Journal of Experimental Botany, Vol. 69, No. 6 pp.

1387–1402, 2018
doi:10.1093/jxb/erx455  Advance Access publication 4 January 2018
This paper is available online free of all access charges (see http://jxb.oxfordjournals.org/open_access.html for further details)

RESEARCH PAPER

Autophagy is activated and involved in cell death with

participation of cathepsins during stress-induced microspore
embryogenesis in barley
Ivett Bárány1,*, Eduardo Berenguer1,*, María-Teresa Solís1, Yolanda Pérez-Pérez1, M. Estrella Santamaría2,
José Luis Crespo3, María C. Risueño1, Isabel Díaz2 and Pilar S. Testillano1,†

Downloaded from https://academic.oup.com/jxb/article/69/6/1387/4791010 by guest on 17 June 2021

1 
Biological Research Center, CIB, CSIC, Ramiro de Maeztu 9, 28040 Madrid, Spain
2 
Center of Plant Biotechnology and Genomics, CBGP, UPM, Autovía M40, km 38, Pozuelo de Alarcón, 28223 Madrid, Spain
3 
Institute of Plant Biochemistry and Photosynthesis, IBVF, CSIC, Avenida Américo Vespucio 49, 41092 Seville, Spain

*These authors contributed equally to this work.

† 
Correspondence: [email protected]
Received 30 August 2017; Editorial decision 29 November 2017; Accepted 30 November 2017

Editor: Zoe Wilson, University of Nottingham, UK

Abstract
Microspores are reprogrammed towards embryogenesis by stress. Many microspores die after this stress, limiting the
efficiency of microspore embryogenesis. Autophagy is a degradation pathway that plays critical roles in stress response
and cell death. In animals, cathepsins have an integral role in autophagy by degrading autophagic material; less is known
in plants. Plant cathepsins are papain-like C1A cysteine proteases involved in many physiological processes, including
programmed cell death. We have analysed the involvement of autophagy in cell death, in relation to cathepsin activation,
during stress-induced microspore embryogenesis in Hordeum vulgare. After stress, reactive oxygen species (ROS) and
cell death increased and autophagy was activated, including HvATG5 and HvATG6 up-regulation and increase of ATG5,
ATG8, and autophagosomes. Concomitantly, cathepsin L/F-, B-, and H-like activities were induced, cathepsin-like genes
HvPap-1 and HvPap-6 were up-regulated, and HvPap-1, HvPap-6, and HvPap-19 proteins increased and localized in the
cytoplasm, resembling autophagy structures. Inhibitors of autophagy and cysteine proteases reduced cell death and
promoted embryogenesis. The findings reveal a role for autophagy in stress-induced cell death during microspore em-
bryogenesis, and the participation of cathepsins. Similar patterns of activation, expression, and localization suggest a
possible connection between cathepsins and autophagy. The results open up new possibilities to enhance microspore
embryogenesis efficiency with autophagy and/or cysteine protease modulators.

Keywords:   Autophagy, barley, caspase-like activity, cathepsins, cell death, cysteine C1A proteases, microspore
embryogenesis, ROS, stress response.

Introduction
Plant cell plasticity and ability to regenerate embryos in of genetic resources (Germaná and Lambardi, 2016).
in vitro culture have been extensively exploited for decades, in In vitro embryogenesis is a fascinating system to study cellular
the areas of plant propagation, breeding, and conservation reprogramming and acquisition of totipotency, as well as an

Abbreviations: ATG, autophagy-related protein; 3-MA, 3-methyladenine; MDC, monodansylcadaverine; PCD, programmed cell death; ROS, reactive oxygen
species.

© The Author(s) 2018. Published by Oxford University Press on behalf of the Society for Experimental Biology.
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 reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
1388 | Bárány et al.

alternative to study early embryogenesis events since zygotes Hirt, 2004). Recent studies in plants and algae have described
and immature embryos produced in planta are surrounded the activation of autophagy in response to several stress con-
by maternal tissues and are difficult to analyse. Microspore ditions that increase ROS production (Liu and Bassham,
embryogenesis is an in vitro system in which the haploid 2012; Pérez-Pérez et al., 2010, 2012). These findings suggest
microspore is reprogrammed by the application of external a strong link between autophagy and ROS production in
stress treatment and enters into an embryogenesis pathway plants.
(Bárány et al., 2005; Prem et al., 2012). The resulting haploid Autophagy and cell death proteases are well characterized
and double-haploid embryos and generated plants are import- in animals, and caspases are thought to be the major pro-
ant biotechnological tools in plant breeding for the rapid gen- teases involved. Although to date no functional homologues
eration of isogenic new varieties as they represent a source of animal caspases have been identified in plants, several in-
of new genetic variability, fixed in complete homozygous direct pieces of evidence suggest the existence of functionally
plants and obtained in only one generation step (Maluszynski related proteases with similar substrate specificity. The in-
et al., 2003). Despite the usefulness of stress-induced in vitro volvement of caspase-3-like enzymatic activity in plant PCD
embryogenesis in breeding programmes, the efficiency of has been well documented, and its specific inhibitors block
the system in many species of economic interest is still lim- completion of PCD (Bonneau et al., 2008; Solís et al., 2014),

Downloaded from https://academic.oup.com/jxb/article/69/6/1387/4791010 by guest on 17 June 2021

ited since it is greatly affected by many factors (El-Tantawy although the identity of the protease(s) responsible has not
et al., 2013; Rodríguez-Sanz et al., 2014, 2015; Solís et al., yet been fully resolved. Recently, it has been reported that
2012, 2015, 2016; Testillano et al., 2010), and primarily by caspase-3 inhibitors reduce PCD in Arabidopsis by targeting
the occurrence of cell death induced by the stress applied to cathepsin B, another major plant protease involved in PCD
trigger embryogenesis. In barley, microspore embryogenesis is (Ge et al., 2016).
efficiently induced by cold stress treatment in isolated micro- Cathepsins are papain-like C1A cysteine proteases, as clas-
spore cultures (Rodríguez-Serrano et al., 2012). sified in the MEROPS peptidase database (Rawlings et  al.,
Autophagy is a universal degradation pathway in all 2016). They are well known lysosomal proteases with a role in
eukaryotes, including plants, that recycles cell materials upon autophagy and cell death, in animals (Turk and Stoka, 2007).
stress conditions or during specific developmental processes, It is well documented in animals that cathepsins are respon-
thereby promoting cell survival (reviewed in Hofius et al., sible for driving proteolytic degradation in lysosomes and
2017; Masclaux-Daubresse et al., 2017). In addition to this have a critical role in the terminal degradation of proteins
survival role, autophagy can also play critical roles as a cell within autolysosomes, following the autophagosome fusion
death initiator and/or executioner. Increasing evidence indi- (Jung et  al., 2015; Kroemer and Jäättelä, 2005; Man and
cates the involvement of autophagy in plant cell death (Minina Kanneganti, 2016). In plants, this enzyme group with proteo-
et al., 2013, 2014b; Yang and Bassham, 2015). In Picea abies lytic activity is involved in many physiological processes such
embryos, it has been demonstrated that autophagy is respon- as senescence, abscission, fruit ripening, and PCD, and in the
sible for cell self-disassembly during programmed cell death mobilization of proteins accumulated in seeds and tubers
(PCD) (Minina et al., 2013). In plants, autophagy-directed Martínez et  al., 2012; Díaz-Mendoza et  al., 2014, 2016).
degradation of cellular components occurs mainly in vacu- Moreover, C1A proteases actively participate in proteolysis
oles. It is initiated by the engulfment of subcellular compo- induced by biotic and abiotic stresses (Díaz-Mendoza et al.,
nents into a double membrane structure, the autophagosome, 2014; Velasco-Arroyo et  al., 2016), but less is known about
the outer membrane of which further fuses with the vacuole the possible role of plant cathepsins in autophagosome deg-
membrane, the tonoplast. This results in the release of the radation. Plant cathepsins are grouped as cathepsin L-, B-,
so-called autophagic bodies (single-membrane structures H-, and F-like according to their gene features and phylogen-
containing the cargo) at the vacuole interior where degrad- etic relationship. The activity of these proteases is regulated
ation takes place via the activity of lytic enzymes. In different by specific inhibitors, termed phytocystatins (Martínez et al.,
plant species, autophagosomes either can fuse directly with 2009; Díaz-Mendoza et al., 2016), and is also the target of syn-
the central vacuole or can first fuse with a smaller vacuole thetic exogenous inhibitors such as E-64. In barley, both C1A
or lysosome-like organelle, which begins content degradation proteases and cystatins have been studied in depth (Martínez
(Bassham, 2007). Activation of autophagy involves the in- et al., 2009, 2012; Díaz-Mendoza et al., 2014, 2016). However,
duction of AuTophaGy-related ATG genes and activation of there is no information on the action of these proteases dur-
specific proteases. In the case of barley, 25 ATG genes have ing stress-induced microspore embryogenesis.
been characterized (Avila-Ospina et al., 2016; Masclaux- During stress-induced microspore embryogenesis, we have
Daubresse et al., 2017). Among them, ATG5, ATG6, and reported increasing cell death levels and caspase-3-like activ-
ATG8 proteins play crucial roles in autophagosome forma- ity, after the inductive stress to trigger microspore reprogram-
tion (Li and Vierstra, 2012; Michaeli et al., 2016). ming, in barley (Rodríguez-Serrano et al., 2012). Nevertheless,
Plant cells actively produce reactive oxygen species (ROS) prior to the present work, no studies had been developed on
at low levels but, as a response to stress, cell production of the participation of autophagy during the induction of micro-
ROS increases. ROS act as signalling molecules to control spore embryogenesis mediated by stress in barley.
processes such as PCD and stress response. Excessive ROS In this work, we have studied the involvement of autophagy
levels may cause irreversible oxidative damage and activate in cell death occurrence during stress-induced microspore
signalling pathways ultimately leading to cell death (Apel and embryogenesis of barley, in relation to cathepsin activation.
 Autophagy and cathepsins in stress-induced microspore embryogenesis  |  1389

