Influence of oxygen deficiency and the role of specific amino acids in cryopreservation of garlic shoot tips
© Subbarayan et al.; licensee BioMed Central. 2015
Received: 21 January 2015
Accepted: 20 May 2015
Published: 28 May 2015
Garlic has lost its ability to form seeds in the course of its domestication. Therefore, the germplasm storage via cryopreservation is increasingly applied. The progression of the various steps within the cryopreservation procedure is accompanied by declining survival rates of the explants. Much of the recent work on cryo-stress has been focussed on osmotic and cold stress components. However, two decades after invention of garlic cryopreservation, the function of metabolites and oxygen in and around the cryopreserved tissues is still largely obscure.
In this study, hypoxia was characterized in cryopreservation of garlic with oxygen sensors and amino acid metabolism. Furthermore, malondialdehyde, soluble sugars and ammonium were quantified to demonstrate the influence of cryo-stress in declining survival rates.
To better understand the possible reasons for a reduction in the survival rate at the subsequent steps of cryopreservation, the concentration of amino acids, ammonium, γ-aminobutyric acid (GABA), soluble sugars, malondialdehyde (MDA), and oxygen were measured in garlic shoot tips undergoing cryopreservation. Using microsensors, a very low oxygen concentration (<0.1 μM) was detected within the central meristem region of the shoot apex. When apices were immersed in cryoprotectant solution, the well-oxygenated peripheral regions (foliage leaf bases) became likewise hypoxic within a few minutes, probably resulting from strongly restricted gaseous diffusion.
Tissue level oxygen measurements supported the occurrence of hypoxia while biochemical analysis indicated adaptive responses, in particular the modulation in alanine and glutamate metabolism. The possible role of serine and glycine metabolism during cryopreservation is also discussed.
Cryopreservation, the storage of germplasm at ultra-low temperature in liquid nitrogen, was successfully applied to many plant species [1, 2]. It is considered as an important tool also for long-term conservation of garlic. Shoot apices are the preferred tissue source for garlic cryopreservation . Meristem cells located in the apical region of shoot tips have higher tolerance towards cryo-storage protocols . Among several cryopreservation protocols, vitrification is characterized by initiating glass transition of the cellular water content. This avoids ice-caused damage during freezing. The plant vitrification solution PVS3 , is the most effective for garlic allowing higher regrowth rates than other vitrification solutions . The genebank at IPK comprises one of the largest cryo-collections worldwide with more than 1500 accessions, the number of garlic accessions amounting to 106. In general, regrowth of plants after cryopreservation varies, and resulting rates range from poor to excellent. Understanding how and why germplasm does or does not survive cooling is needed to further improve plant cryopreservation . In nature, plants are not exposed to such low temperatures as during freezing. Cryogenically-stored plant tissues undergo dehydration, osmotic, oxidative and cold stress conditions in the course of the cryopreservation procedure . While lipid peroxidation was well documented as occurring under cryo-stress by measuring malondialdehyde (MDA) accumulation , none of these studies has discussed the potential role of the amino acid metabolism under the specific physiological conditions occurring during cryopreservation. Furthermore, the oxygen level in the cryopreservation solutions as well as in the tissue exposed to dehydration solution was not yet recorded, while the osmotic effects were demonstrated . Within a plant tissue or organ, cells are frequently challenged with limited levels of oxygen supply due to changes in the external environment or high rates of cellular metabolism [10, 11]. This in turn causes partial or complete inhibition of mitochondrial respiration, also known as hypoxia or anoxia, respectively.
To gain insights into deleterious effects of cryo-stress, we investigated the amino acid content in the tissue in the course of the cryopreservation procedure. We found an altered modulation in alanine and glutamate metabolism at dehydration step that may indicate hypoxic stress in the tissue . Using microsensors, O2 maps were generated for garlic shoot apices providing evidence for severe O2 deficiency in shoot apices. We demonstrated that some of the metabolites are specifically regulated during dehydration and after storage in liquid nitrogen (LN). Furthermore, accumulation of MDA was confirmed to play an important role in cryo-stress. On the basis of our results, the role of amino acid metabolism in response to hypoxic stress in cryopreserved garlic shoot apices is discussed.
