Physiological, molecular and ultrastructural analyses during ripening and over-

ripening of banana (Musa spp., AAA group, Cavendish subgroup) fruit suggest

characteristics of programmed cell death

Maricruz Ramírez-Sánchez1*, Donald J. Huber1, C. Eduardo Vallejos1 and Karen Kelley2

Accepted Article

1
Horticultural Sciences Department, PO Box 110690, IFAS, University of Florida,

Gainesville, FL 32611-0690, USA

2
Electron Microscopy and Bio-imaging Core, ICBR, University of Florida, Gainesville, FL

32611, USA

*Corresponding author e-mail address: maricruz.ramirezsanchez@ucr.ac.cr

Correspondence address: Centro de Investigaciones Agronómicas, Universidad de

Costa Rica, PO Box 2060-11501, San José, Costa Rica

Abstract

BACKGROUND: Programmed cell death (PCD) is a part of plant development that has

been studied for petal senescence and vegetative tissue but has not been thoroughly

investigated for fleshy fruits. The purpose of this research was to examine ripening and

over-ripening in banana fruit to determine if there were processes in common to

previously described PCD.

This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process, which
may lead to differences between this version and the Version of Record. Please cite this
article as doi: 10.1002/jsfa.8505

This article is protected by copyright. All rights reserved.

 RESULTS: Loss of cellular integrity (over 40 %) and development of senescence

related dark spot (SRDS) occurred after day 8 in banana peel. Nuclease and protease

activity in the peel increased during ripening starting from day 2, and decreased during
Accepted Article
over-ripening. The highest activity was for proteases and nucleases with apparent

molecular weights of 86 kDa and 27 kDa, respectively. Images of SRDS showed

shrinkage of the upper layers of cells, visually suggesting cell death. Decrease of

electron dense areas was evident in TEM micrographs of nuclei.

CONCLUSION: This study shows for the first time that ripening and over-ripening of

banana peel share physiological and molecular processes previously described in plant

PCD. SRDS could represent a morphotype of PCD that characterizes a structural and

biochemical failure in the upper layers of the peel, thereafter spreading to lower and

adjacent layers of cells.

Keywords

Banana, ripening, over-ripening, protease, nuclease, programmed cell death

Introduction

Plant programmed cell death (PCD) is an integral part of normal plant

development including embryogenesis, leaf morphology, floral organ abortion, root cap

sloughing, leaf senescence, and the development of gametophytes and vascular

tissue.1,2 PCD is described as an active process that removes cells, tissues, or organs

that are no longer required for subsequent development, where in most of the cases

nutrients are remobilized and transported to other sinks in the plant.3,4,5,6 During PCD,

This article is protected by copyright. All rights reserved.

 developmental or environmental stimuli activate a ‘specific series of events that

culminate in cell death’.7 The initiation and progression of this process is influenced by

both biotic and abiotic factors, the characteristics of the tissue/organ, and
Accepted Article
developmental stage.6

PCD processes have been extensively studied in flower petals2,8,9,10,11 and

vegetative tissues.12,13,14 There are very limited reports on PCD events in fleshy fruits

including heat-induced PCD in tomato15, ethylene induced PCD in immature

cucumber16,17,18, and CI-induced PCD in cucumber.19,20 None have addressed the

occurrence of PCD-like events during ripening and over-ripening in fleshy fruits. Qu et

al.15 made no mention of deleterious effects of the PCD responses on tomato fruit

ripening.

Bananas are fleshy simple fruits with a soft epicarp and a very fleshy, edible

mesocarp and endocarp comprised of starchy parenchyma.21,22 During ripening, the

crisp, hard, and dark green banana is transformed into a yellow fruit with tender and soft

internal pulp at the full ripening stage.23 The ripening and softening are major attributes

of perishability in fleshy climacteric fruit, including bananas. A few days after the fruit

ripens, it is inedible due to over-ripening. The spoilage includes excessive softening and

changes in taste, aroma, and skin color.24 In advanced stages of banana ripening (over-

ripening), senescence-related dark spots (SRDS) develop on the skin.25

There are studies of the onset of SRDS development from environmental

conditions including relative humidity, temperature, and gas concentrations and

exchange rates.26,27,28,29 Moser et al.30 described that these dark spots are generated

from the vicinity of stomata. They also analyzed the breakdown of chlorophyll (Chl) as

This article is protected by copyright. All rights reserved.

 an indicator of senescence and the accumulation of senescence associated dark spots

(SADS) as a sign of localized cell death.

The purpose of the present research was to examine the changes at the
Accepted Article
physiological, biochemical and ultrastructural levels during ripening and over-ripening of

banana fruit to identify and evaluate the presence of PCD indicators on degradative

processes during ripening and over-ripening.

Materials and methods

Plant material

Green, ethylene-treated bananas (Musa spp., AAA group, Cavendish subgroup)

were shipped overnight to the Horticultural Sciences Department, UF, Gainesville, FL.

Ethylene treatment (150-300 μL L-1) for 48 h at 17 °C was performed at Chiquita brands

facilities in Fort Lauderdale, FL immediately prior to shipping to UF. Upon arrival at the

Horticultural Sciences Department, banana clusters were removed from packaging and

transferred to storage at 20 °C and 80-85 % RH to simulate shelf life conditions. Three

randomly selected clusters were taken at 6 h, 12 h, 24 h and every two days until day 8

and every 4 days until day 16. Enzyme analyses were performed through 12 days. The

experiment was repeated two times using different fruit shipments, the data

corresponds to the second experiment.

Respiration rate and ethylene production

Six banana fingers were randomly selected and separated from the clusters at

each ripening/storage interval. Two banana fingers per replicate were sealed in 3.3-L

glass containers fitted with septa for 1 h at 20 °C. Three-mililiter samples of the

headspace atmosphere were withdrawn with a 3-mL gas syringe. Ethylene and carbon

This article is protected by copyright. All rights reserved.

 dioxide were determined using gas chromatography as described in detail by Zhang et

al.31 The GC was calibrated daily using standard gas mixtures of 0.98 μL L-1 ethylene

and 1 g L-1 CO2. Ethylene production and respiration rate are expressed as ng kg-1s-1
Accepted Article
and µg kg-1s-1, respectively.