The results indicated that autophagy is induced in micro- Table 1.  Inhibitors and conditions used for in vitro treatments of
spores after the inductive stress. Concomitantly, cathepsins microspore cultures
are also activated and show similar patterns of expression
and localization to ATGs. Inhibition of autophagy and cath- Name Concentration Brand
epsins reduced cell death levels and increased the embryogen-
MnCl2 4 mM Merck
esis induction rate. Taken together, the results indicate a role Ac-DEVD-CHO Sigma
5 µM
for autophagy in cell death at early stages of stress-induced 3-MA 5 mM Sigma
microspore embryogenesis, a death process in which cath- E-64 1µM Sigma
epsins also participate. ConA 1 µM Sigma

MnCl2, manganese chloride; Ac-DEVD-CHO, N-acetyl-l-α-aspartyl-

Materials and methods l-α-glutamyl-N-(2-carboxyl-1-formylethyl)-l-valinamide; 3-MA,
3-methyladenine; E-64, trans-epoxysuccinyl-l-leucylamido(4-guanidino)
Plant material and in vitro microspore embryogenesis culture butane; ConA, concanamycin A. 
Winter barley cultivars (Hordeum vulgare L.  cv. Igri) were used as
donor plants. Seeds were vernalized in soil for 1 month at 4 °C, grown

Downloaded from https://academic.oup.com/jxb/article/69/6/1387/4791010 by guest on 17 June 2021

(Supplementary Method S1). ATG8 antibody was produced in
at 12 °C with a 12 h light/12 h dark photoperiod (10 000–16 000 lux) for the laboratory of Dr J.L. Crespo (IBFV, Seville, Spain), and was
1 month in a growth chamber (Sanyo MLR-351-H, relative humidity reported to recognize specifically Arabidopsis ATG8 (Álvarez et al.,
70%), and then grown in a greenhouse under a controlled temperature 2012; Pérez-Pérez et al., 2010).
of 18 °C. In vitro cultures of isolated microspores, the most responsive
stage for embryogenesis induction, were performed by stress treatment
of 4 °C, as previously described (Rodríguez-Serrano et al., 2012). Fixation and processing for light and electron microscopy
analyses
Cell death detection For light microscopy, in vitro samples were collected and fixed in
4% paraformaldehyde in phosphate-buffered saline (PBS), pH 6.8,
Microspore culture samples were incubated with a 0.25% (w/v)
overnight at 4 °C. Samples were dehydrated in acetone and embed-
aqueous solution of Evans Blue for 30  min and observed with a
ded in Technovit 8100 resin (Kulzer, Germany) at 4 °C. Semi-thin
light microscope under bright field. Mean percentages of dead cells
resin sections were either stained with toluidine blue and observed
(stained by Evans Blue) were quantified in random micrographs
under bright field, for structural analysis, or stored at 4 °C and used
from two replicates of three independent experiments, as described
for immunofluorescence (Solís et  al., 2016). For electron micros-
(Solís et al., 2015).
copy, samples were fixed in Karnowsky fixative (4% paraformalde-
hyde, 5% glutaraldehyde in 0.025 M cacodylate buffer with 0.5 mg
ROS detection ml–1 calcium chloride) for 4 h at room temperature, post-fixed in 2%
osmium tetroxide, dehydrated in an ethanol series and propylene
Microspore culture samples were incubated for 1 h, in the dark, with oxide, and embedded in Epon 812 resin. Ultrathin sections were
10  µM dihydroethidium (DHE) to detect ROS, specifically super- counterstained by uranyl acetate and lead citrate, and examined in
oxide radicals (Rodríguez-Serrano et al., 2012). As negative control, an electron microscope (JEOL JEM 2100).
samples were incubated for 1 h before DHE with 4 mM MnCl2, an
O2– scavenger. After washing, samples were immediately observed
with a confocal microscope (Leica TCS SP5) and signal captured as Immunofluorescence and confocal microscopy
red fluorescence (490 nm excitation; 520 nm emission). Semi-thin sections were blocked by 5% (w/v) BSA and incubated for
1 h with the corresponding primary polyclonal antibody diluted in
Treatments with inhibitors 1% BSA at 1:100 (HvPap-1), 1:50 (HvPap-6), 1:20 (HvPap-19), and
1:50 (ATG5 and ATG8). After washing in PBS, signal was revealed
At the time of culture initiation, several inhibitors (Table  1) were with Alexa Fluor 488-labelled anti-rabbit IgG antibody (Molecular
added to the microspore culture plates. Ac-DEVD-CHO, E-64, Probes) diluted 1:25 in 1% BSA for 45 min in darkness. Finally, sec-
and concanamycin A (ConA) were added from concentrated stock tions were counterstained with 1 mg ml−1 DAPI for 10 min and ana-
solutions dissolved in ethanol (E-64) and DMSO (Ac-DEVD-CHO lysed by confocal laser microscopy (Leica TCS-SP5-AOBS, Vienna,
and ConA). Controls of solvent effects were performed by add- Austria). Images of maximum projections were obtained with soft-
ing the same volumes of ethanol or DMSO to untreated cultures. ware of the confocal microscope (Leica software LCS version 2.5).
3-Methyladenine (3-MA) and MnCl2 were directly dissolved in the Negative controls were performed avoiding the primary antibody.
culture medium. Short treatments were carried out from culture ini-
tiation during 4 d.  Mean percentages of ‘proembryos’ were quan-
tified from random samples of three independent experiments, as MDC in vivo staining
previously described (Berenguer et al., 2017). Microspore samples from untreated and inhibitor-treated cultures
were stained with 0.05 mM monodansylcadaverine (MDC; Sigma-
Aldrich), at room temperature for 30  min in darkness (Contento
Antibodies et al., 2005). After incubation, cells were washed twice with PBS and
Cathepsin antibodies against HvPap-1, HvPap-6, and Hv-Pap-19 immediately observed by confocal microscopy (Leica TCS SP5).
proteases were produced in rabbits by Pineda Antibody Services Fluorescence of intracellular MDC was observed selecting wave-
(Berlin, Germany), against specific peptide sequences of each pro- lengths of 405 nm for excitation and 435–483 nm for emission.
tease (Supplementary Table S1 at JXB online); they were previously
produced and reported to recognize these proteases in barley leaves
(Díaz-Mendoza et al., 2016). ATG5 antibody was kindly provided Quantitative real-time PCR analysis (RT-qPCR)
by Dr M.F. Suárez (University of Málaga, Spain), produced in rab- Total RNA was extracted from in vitro samples using the RNeasy®
bits with purified recombinant ATG5 protein of spruce as antigen Plant Micro and RNeasy® Plant Mini kits (Qiagen) according to
1390 | Bárány et al.

the manufacturer’s instruction. cDNAs were obtained from 2 µg of so-called microspore-derived ‘proembryos’, were produced
RNA using the Superscript™ II reverse transcriptase (Invitrogen) (Fig.  1B). In 4 d cultures, proembryos were accompanied
according to Solís et al. (2012). RT-qPCR analyses were performed by non-responsive and dead microspores (Fig.  1B). During
using the SsoAdvanced™ Universal SYBR®Green Supermix on
the iQ™5 Real-Time PCR Detection Sytem (Biorad). The oligonu- the following days of culture, microspore embryogenesis
cleotides used are described in Supplementary Table S2, and qPCR progressed, the exine broke down, and embryos developed
conditions were as previously described (Berenguer et al., 2017). All (Fig. 1C) and followed a pathway similar to zygotic embryo-
qPCRs were run in duplicate, and the Cyclophilin gene was used as genesis in monocot species, producing transitional and scu-
the internal reference gene. Transcript levels were normalized to the tellar embryos and then, after 30 d in culture, coleoptilar
vacuolated microspore levels. Data were analysed with the Bio-Rad
CFX Manager 3.0 (3.01224.1015) (Biorad), using the Livak calcula- embryos (Fig. 1D).
tion method (Livak and Schmittgen, 2001). The percentage of dead cells, identified by positive Evans
blue staining (Fig.  2), was quantified at several culture
steps: ‘isolated microspores’ (microspores extracted from
Protein quantification and protease activities
spikes but not treated by stress), ‘stress-treated micro-
Total soluble proteins were extracted from in vitro samples accord-
ing to Velasco-Arroyo et al. (2016), using the method of Bradford
spores’ (isolated microspores after the inductive stress),
(1976) for protein quantification. Enzymatic activity assays were and ‘4 d cultures’ (stage of formation of the proembryo).