Bulbils were treated with 70 % ethanol for 10–20 s, and then shaken in sodium hypochlorite solution (3 % active chlorine [Carl Roth GmbH & Co., Karlsruhe, Germany]) with three drops of Tween 20 [Riedel-de Haën AG, Seelze, Germany] for 25 min. Bulbils were then rinsed five times with sterile water. Shoot apices 1–2 mm in diameter and 3 mm long were excised from bulbils .
Cryogenic treatment procedure
Fresh: Shoot apices were excised from garlic bulbils or in vitro plantlets.
Preculture: Thereafter, the shoot apices were precultured in standard medium containing 10 % sucrose overnight.
Dehydration: A dehydration step followed, in which the explant was treated with loading solution (18.4 % w/v glycerol and 13.7 % w/v sucrose in liquid MS medium) for 20 min and then dehydrated using PVS3 (50 % w/v glycerol and 50 % w/v sucrose in liquid MS) for 120 min at 25 °C.
Storage and recovery (LN storage): Finally, the vitrified explants were rapidly cooled down and kept in liquid nitrogen for 60 min, then thawed at 40 °C in a water bath for 2 min, subsequently immersed in washing solution (1.2 M sucrose in liquid MS medium) for 10 min at 25 °C and finally plated onto solid MS medium.
Determination of survival rates
Shoot apices from each of the above cryopreservation steps were propagated on standard medium for fifteen days and tested for survival rates by counting shoot apices that exhibited visible growth (F and P) or pertained greenish and swollen (D and LN). The data were calculated to display the results in percent of survival ± standard error.
Determination of oxygen concentration in plant tissue and cryopreservation solutions using an optical microsensor
Oxygen concentration profiles across the tissue were measured using a needle-type microsensor according to Rolletschek et al.  at ambient atmosphere and incubation in cryopreservation solutions (vitrification, loading and washing solutions). Briefly, an oxygen-sensitive optodes (Presens GmbH, Regensburg, Germany) was inserted and moved forward into the tissue using a micromanipulator, and the oxygen concentration was measured at 100 μm intervals along the horizontal pathway of sensor penetration. Before and after each analysis the sensor was calibrated using premixed gases (0 and 21 kPa oxygen, balanced by N2). Measurement of oxygen concentration in cryopreservation solutions was done by inserting the microsensor into the gently stirred solution for at least 1 min.
Fresh and dry weight comparison
Quantification of malondialdehyde (MDA) by HPLC
Four shoot apices (40–80 mg fresh weight) were collected from each step of the cryopreservation procedure. MDA was measured according to Lepage et al.  with a few modifications. Shoot apices were added to 1 ml of 5 % TCA solution (trichloroacetic acid) and 0.1 ml of 20 mM BHT (butylated hydroxytoluene) and incubated at 95 °C for 30 min. After centrifugation at 1000 g for 10 min, an equal volume of 0.25 % thiobarbituric acid (TBA) was added to the supernatant. The reaction mixture was heated at 95 °C for 30 min for the formation of MDA-TBA complex. Aliquots of samples were transferred into microvials, which were placed in an autosampler and were automatically injected into the HPLC reverse-phase system (Hypersil ODS C18 with 5 μm particle size, 4.6 × 100 mm). The MDA (TBA)2 adduct was eluted using an isocratic mobile phase consisting of 100 % methanol at a flow rate of 0.2 ml/min at 25 °C. MDA was detected by fluorescence at excitation 532 nm and emission 553 nm. Quantification was done using an external standard of 1,1,3,3-tetraethoxypropane (Sigma, St Louis, MO) prepared using the same method as for the samples. Chromatograms were analyzed by EMPOWER (Waters, Germany).
Determination of soluble sugars
Three shoot apices (30–40 mg fresh weight) were collected from each step of the cryopreservation procedure. The samples were extracted with 80 % ethanol at 60 °C and centrifuged at 14,000 g for 15 min. Soluble sugars (glucose, fructose and sucrose) were measured using an enzyme-coupled assay according to Hajirezaei et al. .
