses/by/4.0/.

Structural and optical variation of pseudoisocyanine aggregates nucleated on DNA substrates

The ability to maximize the range of exciton transport while minimizing energy loss has significant implications for the design of future nanoscale light harvesting, optoelectronic, and sensing applications. *

One method of achieving this would be to densely pack dyes into strongly coupled aggregates such that excitations can be coherently delocalized through the partial or full length of the aggregate. *

Coherently coupled aggregates enable exciton migration over discreet spatial distances with near unitary quantum efficiency. As a result, controlled dye aggregation has long been studied by chemists as a method of tuning the photonic and physical properties of the dyes and pigments in light harvesting devices. An example of this coherent coupling phenomenon can be observed in the cyanine dye family and specifically the prototypical example, pseudoisocyanine (PIC) dye. *

In the article «Structural and optical variation of pseudoisocyanine aggregates nucleated on DNA substrates” Matthew Chiriboga, Christopher M Green, Divita Mathur, David A Hastman, Joseph S Melinger, Remi Veneziano, Igor L Medintz and Sebastián A Díaz show that DNA-nucleated PIC aggregates have properties which correlate to different molecular structures and are directed by changing the DNA scaffold. *

To achieve this, they formed PIC aggregates through heterogeneous nucleation by mixing dissolved PIC dye with various DNA nanostructures ranging from a rigid DX-tile to more flexible DNA duplex (dsDNA) or single strand DNA oligonucleotide (ssDNA). *

Although the aggregates Matthew Chiriboga et al formed required elevated excess of PIC dye relative to previously reported J-bits, they exhibited sharper and brighter fluorescence peaks as well as longer Ncoh. *

Therefore, the authors refer to aggregates formed by this approach as super aggregate (SA) in their article, though they note SA formed with different DNA substrates result in unique properties. *

Complementary circular dichroism (CD) and atomic force microscopy (AFM) characterizations were used to analyze the SA and both indicated distinctions in the way each substrate and subsequent dye aggregate incorporates the individual PIC molecules. *

To achieve high resolution imaging of nucleic acid nanostructures, the DNA is often deposited onto a mica substrate, where mica electrostatically binds the DNA. Once deposited onto the mica, the imaging can be done in a hydrated environment as there is no additional required dehydration or staining of the DNA, a particularly convenient advantage of AFM. *

The AFM imaging was performed under AC fast imaging mode (liquid) with NanoWorld  Ultra-Short Cantilevers (USC) for Fast/High-Speed AFM of the USC-F0.3-k0.3 AFM probe type.*

On a segment of freshly cleaved mica mounted to a magnetic puck, 15 μl of PIC-DANN solution was deposited immediately before measurement. A 25 μl droplet of imaging buffer was deposited on the AFM tip, then the AFM tip mount was lowered into the sample buffer to create a liquid ‘chamber’ for imaging. *

When introducing various DNA scaffolds for SA formation and subsequent AFM imaging, Matthew Chiriboga et al. observed significant changes in the aggregates structure. *

The AFM imaging highlighted the stark differences in aggregate formation resulting from the DNA substrates. *

To the author’s knowledge this is the first visualization of DNA-based PIC aggregates. Results from the field have been pointing towards fiber-like or nanotube-like networks of polymerized PIC as a structural model for aggregates suspended in solution. *

On the other hand, other AFM studies demonstrate that PIC aggregates formed on mica substrates adopt a leafy island morphology. *

Interestingly, Matthew Chiriboga et al. observe evidence of both PIC fibers as well as leafy islands that exhibit distinct growth patterns, again depending on the DNA substrate. *

Although this work contributes to the growing body of evidence that solution-based PIC aggregates form fibrous networks structures, the AFM measurements presented in the article highlight the multiplicity of PIC aggregation modes when introduced to DNA scaffolds. *

The results presented in the research article suggest modification of the DNA substrate results in significant changes to how the DNA and companion dye molecules are integrated into larger form PIC aggregates. *

Bearing in mind that the broader motivation for studying DNA based PIC aggregates is to integrate strongly coupled dyes onto modular DNA structural units, PIC SAs should be given due consideration as a versatile option. *

In fact, similar work is being done with other cyanine dyes where DNA template modification is used to switch between quenching and energy transfer. *

Ultimately this could be a path for the PIC SA and one which possibly leads towards applications in optical microcavities for quantum electrodynamical devices and optical switching, molecular plasmonics, biosensors, and light-harvesting arrays. *