Downloaded from https://academic.oup.com/jxb/article/69/6/1387/4791010 by guest on 17 June 2021

performed as previously described (Velasco-Arroyo et  al., 2016) Results showed the occurrence of cell death in isolated
with minor modifications. Cathepsin L-/F-, B-, and H-like activities microspores, probably due to the isolation procedure and
were assayed using Z-FR-AMC (N-carbobenzoxy-Phe-Arg-AMC), the presence of dead cells in the spike. Dead cell levels sig-
Z-RR-AMC (N-carbobenzoxy-Arg-Arg-AMC), and Bz-FVR-
AMC (Bz-Phe-Val-Arg-AMC) substrates, respectively. The reaction
nificantly increased (P<0.05, ANOVA and Tukey’s tests)
was incubated at 28 °C for 1 h. All assays were carried out in du- after the stress treatment in stress-treated microspores and
plicate. Blanks were used to account for spontaneous breakdown in 4 d cultures (Fig. 2).
of substrates, and the results were expressed as nmol of hydrolysed ROS production was analysed by specific staining with
substrate per mg of protein per min (nmol min–1 mg–1). The system the fluorescence probe DHE which specifically reacts with
was calibrated with known amounts of AMC (7-amino-4-methyl-
coumarin) in a standard reaction mixture.
the intracellular superoxide anion (O2−) in vivo (Zhao et al.,
2003). Under confocal microscopy analysis, almost no signal
was detected in isolated microspores (Fig. 3A) or 4 d cultures
Immunoblot analysis (Fig.  3C), whereas an intense fluorescence was observed in
Protein extracts were prepared from in vitro cultures by grinding many stress-treated microspores (Fig.  3B). If stress-treated
microspore embryos in liquid nitrogen before the addition of 100 μl microspores were incubated with the ROS scavenger MnCl2,
of extraction buffer (150 mM NaCl, 50 mM sodium phosphate, pH
6.0, and 2 mM EDTA). Extracted proteins were quantified by the DHE staining did not provide any signal (Fig. 3D), confirm-
Bradford method (Bradford, 1976) with BSA as standard. Proteins ing the specificity of the probe for ROS. The exine showed
were separated on SDS–polyacrylamide gels (12%,w/v), electro- unspecific autofluorescence of different intensities in all cases
transferred onto nitrocellulose membranes, and blocked in 3% BSA (Fig. 3A'–D').
containing 0.5% Tween-20 for 1 h. Immunoblotting was performed To evaluate the effect on cell death of the elimination of
according to Velasco-Arroyo et  al. (2016) with the polyclonal
antibodies to cathepsins, at 1:2500, 1:2500, and 1:5000 dilutions ROS in stress-treated microspore cultures, treatments were
for HvPap-1, HvPap-6, and HvPap-12 respectively. Peroxidase- performed with the ROS scavenger MnCl2, which specifically
conjugated anti-rabbit IgG (Sigma), diluted at 1:20 000 (v/v), was scavenges superoxide anions. Quantification of dead cells,
used as a secondary antibody for detection with ECL Plus. by Evans blue staining, was carried out in MnCl2-treated
and untreated cultures 4 d after the inductive stress. Results
Data analysis showed a significant reduction (P<0.05, Student’s t-test) in
Statistical differences among several stages were tested by one-way cell death levels in microspore cultures treated with the ROS
ANOVA followed by Tukey’s multiple comparison tests. Significant scavenger in comparison with untreated cultures (Fig. 4A),
differences between untreated and treated cultures were tested by indicating the involvement of ROS in cell death. As a conse-
Student’s t-test, in all cases with P≤0.05. quence, the proportion of proembryos formed in microspore
cultures treated with the ROS scavenger was significantly
higher (P<0.05, Student’s t-test) than in control cultures
Results (Fig. 4B).
Cell death occurrence, ROS production, and caspase- We have previously reported the induction of caspase-
3-like proteolytic activity in stress-induced microspore 3-like proteolytic activity in barley microspore cultures after
embryogenesis the inductive stress (Rodríguez-Serrano et al., 2012). In the
present study, to analyse the role of this enzymatic activity,
Microspore embryogenesis was induced by cold stress (4 °C) treatments with Ac-DEVD-CHO, a specific inhibitor of cas-
in barley using isolated microspore in vitro cultures, as previ- pase-3 activity, were performed in stress-treated microspores.
ously reported (Rodríguez-Serrano et al., 2012). Vacuolated The effects of the treatment with Ac-DEVD-CHO on cell
microspores (Fig. 1A), the most responsive stage for embryo- death and embryogenesis induction were evaluated in con-
genesis induction, were subjected to the inductive stress trol and treated cultures. The proportion of dead cells after
treatment. Four days after induction and culture initiation, the inductive stress was significantly lower (P<0.05, Student’s
multicellular structures still surrounded by the exine, the t-test) in microspore cultures treated with the inhibitor than
 Autophagy and cathepsins in stress-induced microspore embryogenesis  |  1391

Downloaded from https://academic.oup.com/jxb/article/69/6/1387/4791010 by guest on 17 June 2021

Fig. 1.  Stress-induced microspore embryogenesis in Hordeum vulgare. Micrographs of toluidine blue-stained semi-thin sections for general structural
analysis. (A) Vacuolated microspore at culture initiation. (B) Proembryos on microspore culture 4 d after stress, still surrounded by the exine. (C) Early
transitional embryo. (D) Microspore-derived embryos after 30 d in culture observed under the stereomicroscope. Ex, exine. Scale bars represent in (A–C)
20 µm, in (D) 10 mm.

Fig. 2.  Cell death in stress-induced microspore embryogenesis. Histogram showing the percentage of cell death after cell isolation and inductive stress
identified by Evan’s blue staining. Micrographs showing dead microspores as blue cells after Evan’s blue staining. Bars in columns indicate the SE. Scale
bars in micrographs represent 60 µm. Different letters on columns indicate significant differences among stages, according to ANOVA and Tukey’s tests
at P<0.05.
1392 | Bárány et al.

Downloaded from https://academic.oup.com/jxb/article/69/6/1387/4791010 by guest on 17 June 2021

Fig. 3.  ROS staining during stress-induced microspore embryogenesis. Specific staining with dihydroethidium (DHE). Confocal laser scanning
microscopy analysis of (A, A') isolated microspore, (B, B') stress-treated microspore, (C, C') 4 d culture proembryo, (D, D') stress-treated microspore after
incubation with MnCl2 (O2– scavenger). Scale bars represent 25 µm.