Extraction and determination of free amino acids and ammonium
Plant material (30–40 mg fresh weight) was incubated for 60 min at 80 °C in 0.5 ml of 80 % ethanol and thereafter centrifuged for 10 min at 14,000 rpm and 4 °C. Supernatant was evaporated to dryness, re-suspended in purest water and analysed by ultra performance liquid chromatography (UPLC). Prior to UPLC analysis samples were derivatized using the fluorescing reagent AQC (6-aminoquinolyl-N-hydroxysuccinimidylcarbamat). 3 mg of self-made AQC (IPK, Germany) was dissolved in 1 ml acetonitrile and incubated exactly for 10 min at 55 °C. The prepared reagent was stored at 4 °C and used up within four weeks. For derivatization of the sample 0.01 ml of the prepared reagent solution was used for each sample which contained 0.8 ml of a buffer (0.2 M, pH 8.8) and 0.01 ml of the supernatant. Separation of soluble amino acids was achieved by a newly developed UPLC-based method using ultra pressure reversed phase chromatography (AcQuity H-Class, Waters GmbH, Germany). The UPLC system consisted of a quaternary solvent manager, a sample manager-FTN, a column manager and a fluorescent detector (PDA eλ Detector). The separation was carried out on a C18 reversed phase column (ACCQ Tag Ultra C18, 1.7 μm, 2.1x100 mm) with a flow rate of 0.7 ml/min and duration of 10.2 min. The column was heated at 50 °C during the whole run. The detection wavelengths were 266 nm for excitation and 473 nm for emission. The gradient was accomplished with four solutions prepared from two different buffers purchased from Waters GmbH (eluent A concentrate and eluent B for amino acid analysis, Waters GmbH, Germany and LCMS water, Geyer GmbH, Germany). Eluent A was pure concentrate, eluent B was a mixture of 90 % LCMS water and 10 % eluent B concentrate, eluent C was pure concentrate of eluent B and eluent D was LCMS water. The column was equilibrated with eluent A (10 %) and eluent C (90 %) for at least 30 min.
Responses of shoot apices to cryogenic protocol
Oxygen mapping across shoot apices of garlic
Fine fiber optic microsensors (tip diameter ~ 30 μm) were used to measure O2 concentration across fresh shoot apices. The microsensors penetrated parallel to the bulb base targeting the meristem. The O2 level immediately declined to 20 % atmospheric saturation (~50 μM or 4 kPa oxygen) when the microsensor penetrated the foliage leaf bases, whereas a sharp decrease was detected at meristem region (Fig. 1). Within the central meristematic zone the oxygen concentration reached almost zero (<0.1 μM), indicating strong hypoxic conditions.
Submergence of shoot apices in cryopreservation solutions causes significant reductions in endogenous oxygen levels
Changes of malondialdehyde and soluble sugars during cryopreservation
Alteration of soluble amino acids during cryopreservation
The total amount of amino acids did not show any remarkable changes throughout the whole cryo-procedure (Fig. 5), however they were slightly lower (P < 0.05, paired Student’s t-test) at the dehydration step. Interestingly, 2.5 fold of alanine was increased during the dehydration step and it became the major amino acid in dehydrated garlic tips. Total amino acids as well as alanine and glutamic acid showed no significant change in response to LN storage whereas glycine and serine (P < 0.05, paired Student’s t-test) were increased after LN storage. Throughout the cryo-procedure, the level of GABA was unchanged (data not shown). In addition, the concentration of ammonium increased in the cryopreserved explants compared to the fresh ones whereas preculture step showed a notable increase (P < 0.05, paired Student’s t-test).
Overall, the differences in amino acid accumulation in cryopreserved explants lay with the accumulation of alanine under hypoxia and glycine with cold stress in the dehydration and LN storage, respectively.
Plant shoot apices do not survive exposure to liquid nitrogen without cryoprotective treatments . The vitrification-based cryopreservation improves survival of plant material by increasing cell sap viscosity and preventing formation of harmful intracellular ice crystals, but it also produces complex stresses . In general, the success rate of cryopreservation is mainly depending on the resistance of explants towards abiotic stresses underlying cryopreservation. Dehydration of samples with PVS2-based vitrification induces mainly chemical cytotoxicity (DMSO, Ethylene glycol) and additionally osmotic stress (glycerol, sucrose), while PVS3 induces osmotic stress .