Electrolyte leakage and nutrient leakage

Tissue cylinders were excised from the equatorial region of the peel tissue using

a 9 mm diameter (No. 5) cork borer. Thickness of the cylinders varied and

corresponded to the thickness of the peel at the storage interval of the evaluation. A

total of 9 cylinders, 3 cylinders per 3 fruit, were rinsed with distilled H2O and then placed

in 30 mL of a 250 mmol L-1 mannitol solution in 50 mL tubes at 25 °C for 4 h.32 During

incubation the tubes were shaken in a precision reciprocal shaking water bath at 25 °C.

The conductivity of the mannitol bathing solution was measured after 4 h of incubation

using a YSI-31A conductivity bridge equipped with a conductivity cell (Model 3403,

Yellow Springs, OH, USA). Total electrolyte content was determined after freezing (24 h

at -20 °C), thawing, and heating the disks and bathing solutions in a boiling water bath

for 30 min. Electrolyte efflux is expressed as a percentage of total tissue electrolyte

content.

Samples for ion leakage followed the same preparation procedure for electrolyte

leakage. After the 4-h incubation period the incubation solution was filtered using

Whatman filter paper (#42 ashless). The filtrate was stored at -20 °C in 20 mL

scintillation vials until all samples were collected. The samples used to determine total

electrolyte content were also filtered and stored for ion analysis. The samples were

analyzed for phosphorus, potassium, calcium and magnesium content at the University

This article is protected by copyright. All rights reserved.

 of Florida Analytical Research Laboratory (IFAS Analytical Services Laboratories).

Nutrient leakage was expressed as a percentage of total nutrient content.

Protein extraction and quantification

Accepted Article
Three banana fingers were randomly selected from the clusters at each

ripening/storage interval. Peel tissue from the selected fruit was excised (approximately

6 g per finger), wrapped in aluminum foil and stored at -80 °C. Peel tissue was ground

to a fine powder in liquid N2 with an IKA A11 (A) basic analytical mill grinder (IKA®,

Wilmington, NC), and stored at -80 °C. Powdered peel tissue (0.2 g) was transferred to

a 1.5 mL microcentrifuge tube and mixed with 700 µL of protein extraction buffer (0.4

mol L-1 MOPS pH 7.0, 10 mL L-1 glycerol , 1 mL L-1 Triton-X 100 and 6 mmol L-1 DTT).

One mmol L-1 phenylmethanesulfonylfluoride (PMSF; a serine protease inhibitor) was

included only in the extraction buffer for nuclease activity. The mix was vortexed for 30 s

on the highest setting and the samples were then centrifuged at 8,000 x g for 10 min at

4 °C. The supernatant was transferred to a fresh microcentrifuge tube and kept on ice

until assayed for nuclease activity or protease activity. Total protein was determined in a

microplate assay using the Bradford33 method and bovine ɣ-globulin as standard.

Azocasein protease activity assay

Colorimetric protease activity was measured according to Sousa et al.34 with

modifications. Azocasein was used as substrate at a concentration of 5 g kg-1 in 50

mmol L-1 Na-acetate buffer, pH 5.2. One-hundred µL of protein extract prepared as

described above were mixed with 200 µL of the azocasein-buffer mix. For the controls,

500 µL of 100 g L-1 of trichloracetic acid (TCA) were added prior to adding protein. All

samples were incubated at 37 °C for 24 h. After incubation, 500 µL of 100 g L-1 TCA

This article is protected by copyright. All rights reserved.

 were added and, after 10 min at room temperature, the samples were microcentrifuged

at 2000 x g. Afterward, 500 µL of the supernatants were mixed with 500 µL of 1 mol L-1

NaOH and 200 µL of the mixture were analyzed at 440 nm with a microplate
Accepted Article
spectrophotometer (Power Wave™ XS, BioTek Instruments, Inc., VT). One unit of

activity is defined as a change in absorbance of 0.01 under the conditions of the assay.

Data represent the average of nine microplate readings, 3 replicates of the 3 protein

extracts of each ripening/storage interval and are expressed as units kg-1 protein s-1.

In-gel Protease Activity

For in-gel activity detection, a volume of the protein extract of each

ripening/storage interval containing 15 μg of protein was added to an equal volume of a

non-reducing SDS buffer comprised of 0.5 mmol L-1 Tris-HCl, pH 6.8, 100 g L-1 SDS,

100 g L-1 glycerol, and 5 g L-1 bromophenol blue. The procedures for electrophoresis

and removal of SDS prior to enzyme detection followed the methodology described by

Azeez et al.35 with modifications. The gels were 80 g L-1 acrylamide/bis, 1 g L-1 gelatin.

Electrophoresis was carried out at 200 V for approximately 80 min. Afterward, SDS was

removed by washing the gels with 25 mL L-1 Triton X-100 and 50 mmol L-1 Tris-HCl, pH

7.5 for 30 min, changing the washing solution every 15 min, followed by two changes of

renaturing solution that included 50 mmol L-1 Tris-HCl, pH 7.5, 10 mmol L-1 MgCl2 and

10 mmol L-1 CaCl2. Gels were then incubated at 37 °C for 16 hours. Gel were stained

with brilliant blue-R250 (1 g L-1 R-250 in 500 mL L-1 methanol and 100 mL L-1 glacial

acetic acid) and destained with a solution of 400 mL L-1 methanol and 100 mL L-1

glacial. Protease activity was observed as clear bands on a dark background. Gels were

This article is protected by copyright. All rights reserved.

 run in triplicate, one for each protein extract at each ripening/storage interval. Analysis

of band signal intensity was performed using Image Studio Lite (LI-COR®).