Fig. 4.  Effect of treatments with MnCl2 (O2– scavenger) and Ac-DEVD-CHO (caspase-3 inhibitor) in stress-induced microspore embryogenesis.
(A, C) Quantification of cell death levels, identified by Evan’s blue staining, 4 d after stress in untreated microspore cultures and cultures treated with
MnCl2 (A) and Ac-DEVD-CHO (C). (B, D) Quantification of proembryos (as an indicator of microspore embryogenesis initiation) in microspore cultures, 4 d
after stress, treated with MnCl2 (B) and Ac-DEVD-CHO (D). In all histograms, results are expressed as percentages (percent change) and referred to the
mean percentage of dead cells or proembryos in control cultures which has been normalized to 100%. Bars indicate the SE. Asterisks indicate significant
differences between treated and untreated cultures, within each treatment, assessed by Student’s t-test, at P<0.05.

in control cultures (Fig. 4C). After 4 d in culture, the num- the proportion of proembryos produced in microspore cul-
ber of proembryos was determined as an indicator of ini- tures treated with the caspase-3 inhibitor in comparison with
tiation of microspore embryogenesis. The results showed a control cultures (Fig. 4D), probably as a consequence of the
statistically significant increase (P<0.05 Student’s t-test) in reduction in cell death.
 Autophagy and cathepsins in stress-induced microspore embryogenesis  |  1393

ATG gene expression and protein localization in stress- 2005). Microspores were stained by MDC and analysed by
induced microspore embryogenesis confocal microscopy. Stress-treated microspores showed
strong MDC fluorescence on small spherical cytoplasmic
Among the 25 ATG genes characterized in barley (Avila- spots (Fig. 7A, A'). These spots were of different sizes, and
Ospina et  al., 2016; Masclaux-Daubresse et  al., 2017), to were occasionally observed within vacuoles, and, there-
evaluate the possible activation of autophagy in microspore fore, they most probably corresponded to autophagosomes
embryogenesis cultures after the inductive stress, expres- and autophagic bodies. Electron microscopy provided evi-
sion analyses were conducted for two key autophagy genes, dence of autophagic structures in stress-treated microspores.
HvATG5 and HvATG6, identified in barley with only one gene Ultrastructural analysis revealed the presence of early and
isoform each (Avila-Ospina et al., 2016). RT-qPCRs showed mature autophagosomes. Autophagosomes at an early stage
similar expression patterns for both ATG genes: a low level of of their formation appeared as double-membrane structures
expression in isolated microspores, before the stress, and high with semi-dense content, similar to the cytoplasm (Fig.  7B,
gene expression in stress-treated microspores (Fig. 5). Later, inset). Advanced/mature autophagosomes that had engulfed
in 4 d microspore cultures, expression decreased, dropping to cytoplasmic structures/organelles showed double- and multi-
levels similar to those seen with isolated microspores (Fig. 5). ple-membrane structures, with organelle and membrane rem-
ATG5 and ATG8 proteins, which have a crucial role in

Downloaded from https://academic.oup.com/jxb/article/69/6/1387/4791010 by guest on 17 June 2021

nants in their interior (Fig. 7B).
autophagy (Bassham, 2009; Li and Vierstra, 2012; Michaeli 3-MA is an inhibitor of phosphatidylinositol 3-kinase
et  al., 2016), were localized by using specific antibodies. (PtdIns3K) involved in the formation of the autophagosome
Immunofluorescence assays and confocal analyses on semi- (Li and Vierstra, 2012). 3-MA has been reported to block
thin sections showed no labelling in isolated microspores with autophagosome formation in tobacco BY2 culture cells, at
either ATG5 or ATG8 antibodies (Fig. 6A, A', A''), whereas 5  mM concentration (Takatsuka et  al., 2004). To evaluate
in stress-treated microspores ATG5 and ATG8 localized in its effects on microspore cultures, 5  mM 3-MA was added
small punctuate cytoplasmic structures (Fig. 6B, B', B''). Four to the culture media of stress-treated microspores. Inhibition
days after the stress, no significant immunofluorescence label- of autophagy by 3-MA was measured in microspores after
ling was observed with either of these two autophagy anti- the inductive stress by quantifying the autophagosomes and
bodies (Fig. 6C, C', C''). autophagic bodies, as revealed by MDC, in untreated and
3-MA-treated cultures (Fig. 8A, C). Control assays without
Effects of treatments with the inhibitors of autophagy MDC staining in stress-treated microspores did not pro-
3-MA, E-64, and Con A on cell death occurrence and vide fluorescence to any subcellular structure, except for the
microspore embryogenesis initiation microspore wall, the exine, which exhibited unspecific auto-
fluorescence in all microspores (Fig. 8B). 3-MA-treated and
Functional analyses of autophagy in microspore embryogen-
untreated cultures showed cells with MDC-stained spots (one
esis were performed by treating microspore cultures in vitro
or two) and cells without any stained structures (Fig. 8A, C).
with 3-MA, E-64, and ConA, three drugs commonly used to
The results of the quantification showed a significant re-
inhibit autophagy in plants (Matsuoka et al., 1997; Takatsuka
duction of autophagy in microspores treated with 3-MA in
et  al., 2004; Sláviková et  al., 2005; Bassham, 2009, 2015;
comparison with control cultures (Student’s t-test, P<0.05),
Merkulova et al., 2014; Shin et al., 2014; Yano et al., 2015).
as revealed by the reduction in the proportion of cells with
Autophagosomes, autolysosome-like structures, and
autophagosomes (Fig.  9A) and the decrease in the mean
autophagic bodies can be detected by in vivo MDC staining in
number of autophagosomes per cell (Fig. 9B).
plant cell suspensions (Niemann et al., 2000; Contento et al.,
The cysteine protease inhibitor E-64 was also added
to microspore cultures after the inductive stress to trig-
ger embryogenesis. In many plant species, E-64 has been
reported to block autophagy at the step of autophagosome
degradation, which therefore leads to the accumulation of
autophagic bodies in vacuoles or smaller autolysosome-like
organelles in the cytoplasm (Bassham, 2007, 2015; Moriyasu
and Inoue, 2008). Microspore cultures treated with E-64
showed cells with higher numbers of MDC-stained spots in
their cytoplasms (up to 4–6 spots) compared with untreated
cultures, whose cells showed 0–2 spots per cell (Fig.  8A,
D). Moreover, quantitative analyses showed a significantly
higher proportion of cells with autophagic bodies (Fig. 9C),
Fig. 5.  Gene expression patterns of autophagy genes HvATG5 and as well as a significant increase in the mean number of
HvATG6 during stress-induced microspore embryogenesis. Histogram autophagic structures per cell (Fig. 9D) in E-64-treated cul-
showing relative changes of mRNA levels normalized to isolated
tures compared with untreated cultures (Student’s t-test,
microspore levels, as determined by RT-qPCR. Bars indicate the SE.
Different letters indicate significant differences among stages within P<0.05).
the expression of each gene according to ANOVA and Tukey’s tests at ConA, which inhibits vacuolar proton pumps and leads to
P<0.05. increased vacuolar pH and inactivation of acid hydrolases,
1394 | Bárány et al.

Downloaded from https://academic.oup.com/jxb/article/69/6/1387/4791010 by guest on 17 June 2021

Fig. 6.  Immunolocalization of autophagy proteins HvATG5 and HvATG8 during stress-induced microspore embryogenesis. Immunofluorescence and
confocal laser scanning microscopy analysis of isolated microspore (A–A'''), stress-treated microspore (B–B'''), and 4 d culture proembryo (C–C'''). (A–C,
A''–C'') Normarsky’s differential interference contrast (DIC). (A'–C', A'''–C''') Merged images of ATG immunofluorescence (green) and DAPI staining of
nuclei (blue). (A'–C') HvATG5. (A'''–C''') HvATG8. Scale bars represent in (A–B'') 10 µm, in (C–C'') 20 µm.

2014; Bassham, 2015; Yano et al., 2015). ConA treatment of

microspore cultures showed a similar effect to E-64 treatment.
In ConA-treated cultures, MDC staining revealed cells with
a higher number of autophagosomes (3–6) than untreated
cultures (Fig.  8A, E). The quantitative analyses of MDC-
positive spots showed that ConA-treated cultures presented
a significant increase in both the proportion of cells with
autophagosomes and the number of autophagosomes per
cell (Fig. 9E, F). These results indicated that ConA treatment
led to the blocking of autophagic body degradation in stress-
treated microspores. The accumulation of autophagosomes/
autophagic bodies after E-64 and ConA treatments also sug-
gested the existence of autophagic flux in microspores after
the stress.
The effects of the treatments with 3-MA, E-64, and ConA
Fig. 7.  Detection of autophagosomes and autophagic bodies in stress- on cell death and embryogenesis induction were also evalu-
treated microspores by monodansylcadaverine (MDC) staining and ated. After the inductive stress, cell death levels in microspore
ultrastructural analysis. (A, A') MDC staining of an autophagosome/ cultures were significantly reduced by these three inhibitors
autophagic body (green) under confocal microscopy, (A) merged DIC and (Student’s t-test, P<0.05), in comparison with untreated cul-
fluorescence image. (B) Electron microscopy images of autophagosomes.
The main micrograph shows an advanced/mature autophagosome
tures (Fig. 10A, C, E).
that has engulfed cytoplasmic organelles/material. The inset shows an Regarding the quantification of the proembryos in 4 d
autophagosome at an early stage of its formation. Bars represent in (A, A') cultures, the proportion of proembryos in cultures treated
25 µm, in (B) 0.5 µm, in (inset) 0.2 µm. with 3-MA, E-64, and ConA was significantly higher than in
untreated cultures (Fig. 10B, D, F). These results indicated
has been used to inhibit autophagic body degradation and that inhibition of autophagy by blocking autophagosome
to assess autophagic flux in plant tissues and cell suspensions formation, with 3-MA, or by inhibition of autophagic body
(Matsuoka et  al., 1997; Sláviková et  al., 2005; Shin et  al., degradation, with E-64 or ConA, improved embryogenesis
 Autophagy and cathepsins in stress-induced microspore embryogenesis  |  1395

