The effect of desulfurization on the postharvest quality and sulfite metabolism in pulp of sulfitated “Feizixiao” Litchi (Litchi chinensis Sonn.) fruits

Abstract The residual sulfite caused by sulfur fumigation (SF) is a hazard to health and influenced the export trade of litchi. Desulfurization (DS) is a valid chemical method to reduce the residual sulfite. However, the effect of DS on fumigated litchi has not been studied at physiological and molecular level. This study was aimed to evaluate the effect of DS (SF plus 3% desulfurizer) on the postharvest quality, sulfite residue, and the sulfite metabolism in sulfitated “Feizixiao” litchi during the 4°C storage. Results indicated that the DS promoted the color recovery of sulfitated litchi and achieved an effect similar to SF on controlling rot and browning. DS recovered the water content and respiration rate of sulfitated litchi pericarp. Thus, DS improves commodity properties of sulfitated litchi. Moreover, DS greatly reduced sulfite residue especially in pulp and ensured the edible safety of sulfitated litchi. The activities of sulfite oxidase, sulfite reductase, serine acetyltransferase, and O‐acetylserine(thiol) lyase in pulp increased after SF but fell down after DS while the expressions of their encoding genes decreased after SF but then rallied after DS. These results indicated the key role of these enzymes in sulfite metabolism after SF and DS changed the sulfite metabolism at both enzymatic and transcriptional level. It could be concluded that DS used in this study was an effective method for improving the color recovery and ensuring the edible safety of sulfitated litchi by not only chemical reaction but also both of enzymatic and transcriptional regulation.

Sulfur dioxide was used widely as preservative and sanitizing agent to prevent spoilage by microorganisms in fruit juices, syrups, wine or vinegar, dehydrated and dried fruits, vegetables, traditional chinese medicine products, table grapes, kiwifruits, blueberries, litchi fruits, and other fresh fruits due to its strong oxidability, low cost, volatileness, operability, and excellent sterilization ability.
It was also used as an antioxidant and inhibitor of enzyme-catalyzed oxidative discoloration and of nonenzymatic browning during preparation, storage, or distribution of many food products (Joslyn & Braverman, 1954). The utilization of SO 2 had been applied in the marketing of grapes since 1920s and succeeded in commercial preservation of litchi fruits in 1980s (Swarts, 1985). SO 2 fumigation caused the red color to be bleached to yellow, which was slowly and partially restored to pink. SO 2 interacted with the membranes, making the rind pliable and leaky to solutes. In addition, SO 2 directly reacted with anthocyanins and inhibits nonenzymatic formation of colorless quinone-sulfite complexes and enzymatic browning by inactivation of PPO (Jiang et al., 2006). However, due to its reaction with water thereby forming sulfite (the main form of sulfur residue), SO 2 showed toxicity for organisms (Baillie et al., 2016). The approval from Europe, Australia, and Japan for SO 2 was likely to be withdrawn due to concerns over sulfur residues in fumigated litchi fruits (Jiang et al., 2006). Therefore, the solution of residual sulfite is the key to break the export barrier of litchi fruits.
In order to restore the color and reduce the sulfur residue, proper desulfurization was often carried out after SO 2 fumigation. In addition to the chemical desulfurization methods, the enzymatic degradation of sulfite also plays an important role in reducing sulfur residue. SO 2 gas entered the cell apoplast space and formed sulfite (SO 3 2− ) with water. Sulfite is a toxic metabolite that can break disulfide bridges, which is termed sulfitolysis; sulfite inhibits numerous enzymes, and it can attach to aldehydes forming hydroxysulfonates, which are metabolic inhibitors (Hänsch et al., 2006). Therefore, its fast removal by oxidation to nontoxic sulfate is a means to protect the cell against excess of sulfite derived from SO 2 . Plants usually relieve the SO 2 stress in three ways: regulating stomatal conductance to control the amount of SO 2 entering the cell, oxidizing the SO 3 2− into SO 4 2− by sulfite oxidase (SO) in peroxisome, and then was stored in the vacuoles or converting the SO 3 2− into a thiol or other sulfur-containing compounds via a reduction pathway (Aghajanzadeh, Hawkesford, & Kok, 2016;Baillie et al., 2016;Chao et al., 2014 However, these studies focused on the effects of low concentration SO 2 on sulfur metabolism in plant leaves or roots (Aghajanzadeh et al., 2016;Brychkova, Yarmolinsky, Fluhr, & Sagi, 2012;Randewig et al., 2014), but few reports on the effects of high concentration of SO 2 on sulfur metabolism in sulfitated fruits especially the edible parts such as pulp. In this study, we performed a desulfurization (DS) treatment to reduce the residual sulfite and investigated its effect on the quality and postharvest physiology of sulfitated litchi fruits.
In addition, we explored the key regulation steps of sulfur detoxification in litchi pulp by comparing the difference at enzymatic and transcriptional level of five enzymes related to sulfite degradation between fumigated and desulfurization litchi fruits.