Alterations in alanine and glutamate metabolism in garlic apices observed in this study are probably related to the induction of O2 deficiency. In previous studies alanine was shown to be the characteristic amino acid accumulated when plants Arabidopsis thaliana  or Hordeum vulgare  were suffering from hypoxia. It has been suggested that alanine acts as a non-toxic form of carbon and nitrogen storage during hypoxia, since it firstly allows glycolysis to run under severe O2 limitation, secondly it is easily reconverted back to pyruvate (Fig. 7) and/or participates in the synthesis of other amino acids during the recovery period when returning to normoxia [25, 27], and finally its synthesis does not induce shifts in pH. We have demonstrated here that the increase in alanine levels (i.e. alanine synthesis) occurs mainly readily during the dehydration step, which obviously induces hypoxia inside the shoot apex. The accumulation of alanine represents a marker for hypoxia, and alanine synthesis can be used for the diagnosis of hypoxia-related injuries . The relationship concluded here between alanine accumulation and hypoxia in garlic can quite possibly be extended to other plant species in which cryopreservation is being practiced.
After dehydration, alanine became the predominant amino acid (21 %). Similar results were reported by Limami et al. , who detected that after hypoxic stress alanine replaced asparagine as the predominant amino acid. Moreover, glutamate concentration decreased during the hypoxic dehydration step, which is consistent with its role as precursor of alanine synthesis .
MDA is a marker of peroxidation caused by reactive oxygen species (ROS). In the present study MDA content increased significantly in the preculture, reached its maximum during dehydration and slightly increased after LN storage (Fig. 4) indicating that the generation of ROS is highest after the hypoxic dehydration step. This finding is consistent with other cryopreservation studies . The presented results show that the increase of MDA is associated with a decrease of the regeneration potential. This corresponds to the observations of Verleysen et al.  indicating that MDA is a parameter for loss of viability due to cold stress and chilling injury. This is reported to have accumulated in plant tissues under hypoxic conditions  and MDA was enhanced maximally of 1.4 fold during dehydration. Altogether, the cryo-stress obviously had deleterious effects on the shoot apices, because the survival rate determined after cryopreservation was much lower (47.1 %) and the explants grow slower than without these influences (Fig. 2). Apart from the stress markers, soluble sugars have important roles in different protection systems, such as in the reactive oxygen species balance . An increasing pattern was observed in soluble sugars throughout the cryoprevation protocol (Fig. 5). Indeed, sucrose is known to be a very important metabolite that is involved in different pathways and has been observed to be a key molecule in determining the ability of plants to be cryopreserved [32, 33]. Furthermore, sucrose plays an important role as cryoprotectant . The tissue level sucrose enhancement across cryo-steps might also be related to the increased supplementation of sucrose concentration in cryopreservation solutions.
An increase in total amino acid content has already been reported for many plant species submitted to O2 deficiency . In the present study this was not observed during the dehydration step (Fig. 5), which might be due to the short incubation time of only two hours. As vitrification avoids ice crystal formation and protects plant cells from freezing, this process could not be considered as proper cold stress. But vitrification may have features in common with cold stress. The concentration of the amino acids glycine and serine increased after LN storage which resembles changes occurring after cold stress . Serine has been demonstrated to accumulate in Lolium perenne at low temperature conditions .
As serine is closely linked to glycine formation , the concurrent accumulation of glycine was probably measured after LN storage. Furthermore glycine accumulation at this step can be related to a change in lipid peroxidation indicated by a stagnation of the MDA concentration .
This study analyzed the stress factors involved in the different steps of vitrification and documented the occurrence of hypoxia in addition to osmotic stress during dehydration. This might be of high relevance in regard to the reduced regeneration rates after cryopreservation and could lead to improvement of the protocols. Understanding the mechanisms of cryo-stress will also help us to understand how plants respond to various types of abiotic stresses in nature. Further studies are necessary that examine hypoxia during dehydration at molecular level.
This research has been supported by IPK and Leibniz-DAAD fellowships (A/11/94467), which are greatly acknowledged. We extend our thanks to Doris Buechner and Wally Wendt for their technical assistance.