DNA-Methyl green nuclease activity assay

Accepted Article
Nuclease activity was measured according to Bronstein and Weber36 with

modifications. A sample of the protein extract (30 μL) was added to a 96-well

microplate. Fifty μL of 0.2 mol L-1 sodium citrate, pH 7.5, was added to the controls of

each sample prior to incubation to stop the reaction. Then 95 μL of nuclease reaction

buffer (80 mmol L-1 MOPS, pH 7, 1 mmol L-1 MgCl2, and 1 mmol L-1 CaCl2) was added

to the wells. The plates were covered and incubated at 37 °C for 10 min. Afterward, 25

μL of a DNA methyl green solution was added to each well. This solution of DNA methyl

green (1 mg mL-1) (Sigma, St. Louis, MO, USA) was mixed and dissolved previously by

holding in a water bath at 40 °C. The microplate was covered and incubated at 37 °C for

2 h. Afterward, 100 μL of 200 mmol L-1 sodium citrate, pH 7.5 was added to the sample

wells to stop the reaction and stabilize the color. The plate was then read at 655 nm

with a microplate spectrophotometer (Power Wave™ XS, BioTek Instruments, Inc., VT).

One unit of activity was defined as a change in absorbance of 0.01 under the conditions

of the assay. The data represent the average of nine microplate readings, 3 replicates

of the 3 protein extracts of each ripening/storage interval and are expressed as units kg-
1
FW s-1.

In-gel Nuclease Activity

For in-gel activity detection, a volume of the protein extract of each

ripening/storage interval containing 1 μg of protein was added to 15 μL of sodium

dodecyl sulfate (SDS) buffer containing 0.5 mol L-1 Tris-HCl, pH 6.8, 100 g L-1 SDS, 100

This article is protected by copyright. All rights reserved.

 g L-1 glycerol, and 5 g L-1 bromophenol blue. The procedures for electrophoresis and the

removal of SDS from gels prior to enzyme detection were as described by Thelen and

Northcote37 with the following modifications. Gel consisted of 120 g L-1 acrylamide/bis,
Accepted Article
50 μg mL-1 bovine fibrinogen, and 30 μg mL-1 herring sperm DNA. Electrophoresis was

carried out at 200 V for approximately 1 h. Afterward, SDS was removed by washing the

gels with 250 mL L-1 isopropanol, 10 mmol L-1 MOPS, pH 7, 1 mmol L-1 CaCl2, and 1

mmol L-1 MgCl2 for 30 min, changing the washing solution every 15 min, then followed

by two changes of buffer without isopropanol. After washing, gels were incubated in 10

mmol L-1 MOPS, pH 7, 1 mmol L-1 CaCl2, and 0.1 mmol L-1 MgCl2, at 37 °C for 2 h.

Undigested DNA in the gel was stained with a solution of 1 μg mL-1 of ethidium bromide

(dissolved in water) and visualized with a UV trans-illuminator. Nuclease activities

appear as dark spots on a white background. Gels were run in triplicate, one for each

protein extract at each ripening/storage interval. Image analysis to determine signal

intensity of bands was performed using Image Studio Lite (LI-COR®).

Light and Transmission Electron Microscopy Images

Peel tissues were collected at 2, 6 and 10 days during ripening and over-ripening

of fruit treated with ethylene. Epidermal tissues measuring 3 mm3 were excised from the

peel using a blade razor and immersed in a 1.5 mL centrifuge tube containing

McDowells Trump primary fixative (Electron Microscopy Sciences, Hatfield, PA)

consisting of sodium cacodylate, formalin and glutaraldehyde for 1 h at room

temperature then stored for one week at 4 °C. Tissue processing and embedding was

microwave-assisted with a Pelco® BioWave Pro laboratory microwave (Ted Pella,

Redding, CA) equipped with a cold spot and a Pelco® microwave vacuum chamber

This article is protected by copyright. All rights reserved.

 (Ted Pella, Redding, CA). After each microwave irradiation, the vacuum was released

and tissues in microcentrifuge tubes were placed on a test tube rocker (Thermo

Scientific™) for 5 min. The fixed tissues were washed three times in 0.1 mol L-1
Accepted Article
cacodylate, pH 7.24, followed by subsequent water washes, 45 seconds in the

microwave oven (at 220 watts) each. A graded ethanol series using 50 mL L-1

increments from 250 mL L-1 through 1000 mL L-1 were carried out without vacuum for 45

second in the microwave (at 220 watts). Dehydrated tissues were infiltrated with LR

White hard acrylic resin containing Z6040 silane (100 μL/10 mL resin) (Electron

Microscopy Sciences, Hatfield, PA), to assist resin adherence to the waxy epidermal

surface. LR White resin infiltrations were carried out 5 times in the microwave at a

pressure of 67727.8 Pa (220 watts) for 3 min each followed by 10 min on the test tube

rocker. The final 1000 mL L-1 resin infiltration was left over night at room temperature on

the test tube rocker. Tissues were transferred to flat bottom BEAM capsules (Ted Pella,

Redding CA) with fresh 1000 mL L-1 LR White resin containing Z6040 silane (100 μL/10

mL resin), and polymerized in a Labline curing oven (Thermo Scientific, Waltham, MA)

at 60 oC for 48 h. The cured resin blocks were hand-trimmed with a blade razor to

remove excess resin and prepared for ultramicrotomy. Semi-thick sections (500 nm)

were cut using a diamond knife (Micro Star Technologies Inc., Huntsville, Tx) on a UCT

ultramicrotome (Leica Microsystems, Buffalo Grove, IL). Semi-thick sections were dried

onto a glass slide with the aid of a hot plate at 120 oC, stained with 2 g kg-1 toluidine

blue solution for 30 s, rinsed with water, and dried once more before examination using

an Olympus BH-2 bright field microscope (Olympus, Tokyo, Japan). Digital images were

acquired with a Retiga 2000R fast cooled 12-bit camera, equipped with a color filter