| Plant materials, postharvest treatments, and samplings
The "Feizixiao" litchi fruits used for the present study were grown following commercial cultivation practices in the same orchard in Conghua district, Guangzhou, China. Commercial mature "Feizixiao" litchi fruits with no damage and no disease were harvested.

| Color measurement
Fruit color was measured by a color analyzer (KONICA MINOLTA CR-300, Japan). The red to green was expressed as +a* to −a*, yellow to blue was expressed as +b* to −b*, brightness was expressed as L*, and the color index was expressed as CI.

| Determination of browning index and rotting rate
Three bags were randomly selected for examining rotting rate. Ten fruits were randomly selected from each bag, and totally thirty fruits

| Peel electroconductibility measurement
The cell membrane permeability of peel was measured according to a method reported by Duan et al. (2004), with some modifications.
Three peel disks (diameter 0.5 cm) were punched from each fruit, and peel disks of ten fruits were collected and washed three times by deionized water.

| Determination of water content in peel
One gram of the peel which was separated from the pulp and cleaned was added into a rapid moisture meter (Satorious MA150).
The water content in peel of all samplings at each time point was measured in three biological repeats.

| Measurement of respiration rate
Twenty fruits were randomly selected and sealed in a hermetically sealed box for 2 hr at 4°C. The CO 2 concentration in the gas was

| Determination of sulfite residue
The sulfite residue was determined according to a previously re-  (4) and (5):

| Determination of activity of SO
Sample powder (400 mg) grinded by liquid nitrogen was added into 1.6 ml precooled extraction buffer (100 mM Tris-acetic acid, pH 7.5, containing 10 mM DTT, 10 mM KCl, 1 mM EDTA, 1 mM EGTA, and 10% glycerol), fully mixed. After a centrifugation at 15,700 g and 4°C for 20 min, the supernatant was used as crude enzyme extract for assaying activity. Activity of SO was assayed according to a previously reported method (Xia et al., 2012). The absorbance of solution at 420 nm was recorded every minute from 0 to 5 min. The concentration of K 3 Fe(CN) 6 was calculated by a standard curve. The amount of enzyme required for reduction of 2 μmol K 3 Fe(CN) 6 every minute was calculated as one activity unit (U).

| Determination of activity of APR
Sample powder (400 mg) grinded by liquid nitrogen was added into 12 ml precooled extraction buffer (100 mM Tris-HCl, pH 7.7, containing 10 mM Na 2 SO 3 , 0.5 mM AMP, 10 mM DTT, 5 mM EDTA-Na 2 , 10 mM cysteine, 1% Triton X-100, and 2% PVP40), fully mixed. After a centrifugation at 15,700 g and 4°C for 20 min, the supernatant was used as crude enzyme extract for assaying activity. Activity of APR was assayed according to a previously reported method (Scheerer et al., 2010). The absorbance of solution at 420 nm was recorded every minute from 0 to 5 min. The concentration of K 3 Fe(CN) 6 was calculated by a standard curve. The amount of enzyme required for reduction of 1 μmol K 3 Fe(CN) 6 every minute was calculated as one activity unit (U).

| Determination of activity of SiR
Sample powder (400 mg) grinded by liquid nitrogen was added into 1.6 ml precooled extraction buffer (100 mM Tris-acetic acid, pH 7.5, containing 10 mM DTT, 10 mM KCl, 1 mM EDTA and 1 mM EGTA and 10% glycerol), fully mixed. After a centrifugation at 15,700 g and 4°C for 20 min, the supernatant was used as crude enzyme extract for assaying activity. Activity of SiR was assayed according to a previously reported method (Ostrowski et al., 1989). The absorbance of solution at 340 nm was recorded every minute from 0 to 5 min.
Reaction buffer (2.8 ml) mixed with 200 μL deionized water without enzyme extract, and NADPH was setted as control. The concentration of NADPH was calculated by a standard curve. The amount of enzyme required for oxidation of 1 μmol NADPH per minute was calculated as one activity unit (U).