Downloaded from https://academic.oup.com/jxb/article/69/6/1387/4791010 by guest on 17 June 2021

Fig. 8.  Effects of treatments with 3-MA, E-64, and ConA on autophagosome presence in microspore cultures. (A, C–E) MDC staining and confocal
laser scanning microscopy analysis of stress-treated microspores. (B) Control without MDC in stress-treated microspores, which show unspecific
autofluorescence of the microspore wall, the exine. (A) Untreated microspore culture. (C) Microspore culture treated with 3-MA. (D) Microspore culture
treated with E-64. (E) Microspore culture treated with ConA. Scale bars in micrographs represent 20 µm.

initiation yield while reducing cell death levels caused by the To gain more insight into the activation of cathepsins
inductive stress. during cell death in stress-induced microspore embryogen-
esis cultures, the presence and subcellular localization of
the proteins HvPap-1, HvPap-6, and HvPap-19 (a cathepsin
Cathepsin-like activity, gene expression, and
B-like protein in barley) were analysed using specific anti-
subcellular localization in stress-induced microspore
bodies (previously produced in rabbits by Pineda Antibody
embryogenesis
Services; Díaz-Mendoza et  al., 2016). HvPap-12 protein
We analysed the role of cathepsins in the stress response of could not be localized since no antibodies were available. The
microspores because of their relationship with autophagy in specificity of the antibodies in microspore-derived embryos
animals, and their relevant role as plant cell death proteases. was assessed by immunoblot assays. Results revealed that
As a first approach, the enzymatic activity was quantified for each antibody recognized only two bands corresponding
all the cathepsin activities identified in plants. Significant dif- to the inactive (zymogen) and active forms of the corre-
ferences among stages were assessed by ANOVA and Tukey’s sponding C1A protease (Fig.  12). The bands appeared at
tests, with P<0.05 conferring statistical significance. Low lev- the expected molecular weights reported for HvPap-1 (40
els of cathepsin L-/F-, B-, and H-like activities were detected 016  kDa and 26 204  kDa), HvPap-6 (50 226  kDa and 35
in isolated microspores (Fig. 11A). After the stress, especially 158  kDa), and HvPap-19 (37 222  kDa and 29 234  kDa)
in 4 d cultures, cathepsin L-/F-, B-, and H-like proteolytic (Cambra et al., 2012).
activities significantly increased, reaching >2-fold the proteo- Immunofluorescence assays followed by confocal micros-
lytic values detected before the stress (Fig. 11A). copy analyses provided evidence of the induction of cath-
The expression of several cathepsin genes previously charac- epsins and their subcellular localization in microspores after
terized in barley and related to the protease activities detected, the inductive stress to trigger embryogenesis. The results
HvPap-1, HvPap-6, and HvPap-12, which encode cathepsins of showed very low or no detectable signal with the three cath-
type F-, L-,and H-like, respectively, was also analysed in micro- epsin antibodies on isolated microspores before the stress
spore cultures by RT-qPCR. The three cathepsin genes were (Fig.  13A, A', A'', A'''), whereas stress-treated microspores
expressed at low levels in isolated microspores, while after the (Fig.  13B, B', B'', B''') and cells of 4 d cultures (Fig.  13C,
stress treatment (in stress-treated microspores and 4 d cultures), C', C'', C''') exhibited intense and specific labelling in small
HvPap-1 and HvPap-6 were induced (Fig.  11B). The cathep- cytoplasmic spots of different sizes, probably correspond-
sin H-like HvPap-12 gene did not show significant changes, ing to small vacuoles, a pattern that resembled autophagy
suggesting that other genes are likely to contribute to the in- structures. Patterns of labelling were similar for the three
crease of cathepsin H activity detected in microspores after cathepsins, except for HvPap-19, which showed much less
stress. Among the cathepsin genes studied, the cathepsin L-like labelling in stress-treated microspores (Fig.  13B''') than the
HvPap-6 showed the greatest increase in expression after stress, others. Controls without the primary antibody did not show
in both stress-treated microspores and 4 d cultures (Fig. 11B). any labelling.
1396 | Bárány et al.

Downloaded from https://academic.oup.com/jxb/article/69/6/1387/4791010 by guest on 17 June 2021

Fig. 9.  Quantitative analyses of autophagy in microspore cultures after treatments with autophagy inhibitors 3-MA, E64, and ConA. (A, B) 3-MA
treatment. (C, D) E-64 treatment. (E, F) ConA treatment. (A, C, E) Cells with autophagosomes (MDC-positive cells) in untreated and treated microspore
cultures. (B, D, F) Autophagosomes per cell in untreated and treated microspore cultures. In all histograms, results are expressed as percentages
(percent change) and referred to the mean percentage in untreated cultures which has been normalized to 100%. Bars in histograms indicate the SE.
Asterisks indicate significant differences between treated and untreated cultures, within each treatment, assessed by Student’s t-test, at P<0.05.

Discussion die, strongly limiting the efficiency of the process (Rodríguez-

Serrano et  al., 2012).The involvement of plant autophagy in
Autophagy is activated and has a role in the cell PCD processes during development and pathogen infection
death promoted by the inductive stress of microspore is well known (Yang and Bassham, 2015); however, no infor-
embryogenesis mation had been available until now regarding the role of
The results of the present work provide evidence of the activa- autophagy in cell death during stress-induced embryogenesis.
tion and involvement of autophagy in cell death in the response Autophagy has been shown to be a rather general response
of microspores to the inductive stress triggering embryogen- to a variety of abiotic stresses, playing a role in removing dam-
esis in barley. In microspore embryogenesis systems, after the aged proteins and organelles that can be generated as a result
application of the stress treatment, a proportion of the cells of ROS accumulation during oxidative burst. Increasing evi-
present in the in vitro culture are reprogrammed, initiating dence has connected ROS and autophagy in plants and algae
the embryogenesis pathway; these cells are known as respon- (Pérez-Pérez et  al., 2012). In Arabidopsis, it has been dem-
sive cells. Together with the responsive cells, many other cells onstrated that oxidative damage caused by ROS generators
 Autophagy and cathepsins in stress-induced microspore embryogenesis  |  1397

Downloaded from https://academic.oup.com/jxb/article/69/6/1387/4791010 by guest on 17 June 2021

Fig. 10.  Effects of treatments with 3-MA, E-64, and ConA on cell death and embryogenesis induction in microspore cultures. Quantification of the
percentage of dead cells (A, C, E) and proembryos (B, D, F) on microspore cultures 4 d after stress in untreated cultures and cultures treated with 3-MA
(A, B), E-64 (C, D), and ConA (E, F). In all histograms, results are expressed as percentages (percent change) and referred to the mean percentage of
dead cells or proembryos in untreated cultures which has been normalized to 100%. Bars indicate the SE. Asterisks indicate significant differences
between treated and untreated cultures, within each treatment, assessed by Student’s t-test, at P<0.05.

led to a rapid and strong induction of autophagy (Bassham, Furthermore, our results demonstrate that the induct-
2007). Furthermore, when plants are exposed to abiotic ive stress to trigger microspore embryogenesis in barley also
stress conditions, ROS production acts as a common signal induced the activation of autophagy, which is supported by
to activate stress responses, including autophagy (Bassham, the up-regulation of autophagy HvATG5 and HvATG6 genes,
2009). In a previous study, we reported ROS production in the increase of autophagosome-like structures containing
microspores after the inductive stress of embryogenesis, in ATG5 and ATG8 proteins, and the ultrastructural observa-
barley (Rodríguez-Serrano et al., 2012). In the present study, tion of autophagic structures in microspores after the stress.
endogenous ROS production has also been detected in bar- Moreover, when the inhibitors E-64 or ConA were added to
ley microspores after stress, while treatments with ROS scav- the culture medium, complete autophagosomal degradation is
engers lead to a reduction in cell death levels. These results impaired and the proportion of microspores with autophagic
indicate the involvement of these reactive molecules in micro- structures and their number per cell increased, indicating the
spore death in this system. active formation of autophagosomes after the stress.
1398 | Bárány et al.