This article is protected by copyright. All rights reserved.

 slider and QCapture Pro 7.0 (Qimaging, Surrey, BC, Canada). Areas of banana peel

both with and without senescence related dark spots were examined in the semi-thick

sections.
Accepted Article
Transmission electron microscopy analysis followed the same tissue collection,

processing and embedding described for light microscopy, with the difference of the

primary fixative and the staining with osmium tetroxide (Electron Microscopy Sciences,

Hatfield, PA). Samples were fixed in a solution containing 25 g kg-1 glutaraldehyde, 0.2

mol L-1 cacodylate and 30 g L-1 sucrose, pH 7.4. After the fixed tissues were washed

three times in 0.1 mol L-1 cacodylate, pH 7.24, they were transferred to a solution of 20

g L-1 osmium tetroxide and left 1 min under the hood. They were microwaved for 45 s

(at 100 Watts and at a pressure of 67727.8 Pa). Samples were then transferred to a

hood and left for 3 min. The fixed tissues were washed three times in 0.1 mol L-1

cacodylate, 15 g L-1 sucrose, pH 7.4 followed by subsequent water washes, 45 s in the

microwave oven (at 220 watts) each. The graded ethanol dehydration and resin

infiltration followed the same steps described for light microscopy. The resin blocks

were trimmed to a width of 0.5-1 mm X 1 mm long using a double-edge razor blade.

The block faces were trimmed into trapezoidal shapes ensuring that the leading and

trailing edges were parallel to each other. Ultra-thin sections (100-120 nm) were

collected on carbon coated Formvar copper slotted grids, post-stained with 20 g L-1

aqueous uranyl acetate and Reynold’s lead citrate. Ultra-thin tissue sections were

examined on a transmission electron microscope (Hitachi High Technologies America,

Clarksburg MD) and digital images were acquired with a Veleta 2k x 2k side mount

camera operating iTEM software (ResAlta, Golden CO).

This article is protected by copyright. All rights reserved.

 Statistical analysis

One-way analysis of variance (ANOVA) was used to analyze the effect of time

(days) on the evaluated variables. The analysis of variance was performed by PROC
Accepted Article
GLM using SAS statistical software (version 9.2; SAS Institute, Cary, NC, USA). The

means were compared using the least significant differences (LSD) at the 0.05 level of

significance in the ANOVA. Data are expressed as means ± standard error of the mean.

Results

Fruit ripening, respiration rate and ethylene production

Banana fruit ethylene-treated in commercial facilities exhibited a typical ripening

pattern during storage at 20 °C and 80-85 % RH (Fig. 1). Ethylene and carbon dioxide

production of ethylene-treated, ripening-initiated banana fruit showed significant

differences among the evaluation days. Ethylene production increased from day 0,

reaching an initial maximum on day 2 (0.55 ηg kg-1 s-1). Thereafter, ethylene maintained

a steady production until day 8 and then sharply increased with further development.

Carbon dioxide production reached a maximum value (31.5 μg kg-1 s-1) at day 4 and

production rates remained nearly constant for the duration of ripening and over-ripening

(Fig. 1).

Electrolyte leakage and nutrient leakage

Electrolyte leakage of banana peel showed an increase over ripening and over-

ripening, reaching nearly 40 % by day 8. Magnesium leakage was delayed compared

with the other elements, first increasing after day 8 and reaching more than 40 % by day

12. Calcium had the earliest increase in leakage, first evident at 2 d and reaching more

than 60 % by day 6 and 80 % by day 16. Potassium and phosphorus had similar

This article is protected by copyright. All rights reserved.

 patterns of leakage compared to total electrolyte leakage, showing values of

approximately 40 % by day 8 (Fig. 2).

Nuclease and protease activity of banana peel

Accepted Article
Colorimetric assays revealed increases in both nuclease and protease activities

during ripening and over-ripening. Activity levels showed statistical differences

(p<0.001) among the evaluation days. Nuclease and protease activities reached

maximum values of 266.8 units kg-1 FW s-1 and 23,947.9 units kg-1 protein s-1,

respectively, on day 6. Thereafter, activities of both enzymes declined during over-

ripening in parallel with the development of SRDS (Fig. 5).

In-gel DNA-based nuclease activity showed a steady increase from day 0,

showing bands with apparent MW of 25, 27, 29, and 31 kDa (Fig. 6). A maximum of

activity was observed on day 8, which could be due to saturation of the signal. Activity

declined at day 12. Band signal intensity indicated that the nuclease of apparent MW of

∼27 kDa was the most catalytically active of the proteins, reaching a maximum signal

intensity of about 60,000. In-gel protease activity showed a continual increase until day

6. Observed bands exhibited apparent MWs of 79, 86, 123, 151, 169, and 192 kDa (Fig.

7). The 86 kDa protease presented the highest signal, followed by the 169 and 192 kDa

protease, respectively.

Light and Transmission Electron Microscopy Images of banana peel

Banana peel during ripening exhibited morphological changes in the epidermal

and sub-epidermal cells. On day 2, peel was comprised of epidermal cells with bulliform

shape (Fig. 3). By day 6, the upper layers showed horizontal elongation with an

associated decrease in cell height. Tissue collected on day 10, corresponding to a

This article is protected by copyright. All rights reserved.

 senescence (over-ripening)-related dark spot (SRDS), revealed acute collapse at the

surface layers. Vascular tissue at day 10, however, maintained normal appearance,

even in regions of SRDS. Nuclei clearly showed early changes after ethylene
Accepted Article
treatments (Fig. 4). In preclimacteric tissue, nuclei had high electron density that

subsequently declined at 2 and 6 days after ethylene treatment. Lower electron density

and signs of heterochromatin condensation were observed after 10 days at 20 °C while

the SRDS showed loss of resolution of all organelles including nuclei.