- Benson EE. Cryopreservation of phytodiversity: a critical appraisal of theory & practice. Crit Rev Plant Sci. 2008;27(3):141–219.Google Scholar
- Reed BM. Cryopreservation—practical considerations. In: Plant cryopreservation: a practical guide. Springer; 2008; p. 3–13.Google Scholar
- Keller ERJ, Senula A. Micropropagation and cryopreservation of garlic (Allium sativum L.). In: Protocols for Micropropagation of Selected Economically-Important Horticultural Plants. Springer; 2013; p. 353–368.Google Scholar
- Helliot B, Panis B, Poumay Y, Swennen R, Lepoivre P, Frison E. Cryopreservation for the elimination of cucumber mosaic and banana streak viruses from banana (Musa spp.). Plant Cell Rep. 2002;20(12):1117–22.Google Scholar
- Nishizawa S, Sakai A, Amano Y, Matsuzawa T. Cryopreservation of asparagus (Asparagus officinalis L.) embryogenic suspension cells and subsequent plant regeneration by vitrification. Plant Sci. 1993;91(1):67–73.Google Scholar
- Kim HH, Cho EG, Baek HJ, Kim CY, Keller ERJ, Engelmann F. Cryopreservation of garlic shoot tips by vitrification: effects of dehydration, rewarming, unloading and regrowth conditions. CryoLetters. 2004;25(1):59–70.Google Scholar
- Uchendu EE, Shukla MR, Reed BM, Saxena PK. Melatonin enhances the recovery of cryopreserved shoot tips of American elm (Ulmus americana L.). J Pineal Res. 2013;55(4):435–42.Google Scholar
- Whitaker C, Beckett R, Minibayeva F, Kranner I. Production of reactive oxygen species in excised, desiccated and cryopreserved explants of Trichilia dregeana Sond. S Afr J Bot. 2010;76(1):112–8.Google Scholar
- Grapin A, Dorion N, Verdeil JL, Escoute J. Histo-cytological changes in Pelargonium apices during the cryopreservation process: effect of the osmotic agent chosen for the preculture step. Acta Hort. 2007;760:195–202.Google Scholar
- Hsu FC, Chou MY, Peng HP, Chou SJ, Shih MC. Insights into hypoxic systemic responses based on analyses of transcriptional regulation in Arabidopsis. PLoS One. 2011;6(12), e28888.View ArticleGoogle Scholar
- Borisjuk L, Rolletschek H. The oxygen status of the developing seed. New Phytol. 2009;182(1):17–30.Google Scholar
- Limami AM, Glévarec G, Ricoult C, Cliquet JB, Planchet E. Concerted modulation of alanine and glutamate metabolism in young Medicago truncatula seedlings under hypoxic stress. J Exp Bot. 2008;59(9):2325–35.Google Scholar
- Murashige T, Skoog F. A revised medium for rapid growth and bio assays with tobacco tissue cultures. Physiol Plantarum. 1962;15(3):473–97.Google Scholar
- Bhojwani SS. In vitro propagation of garlic by shoot proliferation. Sci Hortic-Amsterdam. 1980;13(1):47–52.Google Scholar
- Rolletschek H, Stangelmayer A, Borisjuk L. Methodology and significance of microsensor-based oxygen mapping in plant seeds–an overview. Ah S Sens. 2009;9(5):3218–27.Google Scholar
- Lepage G, Munoz G, Champagne J, Roy CC. Preparative steps necessary for the accurate measurement of malondialdehyde by high-performance liquid chromatography. Anal Biochem. 1991;197(2):277–83.Google Scholar
- Hajirezaei MR, Takahata Y, Trethewey RN, Willmitzer L, Sonnewald U. Impact of elevated cytosolic and apoplastic invertase activity on carbon metabolism during potato tuber development. J Exp Bot. 2000;51 Suppl 1:439–45.Google Scholar
- Keller ERJ. Cryopreservation of Allium sativum L. (Garlic). In: Towill LE, Bajaj YPS, editors. Biotechnology in Agriculture and Forestry. Vol 50 Cryopreservation of Plant Germplasm II. Heidelberg: Springer; 2002. p. 37–47.View ArticleGoogle Scholar
- Ren L, Zhang D, Jiang XN, Gai Y, Wang WM, Reed BM, et al. Peroxidation due to cryoprotectant treatment is a vital factor for cell survival in Arabidopsis cryopreservation. Plant Sci. 2013;212:37–47.Google Scholar
- Kim HH, Lee SC. Personalisation of droplet-vitrification protocols for plant cells: An systematic approach to optimising chemical and osmotic effects. CryoLetters. 2012;33(4):271–9.Google Scholar