| Determination of activity of SAT
Sample powder (400 mg) grinded by liquid nitrogen was added into 1.6 ml precooled extraction buffer (100 mM Tris-acetic acid, pH 7.5, containing 10 mM DTT, 10 mM KCl, 1 mM EDTA and 1 mM EGTA and 10% glycerol), fully mixed. After a centrifugation at 15,700 g and 4°C for 20 min, the supernatant was used as crude enzyme extract for assaying activity of SAT. Activity of SAT was assayed according to a method reported by Randewig et al. (2014). The absorbance of solution at 412 nm was recorded every minute from 0 to 5 min. The concentration of DTNB was calculated by a standard curve. Reaction buffer without l-serine was setted as control. The amount of enzyme required for reduction of 1 μmol DTNB every minute was calculated as one activity unit (U). (

| Determination of activity of OAS-TL
Sample powder (400 mg) together with 20 mg PVP was grinded by liquid nitrogen and then was added into 1.26 ml precooled extraction buffer (100 mM Tris-HCl, pH 8.0, 100 mM KCl, 20 mM MgCl 2 , 1% Tween 80, and 10 mM DTT), fully mixed. After a centrifugation at 13,400 g and 4°C for 10 min, the supernatant was used as crude enzyme extract for assaying activity. Activity of OAS-TL was assayed according to a method reported by Chronis and Krishnan (2003). The absorbance of solution at 420 nm was recorded every minute from 0 to 5 min. The concentration of K 3 Fe(CN) 6 was calculated by a standard curve. The amount of enzyme required for reduction of 1 μmol K 3 Fe(CN) 6 every minute was calculated as one activity unit (U).

| Statistical analysis
The variance of data was analyzed using SPSS software package release 16.0 (SPSS Inc. Chicago, IL, USA). Multiple comparisons were performed by one-way ANOVA based on Duncan's multiple range tests.

| DS accelerated the color recovery of sulfitated fruits and kept the inhibitory effect of SF on browning and decay
To investigate the effect of desulfurization on the quality of SF "Feizixiao" litchi, the pigmentation, browning index, and rotting rate were compared among CK, SF, and SF + DS groups. The red pigmentation of "Feizixiao" litchi fruits immediately disappeared and turned to be yellowish green after the sulfur fumigation, but it quickly restored after the desulfurization (0 DAS, AF; Figure 1a). The results of chromatic value indicated a trend that the lightness of "Feizixiao" litchi was SF > SF + DS > CK (Figure 1b), while the a* value was totally CK > SF + DS > SF (Figure 1c). During the low temperature storage, the pericarp browning and decay were effectively inhibited by both of the sulfur fumigation and desulfurization treatment (Figure 1a,d,e).
Although the pericarp browning of SF fruits showed a significantly slower increase than that of SF + DS fruits since 16 DAS, the pericarp browning of SF + DS fruits totally showed a far slower increase than that of the CK fruits which showed an obvious browning appearance (pericarp browning index >1) since 16 DAS (Figure 1d; p < 0.05).
Moreover, the rotting rate of the CK fruits showed an obvious increase since 32 DAS and reached 21.67% at 48 DAS, while that of SF and SF + DS fruits showed obvious increase since 40 DAS and reached only 8.33% and 15% at 48 DAS, respectively ( Figure 1e).
These results indicated that desulfurization treatment not only effectively restored the pigmentation, but also performed a significantly inhibitory effect on the pericarp browning and decay.

| DS totally increased the water content, relative electroconductibility, and respiration rate of sulfitated litchi pericarp
The water content of CK litchi pericarp maintained at a relatively stable level (decreased from 69.08% to 67.71%), while that of the SF and SF + DS litchi pericarp respectively decreased by 7.14% and 6.4% after a 48-days storage (Figure 2a). The relative electroconductibility of the CK litchi pericarp fluctuated from 20.3% to 23.4%. Both of the SF and SF + DS treatments lead to a higher relative electroconductibility of pericarp which fluctuated from 60.7% to 74.22% during the storage. The relative electroconductibility of the SF + DS litchi pericarp was higher than that of the SF litchi pericarp at 8-24 DAS and showed no significant difference to that of the SF litchi pericarp at 0 DAS (AF) and 24-48 DAS (Figure 2b). The respiration rate of the CK fruits was 9.14 mg kg −1 hr −1 at 0 DAS (AF) and rapidly decreased to 3.99 mg kg −1 hr −1 at 8 DAS and then gently decreased to 2.94 mg kg −1 hr −1 at 48 DAS. The respiration rates of the SF and SF + DS fruits were, respectively, 5.13 and 7.77 mg kg −1 hr −1 at 0 DAS (AF), and were significantly lower than that of the CK fruits during the whole storage except 48 DAS. The respiration rate of the SF + DS fruits was totally higher than that of the SF fruits except 24 DAS and 48 DAS (Figure 2c). These results indicated that desulfurization might help recovered the water content and respiration rate of SF litchi pericarp, but totally increased the relative electroconductibility.