Discussion

In banana fruit, Domínguez and Vendrell38 and previously Ke and Tsai39 reported

that ripening of the peel depends on the pulp. For the present research, ethylene and

carbon dioxide production measured in intact, whole fruit for this research agreed with

what Dominguez and Vendrell38 observed in intact fruit. The fruit used for these

experiments exhibited uniform ripening since they were sourced and treated with

ethylene in commercial facilities.

Banana peel cell integrity diminished steadily during ripening, resulting in close to

40 % total electrolyte leakage (EL) by day 8, at which time the fruit were fully yellow

corresponding to ripening stage 6 in the scale of von Loesecke.40 Afterward,

senescence-related dark spots (SRDS) appear, this latter ripening phase associated

with a continuous increase in EL in banana peel after day 8, similar to that reported by

Wang et al41 for senescing banana peel. High leakage of calcium and phosphorus,

structural components of cell wall and membranes, observed in the present study is

consistent with loss of membrane integrity and permeability as reflected by the

significant increase in EL. Potassium leakage has been described as an indicator of

This article is protected by copyright. All rights reserved.

 membrane permeability.42 Similar analyses of strawberry fruit disks43 showed more than

30% potassium leakage in harvested strawberry (75 % red color). In the case of calcium

and magnesium, as cofactors for many enzymes, including nucleases and proteases,
Accepted Article
increased leakage might represent a limiting factor for activity. Calcium also plays a role

in signaling and cell death.44 Leakage of magnesium, a component of chlorophyll,

increased after the fruit was fully yellow showing possible correlation with chlorophyll

degradation and plastid differentiation in banana peel tissue.

Colorimetric assays revealed significant increases in peel protease and nuclease

activity. Immature cucumber also showed an increase in both protease and nuclease

activity in response to ethylene exposure.16,18 During ripening and over-ripening of

banana, electrophoresis revealed a smear of degraded DNA particularly after day 8

(data not shown), which would correlate with the activity of nucleases, the initiation of

development of SRDS and more than 40 % loss of cellular integrity. Internal factors that

regulate the relatively fast changes in banana peel cells during ripening and over-

ripening might have the ultimate function of inducing degradation and death at a higher

rate compared to the pulp. The acceleration in the degradation rate of the tissues,

concomitant with elevated respiratory rates through the remainder of over-ripening of

banana, might facilitate the dispersal of seeds, a process that likely persists in the

parthenocarpic banana. The osmotic pressure gradient generated from the rapid

increase in pulp osmotic pressure compared with the more moderate increase in the

peel results in net water transfer from peel to pulp.45,46

Bright field microscopy of banana peel revealed that epidermal and sub-

epidermal cells showed acute morphological changes during ripening. During over-

This article is protected by copyright. All rights reserved.

 ripening and development of SRDS, cell collapse and loss of cell integrity was also

evident in bright field images. Changes in cell shape have been described in banana

peel throughout development. From mature green stage until fruit was fully yellow, both
Accepted Article
the increase in air space and decrease in peel thickness were likely due to modifications

and degradation of the cell wall and cell separation.47 After initiation of ripening of

banana peel there was a reduction of cell size in ground tissue and an increase in the

volume of air space.47 In the present study, both fresh tissue and bright field images of

SRDS revealed sunken layers of tissue with characteristic browning and vascular

bundles scattered throughout. The shrinkage of the surface layers visually suggested

loss of integrity and cell death. Ratule et al.48 also observed loss of integrity of peel cell

wall during banana ripening.

The increase in ethylene production, membrane disruption, and degradative

pathways during ripening and over-ripening could be due to biotic stressors, with an

accelerated response in the peel tissue. Gene expression of stress-, defense-, and

detoxification-related genes were found to be ethylene-induced in banana pulp tissue49,

showing that ethylene mediates and/or communicates a general stress response.

During banana ripening, moisture content decreases, and other tissue fluids including

latex affect water relations.50 Water loss can exacerbate the stress response; when

banana fruit reached 5 % fresh mass loss, respiration and ethylene production

increased more than 50 % compared with non-stressed fruit.51 Even though water loss

was not measured in the present study, water loss is a common phenomenon during

ripening and over-ripening from fruit respiration and diffusion through the fruit cuticle.52

Reported water loss in banana fruit during shelf life normally ranges between 20 and 70

This article is protected by copyright. All rights reserved.

 g kg-1.53,54 In the present research, loss of integrity of these outermost layers were

evident during ripening and over-ripening. In addition to structural and cell integrity

changes in banana peel epidermal cells, ambient conditions including gas composition
Accepted Article
and relative humidity can affect the development of SRDS, particularly in sensitive

cultivars like Sucrier.26 Adverse environmental conditions in combination with the loss of

cell integrity might accelerate over-ripening, tissue degradation and the development of

SRDS.

The SRDS could be the result of a fruit-environment interaction that originates as

a defense response, involving both mitochondria and plastids, as sources and targets of

PCD signals including reactive oxygen and nitrogen species.55,56 Therefore we

speculate that SRDS could represent a morphotype of PCD reflecting structural and

biochemical failure in the superficial layers of the peel, thereafter spreading to lower and

adjacent layers of cells. Moser et al.30 reported using differential interference contrast

that stomata cells are visible in the center of a SRDS. Peel stomata have a more

scattered pattern compared with the linear pattern as on the leaf57, which would

correlate to the quasi-random distribution of SRDS. The development of SRDS could

reflect the culmination of a loss of cellular homeostasis, brought about by an

unfavorable water balance in the vicinity of stomata. This would explain the accelerated

onset of SRDS development by environmental conditions including relative humidity,

temperature, and gas concentrations and exchange rates.26,27,28,29

Ripening and over-ripening in banana peel included morphological changes in

cell structure, increases in electrolyte leakage, and increases in protease and nuclease

activities, features that have previously been described in PCD in other plant systems.