Figure 6 from Matthew Chiriboga et al. “Structural and optical variation of pseudoisocyanine aggregates nucleated on DNA substrates”:Atomic Force Microscopy visualisation of pseudoisocyanine aggregates – super aggregates (SA) nucleated on DNA substrates. AFM visualizations of pseudoisocyanine (PIC) aggregates formed in the (A) AT, (B) dsDNA, and (C) ssDNA nanostructures. Each of the samples was formed immediately before measurement by mixing 160 μM PIC dye with 500 nM DNA normalized to the dye-labeled strand concentration (i.e. 320-fold excess). When the SA was formed using an AT DX-tile template (figures 6(A) the authors observed the formation of large and long rod-like aggregates with a relatively isotropic growth axis. This supports the hypothesis proposed by Yoa et al which suggested aggregation along preferential axis due to a preferred interaction between the PIC and the mica NanoWorld USC-F0.3-k0.3 AFM probes were used for the under AC fast imaging mode in liquid.
Figure 6 from Matthew Chiriboga et al. “Structural and optical variation of pseudoisocyanine aggregates nucleated on DNA substrates”:
AFM visualization of SA formations. AFM visualizations of PIC aggregates formed in the (A) AT, (B) dsDNA, and (C) ssDNA nanostructures. Each of the samples was formed immediately before measurement by mixing 160 μM PIC dye with 500 nM DNA normalized to the dye-labeled strand concentration (i.e. 320-fold excess).

*Matthew Chiriboga, Christopher M Green, Divita Mathur, David A Hastman, Joseph S Melinger, Remi Veneziano, Igor L Medintz and Sebastián A Díaz
Structural and optical variation of pseudoisocyanine aggregates nucleated on DNA substrates
Methods and Applications in Fluorescence (2023) 11 014003
DOI: https://doi.org/10.1088/2050-6120/acb2b4

The article “Active self-assembly of piezoelectric biomolecular films via synergistic nanoconfinement and in-situ poling” by Matthew Chiriboga, Christopher M Green, Divita Mathur, David A Hastman, Joseph S Melinger, Remi Veneziano, Igor L Medintz and Sebastián A Díaz is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third-party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit https://creativecommons.org/licenses/by/4.0/.

Intrinsically disordered regions in TRPV2 mediate protein-protein interactions

Transient receptor potential (TRP) ion channels are gated by diverse intra- and extracellular stimuli leading to cation inflow (Na+, Ca2+) regulating many cellular processes and initiating organismic somatosensation. *

Structures of most TRP channels have been solved. However, structural and sequence analysis showed that ~30% of the TRP channel sequences, mainly the N- and C-termini, are intrinsically disordered regions (IDRs). Unfortunately, very little is known about IDR ‘structure’, dynamics and function, though it has been shown that they are essential for native channel function. *

In the article “Intrinsically disordered regions in TRPV2 mediate protein-protein interactions”, Raghavendar R. Sanganna Gari, Grigory Tagiltsev, Ruth A. Pumroy, Yining Jiang, Martin Blackledge, Vera Y. Moiseenkova-Bell and Simon Scheuring imaged TRPV2 channels in membranes using high-speed atomic force microscopy (HS-AFM). *

The dynamic single molecule imaging capability of HS-AFM allowed the authors to visualize IDRs and revealed that N-terminal IDRs were involved in intermolecular interactions. Their work provides evidence about the ‘structure’ of the TRPV2 IDRs, and that the IDRs may mediate protein-protein interactions. *

In total, 1.5 µl of the TRPV2 reconstituted vesicles were deposited on a 1.5-mm2 freshly cleaved mica surface, which was glued with epoxy to the quartz sample stage. After 20–30 min incubation, the sample was gently rinsed with imaging buffer (20 mM Hepes, pH 8.0, 150 mM NaCl) and mounted in the HS-AFM fluid cell. All images in this study were taken using a HS-AFM operated in amplitude modulation mode using optimized scan and feedback parameters and lab-built amplitude detectors and free amplitude stabilizers. *

Short (8 µm) cantilevers (NanoWorld Ultra-Short Cantilevers for High-Speed AFM USC-F1.2-k0.15) with nominal spring constant of 0.15 N/m, resonance frequency of 0.6 MHz, and a quality factor of ∼1.5 in liquid were used. AFM probes were sharpened using oxygen plasma etching to obtain better resolution. *