| DS significantly reduced the residual sulfite in the pericarp and pulp of sulfitated litchi fruits
To examine the effect of desulfurization on the residual SO 2 of SF "Feizixiao" litchi fruits, the sulfite content in the pericarp and pulp

| DS significantly reduced the activity of enzymes responsible for detoxification of sulfite in the sulfitated litchi pulp
The activity of five enzymes related to metabolically detoxification of sulfite in litchi pulp was detected during the storage. The The SiR activity in the SF litchi pulp was higher than that in the CK pulp by two-to threefold during the storage, while the SiR activity in the SF + DS litchi pulp was totally significantly lower than that in the SF pulp but totally significantly higher than that in the CK litchi pulp (Figure 4c). The SAT activity in the CK litchi pulp showed a smooth fluctuation, while that in the SF litchi pulp was higher than that in the CK pulp by more than fourfold at 0 DAS (AF), increased and reached the peak value at 8 DAS and then decreased. The SAT activity in the SF + DS litchi pulp was significantly lower than that in the SF litchi pulp through the whole storage. It is noteworthy that the SAT activity in the SF + DS litchi pulp was higher than that in the CK litchi pulp only at 0 DAS (AF) to 24 DAS (Figure 4d). Also, the OAS-TL activity in the SF litchi pulp was parallelly higher than that in the CK litchi pulp, while the OAS-TL activity in the SF + DS litchi pulp was totally lower than that in the SF litchi pulp but totally higher than CK (Figure 4e). The concentration of SO 3 2− in the sulfitated pulp rapidly decreased after desulfurization, while the activity of enzymes for the oxidation and reduction of SO 3 2− decreased simultaneously. This result indicated that the litchi fruits were able to respond quickly to the SO 2 stress signal, which was transmitted from pericarp to pulp; the upregulated enzyme activity of SO and SAT played an important role in the sulfite metabolism in litchi pulp. Besides the chemical reaction with the sulfite, the DS treatment might also influence the sulfite metabolism by regulation at enzymatic level.

| DS significantly recovered the expression of the genes related to detoxification of sulfite in the sulfitated litchi pulp
The expression of five genes related to metabolically detoxification of sulfite in the litchi pulp was detected during the storage.
Interestingly, the expression of all of the five genes in the SF litchi pulp was totally lower than that in the CK litchi pulp, while the desul- ) which was cytotoxicity. The sulfate was mainly metabolically detoxified by oxidative reaction into sulfate with increased sulfite oxidase activity or by reductive reaction into S-metabolites like thiols (Baillie et al., 2016). In this work, our results showed that the activity of SiR, SAT, and OAS-TL in the litchi pulp increased by folds rapidly after sulfur fumigation.
Especially, the activity of SAT increased by more than 10 times after a storage of 8 days and maintained at a relatively stable level later. However, the activity of SO increased slowly after sulfur fumigation. These results were not consistent with the previously reported result that more than 80% of the injected sulfite in arabidopsis and 91% in tomato were oxidized to sulfate which demonstrating the high capacity of the sulfite oxidation mechanisms in plants (Brychkova et al., 2012). Thus, the detoxification

| CON CLUS IONS
A comprehensive evaluation of the effect of desulfurization on the storability and the sulfite metabolism in sulfitated and desulfurized fruits is still lacking, while controversial results have been reported in the literature focused on the effects of low concentration SO 2 on F I G U R E 4 Effect of sulfur fumigation and desulfurization treatment on the activity of five enzymes (SO, APR, SiR, SAT, and OAS-TL) related to sulfur metabolism in pulp of "Feizixiao" litchi (stored at 4°C). Lower case letters after the means designate significance at p < 0.05 sulfur metabolism in plant leaves or roots. In the present study, our results demonstrate that the optimized desulfurization treatment accelerated the color recovery of sulfitated litchi fruits, achieved an effect similar to sulfur fumigation on controlling rot and freshkeeping, and reduced the sulfite residue in the sulfitated litchi so as to ensure its edible safety. The upregulated enzyme activity of SO and SAT, rather than the expression level, plays a central role in the sulfite metabolism of sulfitated litchi pulp. More importantly, DS ensured the edible safety of sulfitated litchi by not only chemical reaction but also both of enzymatic and transcriptional regulation of sulfite metabolism. Conclusions from this study will be helpful to optimize the strategies of sulfur fumigation and desulfurization, and F I G U R E 5 Effect of sulfur fumigation and desulfurization treatment on the expression of five genes (SO, APR, SiR, SAT, and OAS-TL) related to sulfur metabolism in pulp of "Feizixiao" litchi (stored at 4°C). Lower case letters after the means designate significance at p < 0.05 provide a theoretical basis for controlling the sulfur residue in the practices of litchi export.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.

E TH I C A L A PPROVA L
The article does not contain any studies with human participants or animals performed by any of the authors.