This article is protected by copyright. All rights reserved.

 Banana fruit would seem to represent a model for further study of programmed cell

death morphology in fleshy fruits. More research is necessary in other fleshy fruits.
Accepted Article
References

1 Jones AM and Dangl JL, Logjam at the Styx: programmed cell death in plants.
Trends Plant Sci 1:114-119 (1996).
2 Xu Y and Hanson MR, Programmed cell death during pollination-induced petal
senescence in petunia. Plant Physiol 122:1323-1333 (2000).
3 van Doorn WG and Woltering EJ, Many ways to exit? Cell death categories in plants.
Trends Plant Sci 10:117-122 (2005).
4 Gunawardena AHLAN, Programmed cell death and tissue remodeling in plants. J
Exp Bot 59:445-451 (2008).
5 Reape TJ, Molony EM and McCabe PF, Programmed cell death in plants:
distinguishing between different modes. J Exp Bot 59:435-444 (2008).
6 Trobacher CP, Ethylene and programmed cell death in plants. Bot 87:757-769 (2009).
7 van Doorn WG and Woltering EJ, Senescence and programmed cell death:
substance or semantics?. J Exp Bot 55:2147-2153 (2004).
8 Yamada T, Takatsu Y, Kasumi M, Manabe T, Hayashi M, Marubashi W and Niwa M,
Novel evaluation method of flower senescence in Freesia (Freesia hybrida)
based on apoptosis as an indicator. Plant BioTech 18:215-218 (2001).
9 Yamada T, Ichimura K and van Doorn WG, DNA degradation and nuclear
degeneration during program cell death in petals of Antirrhinum, Argyranthemun,
and Petunia. J Exp Bot 57:3543-2552 (2006a).
10 Yamada T, Takatsu Y, Kasumi M, Ichimura K and van Doon WG, Nuclear
fragmentation and DNA degradation during programmed cell death in petals of
morning glory (Ipomoea nil). Planta 224:1279-1290 (2006b).
11 Rubinstein B, Regulation of cell death in flower petals. Plant Mol Biol 44:303-318
(2000).
12 Coupe SA, Watson LM, Ryan DJ, Pinkney TT and Eason JR, Molecular analysis of
programmed cell death during senescence in Arabidopsis thaliana and Brassica
oleracea: cloning broccoli LSD1, Bax inhibitor and serine palmitoyltransferase
homologues. J Exp Bot 55:59-68 (2004).
13 Eason JR, Pinkney TT and Johnston JW, DNA fragmentation and nuclear
degradation during harvest-induced senescence of asparagus spears. Postharv
Biol Tech 26:231-235 (2002).
14 Fukuda H, Programmed cell death of tracheary elements as a paradigm in plants.
Plant Mol Biol 44:245-253 (2000).
15 Qu G-Q, Liu X, Zhang Y-L, Yang D, Ma Q-M, Yang M-Y, Zhu W-H, Yu S and Luo Y-
B, Evidence for programmed cell death and activation of specific caspase-like
enzymes in the tomato fruit heat stress response. Planta 229:1269-1279 (2009).

This article is protected by copyright. All rights reserved.

 16 Hurr B, Huber DJ and Vallejos CE, Features of programmed cell death precede
watersoaking development in ethylene-treated immature cucumber fruit. Postharv
Biol Tech 58:13-20 (2010).
17 Hurr BM, Huber DJ, Vallejos CE, Lee E and Sargent SA, Ethylene-induced
overproduction of reactive oxygen species is responsible for the development of
watersoaking in immature cucumber fruit. J Plant Physiol 170:56-62 (2013).
Accepted Article
18 Lee JS, Hurr BM, Huber DJ, Vallejos CE and Sargent SA, Characterization of
proteases and nucleases associated with ethylene-induced programmed cell
death in immature cucumber fruit. Postharv Biol Tech 110:190-196 (2015).
19 Zhao Y, Chen J, Tao X, Zheng X and Mao L, The possible role of BAX and BI-1
genes in chilling induced cell death in cucumber fruit. Acta Physiol Plant 36:
1345-1351 (2014).
20 Chen J, Zhao Y, Chen X, Peng Y, Hurr BM and Mao L, The role of ethylene and
calcium programmed cell death of cold stored cucumber fruit. J Food Biochem
38:337-344 (2014).
21 Simmonds NW, The development of the banana fruit. J Exp Bot 4:87-105. (1953).
22 Turner DW, Bananas and Plantains, in: Postharvest and storage of tropical and
subtropical fruits, ed. by Mitra S. CAB International, Oxon, pp:47-83 (1997).
23 Chen CR and Ramaswamy HS, Color and texture change kinetics in ripening
bananas. LWT Food Sci Tech 35:415-419 (2002).
24 Bapat VA, Trivedi PK, Ghosh A, Sane VA, Ganapathi TR and Nath P, Ripening of
fleshy fruit: Molecular insight and the role of ethylene. Biotech Adv 8:94-107
(2010).
25 Thompson AK, Banana (Musa spp.). In: Postharvest biology and technology of
tropical and subtropical fruits. Volume 2: Açai to citrus, ed by Yahia EM.
Woodhead Publishing, Cambridge, pp: 216-243 (2011).
26 Uthaichay N, Ketsa S and van Doorn WG, Effect of gas composition and relative
humidity on senescent spotting in banana (Musa acuminata AA Group) Fruit cv.
Sucrier. Thai J Agric Sci 3-4:123-127 (2005).
27 Choehom R, Ketsa S and van Doorn WG, Senescent spotting of banana peel is
inhibited by modified atmosphere packaging. Postharv Biol Tech 31:167-175
(2004).
28 Trakulnaleumsai C, Ketsa S and van Doorn WG, Temperature effects on peel
spotting in ‘Sucrier’ banana fruit. Postharv Biol Tech 39:285-290 (2006).
29 Maneenuam T, Ketsa S and van Doorn WG, High oxygen levels promote peel
spotting in banana fruit. Postharv Biol Tech 43:128-132 (2007).
30 Moser S, Müller T, Holzinger A, Lütz C, Jockusch S, Turro NJ and Kräutler B,
Fluorescent chlorophyll catabolites in bananas light up blue halos of cell death.
Proc Nat Assoc Sci 106:15538-15543 (2009).
31 Zhang Z, Huber DJ and Rao J, Ripening delay of mid-climacteric avocado fruit in
response to elevated doses of 1-methylcyclopropane and hypoxia-mediated
reduction in internal ethylene concentration. Postharv Biol Tech 60:83-91 (2011).
32 Villalta A and Sargent SA, Response to beit alpha-type cucumbers (Cucumis sativus
L., ‘Manar’) to continuous ethylene exposure. Proc Fla State Hort Soc 117:368-
372 (2004).