- Colmer TD, Flowers TJ. Flooding tolerance in halophytes. New Phytol. 2008;179(4):964–74.Google Scholar
- Nishiuchi S, Yamauchi T, Takahashi H, Kotula L, Nakazono M. Mechanisms for coping with submergence and waterlogging in rice. Rice J. 2012;5(1):1–14.Google Scholar
- Greve TM, Borum J, Pedersen O. Meristematic oxygen variability in eelgrass (Zostera marina). Limnol and Oceanogr. 2003;48(1):210–6.Google Scholar
- Oliveira HC, Salgado I. Role of plant mitochondria in nitric oxide homeostasis during oxygen deficiency. In: Nitric Oxide in Plants: Metabolism and Role in Stress Physiology. Springer International Publishing; 2014. p. 57–74.Google Scholar
- Miyashita Y, Dolferus R, Ismond KP, Good AG. Alanine aminotransferase catalyses the breakdown of alanine after hypoxia in Arabidopsis thaliana. Plant J. 2007;49(6):1108–21.Google Scholar
- Rolletschek H, Melkus G, Grafahrend-Belau E, Fuchs J, Heinzel N, Schreiber F, et al. Combined noninvasive imaging and modeling approaches reveal metabolic compartmentation in the barley endosperm. Plant Cell. 2011;23(8):3041–54.Google Scholar
- Oliveira HC, Sodek L. Effect of oxygen deficiency on nitrogen assimilation and amino acid metabolism of soybean root segments. Amino Acids. 2013;44(2):743–55.View ArticleGoogle Scholar
- Uchendu EE, Muminova M, Gupta S, Reed BM. Antioxidant and anti-stress compounds improve regrowth of cryopreserved Rubus shoot tips. In Vitro Cell Dev-Pl. 2010;46(4):386–93.Google Scholar
- Verleysen H, Samyn G, Van Bockstaele E, Debergh P. Evaluation of analytical techniques to predict viability after cryopreservation. Plant Cell Tiss Org. 2004;77(1):11–21.Google Scholar
- Ershova AN, Popova NV, Berdnikova OS. Production of reactive oxygen species and antioxidant enzymes of pea and soybean plants under hypoxia and high CO2 concentration in medium. Russ J Plant Physl. 2011;58(6):982–90.Google Scholar
- Couée I, Sulmon C, Gouesbet G, El Amrani A. Involvement of soluble sugars in reactive oxygen species balance and responses to oxidative stress in plants. J Exp Bot. 2006;57:449–59.Google Scholar
- Zhu GY, Geuns JMC, Dussert S, Swennen R, Panis B. Change in sugar, sterol and fatty acid composition in banana meristems caused by sucrose-induced acclimation and its effects on cryopreservation. Physiol Plant. 2006;128:80–94.Google Scholar
- Ramon M, Geuns J, Swennen R, Panis B. Polyamines and fatty acids in sucrose precultured banana meristems and correlation with survival rate after cryopreservation. Cryo Letters. 2002;23:345–52.Google Scholar
- Folgado R, Sergeant K, Renaut J, Swennen R, Hausman JF, Panis B. Changes in sugar content and proteome of potato in response to cold and dehydration stress and their implications for cryopreservation. J Proteomics. 2014;98:99–111.Google Scholar
- Van Dongen JT, Fröhlich A, Ramírez-Aguilar SJ, Schauer N, Fernie AR, Erban A, et al. Transcript and metabolite profiling of the adaptive response to mild decreases in oxygen concentration in the roots of Arabidopsis plants. Ann Bot-London. 2009;103(2):269–80.Google Scholar
- Chawade A, Lindén P, Bräutigam M, Jonsson R, Jonsson A, Moritz T, et al. Development of a model system to identify differences in spring and winter oat. PLoS One. 2012;7(1), e29792.View ArticleGoogle Scholar
- Draper S. Amino acid changes associated with low temperature treatment of Lolium perenne. Phytochemistry. 1972;11(2):639–41.View ArticleGoogle Scholar
- Bauwe H, Kolukisaoglu Ü. Genetic manipulation of glycine decarboxylation. J Exp Bot. 2003;54(387):1523–35.Google Scholar
- Nayyar H, Chander K, Kumar S, Bains T. Glycine betaine mitigates cold stress damage in chickpea. Agron Sustain Dev. 2005;25(3):381.Google Scholar
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.