inductive stress to trigger embryogenesis (Rodríguez-Serrano

et  al., 2012). A  recent report has demonstrated that the
Arabidopsis cathepsin B protease has caspase-3-like activity
and is inhibited by caspase-3-specific inhibitors; AtCathepsin
B triple mutants showed a strong reduction in PCD induced
by several abiotic stresses, including oxidative stress, indicat-
ing a central role for this protease in stress-induced PCD in
Arabidopsis (Ge et al., 2016). The results of the present study
show the effects of a specific caspase-3 inhibitor, namely the
reduction of cell death levels in microspore cultures, provid-
ing additional evidence for the involvement of caspase-like
activities in the stress-induced cell death in microspores. In
barley caryopsis, a VEIDase protease was found to have a
caspase-like activity; it was localized to autophagosomes,
linking the caspase activity to autophagic PCD (Borén et al.,

Downloaded from https://academic.oup.com/jxb/article/69/6/1387/4791010 by guest on 17 June 2021

2006). The results presented here show that autophagy and
cell death are also connected to caspase-3-like proteolytic
activity in microspores treated with the inductive stress of
embryogenesis.
The role of autophagy in degradation of cellular compo-
nents during PCD execution has been reported in various
plant PCD processes during development, such as suspensor
degradation in spruce somatic embryogenesis (Minina et al.,
Fig. 11.  Patterns of cathepsin proteolytic activities and gene expression
2013, 2014a, b), ovary degradation in wheat, petal senescence,
during stress-induced microspore embryogenesis. (A) Proteolytic pattern and xylogenesis (reviewed in Bassham, 2009). Autophagy has
of cathepsin L-/F-like, cathepsin B-like, and cathepsin H-like cysteine also been implicated in PCD induced by pathogens and other
proteases. Specific activity, in nmol mg–1 min–1. (B) Transcript levels of the injuries (Hofius et al., 2011, 2017). New results are reported
HvPap-1 gene (cathepsin F-like protease), HvPap-6 gene (cathepsin L-like here on the activation of autophagy associated with cell
protease), and HvPap-12 gene (cathepsin H-like protease) normalized to
the isolated microspore within each gene. Bars indicate the SE. Different
death occurrence, in response to the inductive stress trigger-
letters indicate significant differences among stages within each activity/ ing microspore embryogenesis. Nevertheless, the exact role
gene studied, according to ANOVA and Tukey’s tests at P<0.05. of autophagy in cell death is still not completely understood.
The functional analyses performed in the present study with
several autophagy inhibitors have revealed the implication
of autophagy in cell death occurrence. The results presented
here reveal that 3-MA inhibited autophagy in stress-treated
microspores of barley, as in other plant systems, most prob-
ably impairing autophagosome formation (Takatsuka et al.,
2004). Moreover, 3-MA treatment resulted in the reduction
of cell death levels of microspores after stress. Secondary
effects of 3-MA have been reported in some systems, such as
in Arabidopsis root hairs, where 3-MA could inhibit mito-
chondrial-activated PCD rather than autophagy (Kacprzyk
et  al., 2014). Although possible secondary effects of 3-MA
in microspores cannot be completely ruled out, there is no
evidence in microspores of mitochondrial activation of PCD,
and our results indicate that in microspores 3-MA inhibits
autophagy and leads to a reduction of cell death. Treatments
with other autophagy inhibitors, such as E-64 and ConA,
which block autophagosome degradation, have been used
Fig. 12.  Protein patterns of cathepsins HvPap-1, HvPap-6, and HvPap-19
in microspore-derived embryos detected by immunoblot. Arrows indicate
in plant tissues and suspension cells (Matsuoka et al., 1997;
bands corresponding to the inactive (upper) and active (lower) forms of Sláviková et al., 2005; Moriyasu and Inoue, 2008; Shin et al.,
each protease. 2014; Bassham, 2015; Yano et  al., 2015). Their application
in microspore cultures also leads to impaired autophagy ac-
On the other hand, plant proteases with caspase-3-like tivity in microspores after the stress, and to a decrease in cell
activity are well known to participate in many PCD processes death levels. Therefore, these results indicate the involvement
(Bonneau et al., 2008). In a previous report, we have shown of autophagy, at least in part, in the death of the microspores
this proteolytic activity to be induced in microspores after the in response to the stress.
 Autophagy and cathepsins in stress-induced microspore embryogenesis  |  1399

Downloaded from https://academic.oup.com/jxb/article/69/6/1387/4791010 by guest on 17 June 2021

Fig. 13.  Immunolocalization of barley cysteine proteases (HvPap-1, -6, and -19) during stress-induced microspore embryogenesis. Immunofluorescence
and confocal laser scanning microscopy analysis of (A–A''') isolated microspore, (B–B''') stress-treated microspore, and (C–C''') 4 d culture proembryo
confined by the exine. (A-C) Normarsky’s differential interference contrast (DIC). (A'–C''') Merged images of cysteine protease immunofluorescence
(green) and DAPI staining of nuclei (blue). (A'–C') HvPap-1 cysteine protease. (A''-C'') HvPap-6 cysteine protease. (A'''-C''') HvPap-19 cysteine protease.
Scale bars represent in (A–B''') 10 µm, in (C–C''') 20 µm.

Autophagy is activated in response to many physiological proportion of the heterogeneous cell population of the micro-
cues and stress conditions, and has been associated with both spore cultures tolerates the stress and is responsive to embryo-
cell survival and cell death. Depending on the context and genesis induction, whereas many other cells cannot tolerate
intensity, autophagy can protect cells or mediate cell death the stress and die. Our results show that the application of
(Kroemer and Jäättelä, 2005). In the case of microspore the inductive stress leads to the activation of autophagy that
embryogenesis, the response of microspores to the stress plays a role in the death of cells, since autophagy inhibition
treatment depends on many factors, such as the physiological reduces cell death levels. As a consequence of this reduction
state and the developmental stage of the cell. Only a certain in cell death, embryogenesis induction was enhanced. On the
1400 | Bárány et al.

other hand, the possibility that autophagy activity could also as cathepsins B- and L-like proteases, impairs autophagy,
have a prosurvival function in some other stress-treated cells and mutants with reduced cathepsins B and D show impaired
cannot be completely ruled out. autophagic degradation (Tatti et al., 2012). A previous study
connected autophagy and cathepsins to the promotion of
Together with autophagy, cathepsins are induced and cell death associated with the hypersensitive response to
participate in the cell death of microspores after the pathogens in Arabidopsis (Hofius et al., 2009), and it was
inductive stress of embryogenesis suggested that they could contribute to different cell death
pathways operating in plant immunity responses. The pat-
Because cathepsins are well known lysosomal proteases with tern of localization of HvPap cathepsins in microspores as
a role in autophagy and cell death, in animals (Turk and cytoplasmic spots of different sizes is consistent with them
Stoka, 2007), and as they are major proteases with reported being located in vesicles, lysosomal-like structures and
functions in cell death also in plants, we have analysed the small vacuoles, some of which could represent autolyso-
participation of cathepsins in the microspore response to the some-like structures and small autophagic vacuoles. This
inductive stress. In animals, numerous reports have docu- fact, together with the activation of both autophagy and
mented the critical role of cathepsins in the degradation of cathepsins after the stress, the similar induction of cath-

Downloaded from https://academic.oup.com/jxb/article/69/6/1387/4791010 by guest on 17 June 2021