This article is protected by copyright. All rights reserved.

 33 Bradford MM, A rapid and sensitive for the quantitation of microgram quantities of
protein utilizing the principle of protein-dye binding. Ann Biochem 72:248-254
(1976).
34 Sousa F, Jus S, Erbel A, Kokol V, Cavaco-Paulo A and Gubitz GM, A novel
metalloprotease from Bacillus cereus for protein fibre processing. Enz Microb
Tech 40:1772-1781 (2007).
Accepted Article
35 Azeez A, Sane AP, Bhatnagar D and Nath P, Enhanced Expression of serine
proteases during floral senescence in Gladiolus. Phytochem 68:1352-1357.
(2007).
36 Bronstein JC and Weber PC, A colorimetric assay for high-throughput screening of
inhibitors of herpes simplex virus type 1 alkaline nuclease. Anal Biochem
293:239-245 (2001).
37 Thelen MP and Northcote DH, Identification and purification of a nuclease from
Zinnia elegans L.: a potential molecular marker for xylogenesis. Planta 179:181-
195 (1989).
38 Domínguez Mand Vendrell M, Effect of ethylene treatment on ethylene production,
EFE activity and ACC levels in peel and pulp of banana fruit. Postharv Biol Tech
4:167-177 (1994).
39 Ke LS and Tsai PL, Changes of ACC contents and EFE activity in peel and pulp of
banana fruit during ripening in relation to ethylene production. J Agr Assoc Chin
143:48-60 (1988).
40 von Loesecke HW, Bananas. 2nd Edition. InterScience Publishers, New York (1950).
41 Wang Y, Luo Z, Du R, Liu Y, Ying T and Mao L, Effect of nitric oxide on antioxidative
response and proline metabolism in banana during cold storage. J Agric Food
Chem 61:8880-8887 (2013).
42 Demidchik V, Straltsova D, Medvedev SS, Pozhvanov GA, Sokolik A and Yurin V,
Stress-induced electrolyte leakage: the role of K+-permeable channels and
involvement in programmed cell death and metabolic adjustment. J Exp Bot
DOI:10.1093/jxb/eru004 (2014).
43 Vicente AR, Martínez GA, Chaves AR and Civello PM, Effect of heat treatment on
strawberry fruit damage and oxidative metabolism during storage. Postharv Biol
Tech 40:116-122 (2006).
44 Zhivotovski B and Irrenius S, Calcium and cell death mechanisms: a perspective
from the cell death community. Cell Calcium 50:211-221 (2011).
45 Stratton FC and von Loesecke H, Changes in osmotic pressure of bananas during
ripening. Plant Physiol 6:361-365 (1931).
46 Burg SP, Postharvest physiology and hypobaric storage of fresh produce. CAB
International, London (2004).
47 Amnuaysin N, Seraypheap K and Kidyoo M, Anatomical changes in peel structure of
‘Hom Thong’ banana during fruit development and ripening. Trop Nat Hist
12:127-136 (2012).
48 Ratule MT, Osman A, Saari N and Ahmad SH, Microstructure of peel cell wall and
selected physico-chemical characteristics of ‘Berangan’ banana (Musa cv.
Berangan [AAA]) ripened at high temperature. Asia Pacific J Mol Biol BioTech
15:8-13 (2007).

This article is protected by copyright. All rights reserved.

 49 Kesari R, Trivedi PK, Nath P, Ethylene-induced ripening in banana evokes
expression of defense and stress related genes in fruit tissue. Postharv Biol Tech
46:136-143 (2007).
50 Baker DA, Kallarackal J and Milburn JA, Water relations of the banana. II.
Physicochemical aspects of the latex and other tissue fluids. Aus J Plant Phys
17:69-77 (1990).
Accepted Article
51 Finger, FL, Puschmann R and Santos Barros R, Effects of water loss on respiration,
ethylene production and ripening of banana fruit. R Bras Fisiol Veg 7:115-118
(1995).
52 Lara I, Belge B and Goulao LF, The fruit cuticle as a modulator of postharvest
quality. Postharv Biol Tech 87:103-111 (2014).
53 Klieber A, Bagnato N, Barrett, R and Sedgley M, Effect of post-ripening nitrogen
atmosphere storage on banana shelf life, visual appearance and aroma.
Postharv Biol Tech 25:15-24 (2002).
54 Baez-Sañudo M, Siller-Cepeda J, Muy-Rangel D and Heredia JB, Extending the
shelf-life of bananas with 1-methylcyclopropene and a chitosan-based edible
coating. J Sci Food Agric 89:2343-2349 (2009).
55 Coll NS, Epple P and Dangl JL, Programmed cell death in the plant immune system.
Cell Death Diff 18:1247-1256 (2011).
56 Hofius D, Tsitsigiannis DI, Jones JDG and Mundy J, Inducible cell death in plant
immunity. Sem Canc Biol 17:166-187 (2007).
57 Johnson BE and Brun WA, Stomatal density and responsiveness of banana fruit
stomates. Plant Physiol 41:99-101 (1966).