Fig. 1 from “Intrinsically disordered regions in TRPV2 mediate protein-protein interactions” by Raghavendar R. Sanganna Gari et al. :TRPV2 reconstitution for HS-AFM analysis. b Overview HS-AFM images (Supplementary Movie 1) of TRPV2 (windmill-shaped molecules) in soy polar lipid membranes on mica (dark background areas). False color scale: 0–9 nm. The white oversaturated areas have a height of ~26 nm and represent likely non-ruptured small vesicles. NanoWorld-USC-F1.2-k0.15 AFM probes were used for the HS-AFM
Fig. 1 from “Intrinsically disordered regions in TRPV2 mediate protein-protein interactions” by Raghavendar R. Sanganna Gari et al. :
TRPV2 reconstitution for HS-AFM analysis.
b Overview HS-AFM images (Supplementary Movie 1) of TRPV2 (windmill-shaped molecules) in soy polar lipid membranes on mica (dark background areas). False color scale: 0–9 nm. The white oversaturated areas have a height of ~26 nm and represent likely non-ruptured small vesicles.
Fig. 1 from “Intrinsically disordered regions in TRPV2 mediate protein-protein interactions” by Raghavendar R. Sanganna Gari et al. : TRPV2 reconstitution for HS-AFM analysis. a Negative-stain EM of TRPV2 reconstituted into soy polar lipids at a lipid-to-protein ratio of 0.7. Protruding features (arrow) at the vesicle periphery and the strong contrast of the proteins in the vesicle in the negative-stain EM are indicative of inside-out reconstitution of the TRPV2 channels with the large cytoplasmic domains exposed to the outside of the vesicle. b Overview HS-AFM images (Supplementary Movie 1) of TRPV2 (windmill-shaped molecules) in soy polar lipid membranes on mica (dark background areas). False color scale: 0–9 nm. The white oversaturated areas have a height of ~26 nm and represent likely non-ruptured small vesicles. c Height distribution of TRPV2 above mica from (b). TRPV2 has a full height of 9.5 ± 0.1 nm above mica, in good agreement with the TRPV2 cryo-EM structure. Inset: Cryo-EM structure PDB 6U84 shown with the intracellular side up (as imaged by HS-AFM), membrane indicated in light gray. Short (8 µm) cantilevers (NanoWorld Ultra-Short Cantilevers for High-Speed AFM USC-F1.2-k0.15,) with nominal spring constant of 0.15 N/m, resonance frequency of 0.6 MHz, and a quality factor of ∼1.5 in liquid were used. AFM probes were sharpened using oxygen plasma etching to obtain better resolution. *
Fig. 1 from “Intrinsically disordered regions in TRPV2 mediate protein-protein interactions” by Raghavendar R. Sanganna Gari et al. :
TRPV2 reconstitution for HS-AFM analysis.
a Negative-stain EM of TRPV2 reconstituted into soy polar lipids at a lipid-to-protein ratio of 0.7. Protruding features (arrow) at the vesicle periphery and the strong contrast of the proteins in the vesicle in the negative-stain EM are indicative of inside-out reconstitution of the TRPV2 channels with the large cytoplasmic domains exposed to the outside of the vesicle. b Overview HS-AFM images (Supplementary Movie 1) of TRPV2 (windmill-shaped molecules) in soy polar lipid membranes on mica (dark background areas). False color scale: 0–9 nm. The white oversaturated areas have a height of ~26 nm and represent likely non-ruptured small vesicles. c Height distribution of TRPV2 above mica from (b). TRPV2 has a full height of 9.5 ± 0.1 nm above mica, in good agreement with the TRPV2 cryo-EM structure. Inset: Cryo-EM structure PDB 6U84 shown with the intracellular side up (as imaged by HS-AFM), membrane indicated in light gray.

 

*Raghavendar R. Sanganna Gari, Grigory Tagiltsev, Ruth A. Pumroy, Yining Jiang, Martin Blackledge, Vera Y. Moiseenkova-Bell and Simon Scheuring
Intrinsically disordered regions in TRPV2 mediate protein-protein interactions
Communications Biology volume 6, Article number: 966 (2023)
DOI: https://doi.org/10.1038/s42003-023-05343-7

Please follow this external link to read the full article: https://rdcu.be/dnNba

The article “Phosphorylation of phase-separated p62 bodies by ULK1 activates a redox-independent stress response” by Raghavendar R. Sanganna Gari, Grigory Tagiltsev, Ruth A. Pumroy, Yining Jiang, Martin Blackledge, Vera Y. Moiseenkova-Bell and Simon Scheuring is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third-party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit https://creativecommons.org/licenses/by/4.0/.