cytoplasmic organelles and components through autophagy, epsin and ATG gene expression after the stress, and the
being responsible for the terminal degradation of proteins observation of the same effects in reducing cell death by the
within autolysosomes (Kroemer and Jäättelä, 2005; Jung inhibition of both actions, suggests a connection between
et al., 2015; Man and Kanneganti, 2016). Nevertheless, much C1A proteases (cathepsins) and autophagy in stress-treated
less information is available on plant cathepsins. microspores, as has been widely demonstrated in mamma-
The results of our study revealed the participation of C1A lian cells (Man and Kanneganti, 2016), although further
proteases (cathepsins) in stress-induced microspore embryo- work will be required to prove this connection. The induc-
genesis, with the up-regulation of cathepsin genes HvPap-1 tion of autophagy after the stress, together with the acti-
and HvPap-6, which encode cathepsins F- and L-like, respect- vation of cathepsins, may be crucial in the orchestration
ively, after the stress. Concomitantly, the cathepsin L-/F-, B-, of cell death among other cell responses to the inductive
and H-like proteolytic activities increase in stress-treated stress, therefore participating in the control of success of
microspores, as does the presence in their cytoplasm of pro- embryogenesis initiation.
teins HvPap-1, HvPap-6, and HvPap-19 (a cathepsin B-like
protein in barley; Díaz-Mendoza et al. 2016). These proteases
Conclusions
localized in small cytoplasmic spots of various sizes, prob-
ably corresponding to vesicles, lysosomal-like organelles, and In summary, the results reported here reveal that autophagy
small vacuoles of stress-treated microspores, a localization is activated after the inductive stress used to trigger micro-
pattern that resembles that of autophagy structures. These spore embryogenesis in barley, and its pharmacological
results indicate the role of cysteine C1A proteases in the inhibition reduces cell death levels, indicating a role for
microspore response to stress. autophagy in the stress-induced cell death of microspores.
In barley, C1A proteases, specifically HvPap-1, have been Cathepsin protease activities are concomitantly induced,
reported to participate in the proteolysis induced in leaves and their inhibition also impaired cell death. The simi-
by abiotic stresses such as darkness and nitrogen starvation lar patterns of activation, expression, and localization of
(Velasco-Arroyo et al., 2016), as well as in the development autophagy and cathepsins suggest a connection between
and germination of barley grains (Díaz-Mendoza et al., 2016). both activities in stress-induced cell death during micro-
Nevertheless, the role of HvPap-1 in PCD has not been previ- spore embryogenesis induction, a hypothesis that needs
ously described. Likewise, little is known about the function further analyses. The findings provide new insights into
of HvPap-6 and HvPap-19 in this process. In Arabidopsis, the mechanisms underlying the microspore response to the
Zhang et al. (2014) reported that CEP1, a C1A cysteine pro- inductive stress, opening up new possibilities to enhance
tease, plays a key role in tapetal PCD, a process that critically microspore embryogenesis efficiency in recalcitrant spe-
regulates pollen development. Our results demonstrate that cies while reducing cell death levels with modulators of
these proteases contribute to the response to stress of micro- autophagy and cysteine proteases.
spores. Moreover, when we treated microspores with E-64,
which inhibits intracellular cysteine proteases, the levels of
cell death decreased, suggesting the involvement of these pro-
Supplementary data
teases in cell death in stress-treated microspore cultures. As
a consequence of this reduction in cell death, embryogenesis Supplementary data are available at JXB online.
induction was enhanced, which opens up new possibilities for Table S1. Cathepsin-like protease amino acid sequences
biotechnological manipulation of the process with cysteine (peptides) used for specific antibody production.
protease modulators to improve the yield of in vitro embryo- Table S2. Primer sequences used for the amplification of
genesis systems. genes by RT-qPCR assays.
Several reports in animals have demonstrated that treat- Method S1. Methods for the production of ATG5 antibod-
ment with inhibitors of the lysosomal cysteine proteases, such ies with the recombinant ATG5 protein of Picea abies.
 Autophagy and cathepsins in stress-induced microspore embryogenesis  |  1401

Acknowledgements Hofius D, Munch D, Bressendorff S, Mundy J, Petersen M. 2011.

Role of autophagy in disease resistance and hypersensitive response-
The authors thank Dr M.F. Suárez (University of Málaga, Málaga, Spain) associated cell death. Cell Death and Differentiation 18, 1257–1262.
for the generous gift of the anti-ATG5 antibody. This work is supported by Hofius D, Schultz-Larsen T, Joensen J, Tsitsigiannis DI, Petersen
projects AGL2014-52028-R and AGL2017-82447-R funded by the Spanish NH, Mattsson O, Jørgensen LB, Jones JD, Mundy J, Petersen M.
Ministry of Economy and Competitiveness (MINECO) and the European 2009. Autophagic components contribute to hypersensitive cell death in
Regional Development Fund (ERDF/FEDER). YPP is the recipient of a Arabidopsis. Cell 137, 773–783.
grant (PEJ15/BIO/AI-01S8) funded by Comunidad de Madrid and European
Commission through ERDF/FEDER. Jung M, Lee J, Seo HY, Lim JS, Kim EK. 2015. Cathepsin inhibition-
induced lysosomal dysfunction enhances pancreatic beta-cell apoptosis in
high glucose. PLoS One 10, e0116972.
Kacprzyk J, Devine A, McCabe PF. 2014. The root hair assay facilitates
References the use of genetic and pharmacological tools in order to dissect multiple
signalling pathways that lead to programmed cell death. PLoS One 9,
Álvarez C, García I, Moreno I, Pérez-Pérez ME, Crespo JL, e94898.
Romero LC, Gotor C. 2012. Cysteine-generated sulfide in the cytosol
negatively regulates autophagy and modulates the transcriptional profile in Kroemer G, Jäättelä M. 2005. Lysosomes and autophagy in cell death
Arabidopsis. The Plant Cell 24, 4621–4634. control. Nature Reviews Cancer 5, 886–897.
Apel K, Hirt H. 2004. Reactive oxygen species: metabolism, oxidative Li F, Vierstra RD. 2012. Autophagy: a multifaceted intracellular system for
stress, and signal transduction. Annual Review of Plant Biology 55, 373–399. bulk and selective recycling. Trends in Plant Science 17, 526–537.

Downloaded from https://academic.oup.com/jxb/article/69/6/1387/4791010 by guest on 17 June 2021