This article is protected by copyright. All rights reserved.

Accepted Article

Figure
e 1. (A) Rip
pening and over-ripening of bana
ana fruit at 2
20 °C and 8
80-85% RH
H.

(B) Etthylene and

d carbon dio
oxide produ
uction ratess during ripe
ening and o
over-ripenin
ng

of ban
nana fruit sttored at 20 °C and 80--85% RH. M
Mean ± standard errorr of 3

measurements. When
W abse
ent, standarrd error barrs fall within nsions of the
n the dimen

symbo
ol.

This article is protected by copyright. All rights reserved.

Accepted Article

Figure
e 2. Total electrolyte
e leakage
l and ion leaka
age of bana
ana peel from ripening

and over-ripenin
o g banana fruit
f stored at
a 20 °C an
nd 80-85% RH. Fruit w
were treated
d

with 150-300 µL L-1 of ethyllene for 48 h at 17 °C prior to sto

orage. Mean
n ± standarrd

error of
o 3 measu
urements. When
W absen
nt, standard
d error barss fall within the

dimen
nsions of th
he symbol.

This article is protected by copyright. All rights reserved.

Accepted Article

Figure
e 3. Peel structure
s of banana peel tissue exxcised befo
ore ethylene
e treatmentt

(ET) and
a after ET 85% RH, and
T at 2 d, 6 d and 10 d during storrage at 20 °°C and 80-8

after exposure
e to
o freeze and thaw at 6 d after ET to induce ccell death. Bright field

microscopy imag
ges of cross
s-sections (500
( nm thiick) of fresh
h tissue sta
ained with 10

g L-1 toluidine
t blu
ue. (A) Green, prior to ethylene trreatment (E
ET), (B) 2 d after ET,

This article is protected by copyright. All rights reserved.

 (C) 6 d after ET, (D) 6 d after ET exposed to freeze and thaw, (E) 10 d after ET

non-senescence-related dark spot, (F) 10 d after ET senescence-related dark spot.

Chlorenchyma ({), epidermis (→), fiber cells (*), void spaces (º). Black horizontal
Accepted Article
bars on the images indicate 120 μm.

This article is protected by copyright. All rights reserved.

Accepted Article

Figure 4. Nuclei of epidermal cells of banana peel tissue excised before ethylene

treatment (ET) and after ET at 2 d, 6 d, and 10 d during storage at 20 °C and 80-

85% RH, and after exposure to freeze and thaw on day 6 after ET to induce cell

death. TEM images from cross-sections (90 nm-thin) stained with 20 g L-1 uranyl

acetate and Reynold’s lead citrate. (A) Green, prior to ethylene treatment (ET), (B)

2 d after ET, (C) 6 d after ET, (D) Freeze and thaw 6 d after ET, (E) 10 d after ET

This article is protected by copyright. All rights reserved.

 non-senescence-related dark spot, (F) 10 d after ET senescence-related dark spot.

Bars represent 5 μm. N nucleus. Nuclei in figures D and F are not labeled since

they cannot be resolved.

Accepted Article

This article is protected by copyright. All rights reserved.

Accepted Article

Figure
e 5. Nuclea y (units kg-1 s-1) and pro
ase activity otease actiivity (units kkg-1 protein s-
1
) of banana
b pee
el during rip ng of banan
pening and over-ripenin na fruit storred at 20 °C
C

and 80-85%
8 RH. Mean ± sttandard erro
or of 3 mea
asurementss.

This article is protected by copyright. All rights reserved.

Accepted Article

e 6. In-gel nuclease activity

Figure a assa
ay of peel d
during ripen
ning and ove
er-ripening

nana fruit sttored at 20 °C and 80--85% RH. ((A) Protein (1 μg) wass resolved
of ban

using SDS-PAGE in 120 g L-1 polyacry ng 30 μg mL

ylamide gells containin L-1 heat-

denattured herrin
ng sperm DNA as in-ge
el substrate
e. Gels werre incubated
d at 37 °C ffor

on buffer (10 mmol L-1 MOPS-Na

2 h in renaturatio aOH, pH 7, 1 mmol L-1 CaCl2, 1

mmol L-1 MgCl2). (B) Signall intensity of

o nuclease activity of tthe 4 nucle
eases

This article is protected by copyright. All rights reserved.

 detected by SDS-PAGE, nucleases MaN25, MaN27, MaN29, and MaN31 had

apparent sizes of ∼25 kDa, ∼27 kDa, ∼29 kDa, and ∼31 kDa, respectively. Mean ±

standard error of 3 measurements. When absent, standard error bars fall within the
Accepted Article
dimensions of the symbol.

This article is protected by copyright. All rights reserved.

Accepted Article

Figure
e 7. In-gel protease activity assa
ay of peel during ripening and ove
er-ripening of

banan
na fruit storred at 20 °C
C and 80-85 5 μg) was rresolved
5% RH. (A) Protein (15

using SDS-PAGE in 80 g L-1

-
g 1 g L-1 gellatin as in-g
polyacrylamide gelss containing gel

This article is protected by copyright. All rights reserved.

 substrate. Gels were incubated at 37 °C for 16 h in renaturation buffer (50 mmol L-1

Tris-HCl, pH 7.5, 10 mmol L-1 CaCl2, 10 mmol L-1 MgCl2). (B) Signal intensity of

protease activity of the 6 proteases detected by SDS-PAGE, proteases MaP79,

Accepted Article
MaP86, MaP123, MaP151, MaP169, and MaP192 had apparent sizes of ∼79 kDa,

∼86 kDa, ∼123 kDa, ∼151 kDa, ∼169 kDa, and ∼192 kDa, respectively. Mean ±

standard error of 3 measurements. When absent, standard error bars fall within the

dimensions of the symbol.

This article is protected by copyright. All rights reserved.