Avila-Ospina L, Marmagne A, Soulay F, Masclaux-Daubresse C. Liu Y, Bassham DC. 2012. Autophagy: pathways for self-eating in plant
2016. Identification of barley (Hordeum vulgare L.) autophagy genes and cells. Annual Review of Plant Biology 63, 215–237.
their expression levels during leaf senescence, chronic nitrogen limitation Livak KJ, Schmittgen TD. 2001. Analysis of relative gene expression
and in response to dark exposure. Agronomy 6, 15. data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method.
Bárány I, González-Melendi P, Fadón B, Mitykó J, Risueño MC, Methods 25, 402–408.
Testillano PS. 2005. Microspore-derived embryogenesis in pepper Maluszynski M, Kasha K, Forster B, Szarejko I, eds. 2003. Doubled
(Capsicum annuum L.): subcellular rearrangements through development. haploid production in crop plants: a manual. Dordrecht: Kluwer.
Biology of the Cell 97, 709–722. Man SM, Kanneganti TD. 2016. Regulation of lysosomal dynamics and
Bassham DC. 2007. Plant autophagy—more than a starvation response. autophagy by CTSB/cathepsin B. Autophagy 12, 2504–2505.
Current Opinion in Plant Biology 10, 587–593. Martinez M, Cambra I, Carrillo L, Diaz-Mendoza M, Diaz I. 2009.
Bassham DC. 2009. Function and regulation of macroautophagy in Characterization of the entire cystatin gene family in barley and their target
plants. Biochimica et Biophysica Acta 1793, 1397–1403. cathepsin L-like cysteine-proteases, partners in the hordein mobilization
Bassham DC. 2015. Methods for analysis of autophagy in plants. during seed germination. Plant Physiology 151, 1531–1545.
Methods 75, 181–188. Martínez M, Cambra I, González-Melendi P, Santamaría ME, Díaz
Berenguer E, Bárány I, Solís MT, Pérez-Pérez Y, Risueño MC, I. 2012. C1A cysteine-proteases and their inhibitors in plants. Physiologia
Testillano PS. 2017. Inhibition of histone H3K9 methylation by BIX- Plantarum 145, 85–94.
01294 promotes stress-induced microspore totipotency and enhances Masclaux-Daubresse C, Chen Q, Havé M. 2017. Regulation of nutrient
embryogenesis initiation. Frontiers in Plant Science 8, 1161. recycling via autophagy. Current Opinion in Plant Biology 39, 8–17.
Bonneau L, Ge Y, Drury GE, Gallois P. 2008. What happened to plant Matsuoka K, Higuchi T, Maeshima M, Nakamura K. 1997. A vacuolar-
caspases? Journal of Experimental Botany 59, 491–499. type H+-ATPase in a nonvacuolar organelle is required for the sorting of soluble
Borén M, Höglund AS, Bozhkov P, Jansson C. 2006. Developmental vacuolar protein precursors in tobacco cells. The Plant Cell 9, 533–546.
regulation of a VEIDase caspase-like proteolytic activity in barley caryopsis. Merkulova EA, Guiboileau A, Naya L, Masclaux-Daubresse C,
Journal of Experimental Botany 57, 3747–3753. Yoshimoto K. 2014. Assessment and optimization of autophagy
Bradford MM. 1976. A rapid and sensitive method for the quantitation monitoring methods in Arabidopsis roots indicate direct fusion of
of microgram quantities of protein utilizing the principle of protein–dye autophagosomes with vacuoles. Plant and Cell Physiology 55, 715–726.
binding. Analytical Biochemistry 72, 248–254. Michaeli S, Galili G, Genschik P, Fernie AR, Avin-Wittenberg T. 2016.
Cambra I, Hernández D, Diaz I, Martinez M. 2012. Structural basis Autophagy in plants—what’s new on the menu? Trends in Plant Science
for specificity of propeptide–enzyme interaction in barley C1A cysteine 21, 134–144.
peptidases. PLoS One 7, e37234. Minina EA, Bozhkov PV, Hofius D. 2014a. Autophagy as initiator or
Contento AL, Xiong Y, Bassham DC. 2005. Visualization of autophagy executioner of cell death. Trends in Plant Science 19, 692–697.
in Arabidopsis using the fluorescent dye monodansylcadaverine and a Minina EA, Filonova LH, Fukada K, et al. 2013. Autophagy and
GFP–AtATG8e fusion protein. The Plant Journal 42, 598–608. metacaspase determine the mode of cell death in plants. Journal of Cell
Díaz-Mendoza M, Dominguez-Figueroa JD, Velasco-Arroyo B, et al. Biology 203, 917–927.
2016. HvPap-1 C1A protease and HvCPI-2 cystatin contribute to barley Minina EA, Smertenko AP, Bozhkov PV. 2014b. Vacuolar cell death in
grain filling and germination. Plant Physiology 170, 2511–2524. plants: metacaspase releases the brakes on autophagy. Autophagy 10,
Díaz-Mendoza M, Velasco-Arroyo B, González-Melendi P, Martínez 928–929.
M, Díaz I. 2014. C1A cysteine protease–cystatin interactions in leaf Moriyasu Y, Inoue Y. 2008. Use of protease inhibitors for detecting
senescence. Journal of Experimental Botany 65, 3825–3833. autophagy in plants. Methods in Enzymology 451, 557–580.
El-Tantawy AA, Solís MT, Da Costa ML, Coimbra S, Risueño MC, Niemann A, Takatsuki A, Elsässer HP. 2000. The lysosomotropic agent
Testillano PS. 2013. Arabinogalactan protein profiles and distribution monodansylcadaverine also acts as a solvent polarity probe. Journal of
patterns during microspore embryogenesis and pollen development in Histochemistry and Cytochemistry 48, 251–258.
Brassica napus. Plant Reproduction 26, 231–243. Pérez-Pérez ME, Florencio FJ, Crespo JL. 2010. Inhibition of target
Ge Y, Cai YM, Bonneau L, Rotari V, Danon A, McKenzie EA, of rapamycin signaling and stress activate autophagy in Chlamydomonas
McLellan H, Mach L, Gallois P. 2016. Inhibition of cathepsin B by reinhardtii. Plant Physiology 152, 1874–1888.
caspase-3 inhibitors blocks programmed cell death in Arabidopsis. Cell Pérez-Pérez ME, Lemaire SD, Crespo JL. 2012. Reactive oxygen
Death and Differentiation 23, 1493–1501. species and autophagy in plants and algae. Plant Physiology 160, 156–164.
Germaná MA, Lambardi M. 2016. In vitro embryogenesis in higher Prem D, Solís MT, Bárány I, Rodríguez-Sanz H, Risueño MC,
plants. New York, Heidelberg: Humana Press, Springer. Testillano PS. 2012. A new microspore embryogenesis system under low
Hofius D, Li L, Hafrén A, Coll NS. 2017. Autophagy as an emerging temperature which mimics zygotic embryogenesis initials, expresses auxin
arena for plant–pathogen interactions. Current Opinion in Plant Biology 38, and efficiently regenerates doubled-haploid plants in Brassica napus. BMC
117–123. Plant Biology 12, 127.
1402 | Bárány et al.
Rawlings ND, Barrett AJ, Finn R. 2016. Twenty years of the MEROPS and MET1a-like gene expression changes during stress-induced pollen
database of proteolytic enzymes, their substrates and inhibitors. Nucleic reprogramming to embryogenesis. Journal of Experimental Botany 63,
Acids Research 44, D343–D350. 6431–6444.
Rodríguez-Sanz H, Moreno-Romero J, Solís MT, Köhler C, Risueño Takatsuka C, Inoue Y, Matsuoka K, Moriyasu Y. 2004.
MC, Testillano PS. 2014. Changes in histone methylation and acetylation 3-Methyladenine inhibits autophagy in tobacco culture cells under sucrose
during microspore reprogramming to embryogenesis occur concomitantly starvation conditions. Plant and Cell Physiology 45, 265–274.
with Bn HKMT and Bn HAT expression and are associated with cell Tatti M, Motta M, Di Bartolomeo S, Scarpa S, Cianfanelli V, Cecconi
totipotency, proliferation, and differentiation in Brassica napus. Cytogenetic F, Salvioli R. 2012. Reduced cathepsins B and D cause impaired
and Genome Research 143, 209–218. autophagic degradation that can be almost completely restored by
Rodríguez-Sanz H, Solís MT, López MF, Gómez-Cadenas A, overexpression of these two proteases in Sap C-deficient fibroblasts.
Risueño MC, Testillano PS. 2015. Auxin biosynthesis, accumulation, Human Molecular Genetics 21, 5159–5173.
action and transport are involved in stress-induced microspore Testillano PS, Coronado MJ, Thierry AM, Matthys-Rochon E,
embryogenesis initiation and progression in Brassica napus. Plant and Cell Risueño MC. 2010. In situ detection of Esr proteins secretion during
Physiology 56, 1401–1417. maize microspore embryogenesis and their secretion blockage show
Rodríguez-Serrano M, Bárány I, Prem D, Coronado MJ, Risueño effects on the culture progression. Functional Plant Biology 37, 985–994.
MC, Testillano PS. 2012. NO, ROS, and cell death associated with Turk B, Stoka V. 2007. Protease signalling in cell death: caspases versus
caspase-like activity increase in stress-induced microspore embryogenesis cysteine cathepsins. FEBS Letters 581, 2761–2767.
of barley. Journal of Experimental Botany 63, 2007–2024.
Velasco-Arroyo B, Diaz-Mendoza M, Gandullo J, Gonzalez-Melendi
Shin KD, Lee HN, Chung T. 2014. A revised assay for monitoring

Downloaded from https://academic.oup.com/jxb/article/69/6/1387/4791010 by guest on 17 June 2021

P, Santamaria ME, Dominguez-Figueroa JD, Hensel G, Martinez M,
autophagic flux in Arabidopsis thaliana reveals involvement of Kumlehn J, Diaz I. 2016. HvPap-1 C1A protease actively participates in
AUTOPHAGY-RELATED9 in autophagy. Molecules and Cells 37, 399–405. barley proteolysis mediated by abiotic stresses. Journal of Experimental
Sláviková S, Shy G, Yao Y, Glozman R, Levanony H, Pietrokovski Botany 67, 4297–4310.
S, Elazar Z, Galili G. 2005. The autophagy-associated Atg8 gene family Yang X, Bassham DC. 2015. New insight into the mechanism and
operates both under favourable growth conditions and under starvation function of autophagy in plant cells. International Review of Cell and
stresses in Arabidopsis plants. Journal of Experimental Botany 56, 2839–2849. Molecular Biology 320, 1–40.
Solís MT, Berenguer E, Risueño MC, Testillano PS. 2016. BnPME is Yano K, Yanagisawa T, Mukae K, Niwa Y, Inoue Y, Moriyasu Y. 2015.
progressively induced after microspore reprogramming to embryogenesis, Dissection of autophagy in tobacco BY-2 cells under sucrose starvation
correlating with pectin de-esterification and cell differentiation in Brassica conditions using the vacuolar H(+)-ATPase inhibitor concanamycin A and
napus. BMC Plant Biology 16, 176. the autophagy-related protein Atg8. Plant Signaling and Behavior 10,
Solís MT, Chakrabarti N, Corredor E, Cortés-Eslava J, Rodríguez- e1082699.
Serrano M, Biggiogera M, Risueño MC, Testillano PS. 2014. Zhang D, Liu D, Lv X, Wang Y, Xun Z, Liu Z, Li F, Lu H. 2014. The
Epigenetic changes accompany developmental programmed cell death in cysteine protease CEP1, a key executor involved in tapetal programmed
tapetum cells. Plant and Cell Physiology 55, 16–29. cell death, regulates pollen development in Arabidopsis. The Plant Cell 26,
Solís MT, El-Tantawy AA, Cano V, Risueño MC, Testillano PS. 2939–2961.
2015. 5-Azacytidine promotes microspore embryogenesis initiation by Zhao H, Kalivendi S, Zhang H, Joseph J, Nithipatikom K, Vásquez-
decreasing global DNA methylation, but prevents subsequent embryo Vivar J, Kalyanaraman B. 2003. Superoxide reacts with hydroethidine
development in rapeseed and barley. Frontiers in Plant Science 6, 472. but forms a fluorescent product that is distinctly different from ethidium:
Solís MT, Rodríguez-Serrano M, Meijón M, Cañal MJ, Cifuentes potential implications in intracellular fluorescence detection of superoxide.
A, Risueño MC, Testillano PS. 2012. DNA methylation dynamics Free Radical Biology and Medicine 34, 1359–1368.