Environmental and Experimental Botany 140 (2017) 96–109

Contents lists available at ScienceDirect

Environmental and Experimental Botany

journal homepage: www.elsevier.com/locate/envexpbot

Impact of seasonal warming on overwintering and spring phenology of MARK

blackcurrant
Uffe Brandt Andersena, Katrine Heinsvig Kjaera, Alexander Erbanb, Jessica Alpersb,
⁎
Dirk K. Hinchab, Joachim Kopkab, Ellen Zutherb, Majken Pagterc,
a
Department of Food Science, Aarhus University, Kirstinebjergvej 10, Aarslev, DK-5792, Denmark
b
Max-Planck-Institut für Molekulare Pflanzenphysiologie, Am Mühlenberg 1, D-14476 Potsdam, Germany
c
Department of Chemistry and Bioscience, Aalborg University, Fredrik Bajers Vej 7H, DK-9220 Aalborg East, Denmark

A R T I C L E I N F O A B S T R A C T

Keywords: The rate of global warming varies on a seasonal timescale, with temperatures increasing faster in winter and
Bud break spring than in summer or autumn. Temperature is a major driver of phenological events, but little information is
Climate change available about the effects of non-uniform seasonal warming on plant development. Here we report phenolo-
Dormancy gical, metabolic and transcriptional responses of two Ribes nigrum cultivars to slightly elevated temperatures in
Fruit yield
winter and spring. Winter warming delayed floral bud break of ‘Narve Viking’, but advanced bud break of
Gene expression
Metabolite profiling
‘Zusha’, implying genotypic differences in the magnitude of responses to warming during dormancy and in the
heat accumulation phase in spring. Accordingly, more genes putatively associated with dormancy release,
growth or development were up-regulated earlier in ‘Zusha’ than in ‘Narve Viking’ or showed significantly
different transcript abundance in buds of plants at elevated or ambient temperatures. Primary metabolism of
flower buds was largely unaffected by warming, but pronounced seasonal accumulation patterns of specific
metabolites were evident. Especially bud break was associated with a dramatic remodeling of the metabolome,
with differential seasonal regulation of metabolism between the two cultivars. Asymmetric warming did not
increase the risk of freeze-induced damage to stems and flower primordia. The expression of several genes
putatively related to freezing tolerance appeared coordinated during seasonal transitions in freezing tolerance.
These data provide new insights into the timing of metabolic and transcriptional regulation during deacclima-
tion, dormancy release and bud burst in a woody perennial, and contribute to our understanding of plant
phenological responses to climate change.

1. Introduction of deacclimation with increasing temperatures and heat sum require-

The Earth́ s mean global surface temperature is increasing and is
projected to increase further in the foreseeable future (IPCC, 2013).
Except for some tropical regions, meteorological records and climate
model projections have shown that the rate of warming varies on a
seasonal timescale, with temperatures increasing faster in winter and in
spring than in summer or autumn (Xia et al., 2014). However, current
research on plant responses to climate change remains seasonally
biased towards the growing season and little effort has been directed
towards understanding the effects of non-uniform seasonal warming on
plant development (Liu et al., 2012; Ladwig et al., 2016). Temperature
is a major driver of phenological events in temperate woody perennials
(Fitter and Fitter, 2002; Parmesan, 2006), including chilling require-
ment for the transition from endodormancy to ecodormancy, induction
ment to initiate bud break, which all takes place in winter and spring.
Seasonal asymmetric warming is therefore likely to induce changes in
most of these events. Proper timing of phenological events is especially
important for perennial fruit crops where it plays a critical role in the
selection of cultivars that are appropriate for a specific region and
where the inability of a cultivar to adequately respond to environ-
mental conditions may have severe consequences in terms of plant
mortality, annual growth and fruit yield (Fu et al., 2012; Chung et al.,
2013).
Woody perennials from temperate climate regions, such as black-
currants (Ribes nigrum), develop endodormancy at the end of the
growing season as an adaptive strategy to avoid growth and flowering
under unfavorable conditions. Endodormancy is defined as the inability
to initiate growth from meristems under growth promoting conditions

⁎
Corresponding author.
E-mail addresses: katrine.kjaer@food.au.dk (K.H. Kjaer), Erban@mpimp-golm.mpg.de (A. Erban), Alpers@mpimp-golm.mpg.de (J. Alpers),
Hincha@mpimp-golm.mpg.de (D.K. Hincha), Kopka@mpimp-golm.mpg.de (J. Kopka), Zuther@mpimp-golm.mpg.de (E. Zuther), mp@bio.aau.dk (M. Pagter).

http://dx.doi.org/10.1016/j.envexpbot.2017.06.005
Received 5 May 2017; Received in revised form 9 June 2017; Accepted 9 June 2017
Available online 12 June 2017
0098-8472/ © 2017 Elsevier B.V. All rights reserved.
U.B. Andersen et al.
and is gradually overcome by chilling (Rohde and Bhalerao, 2007).
Thus, temperature has a dual effect on bud development. On the one
hand, low temperatures are necessary to break bud dormancy and, on
the other hand, higher temperatures are necessary to promote bud
growth afterward. One likely impact of warming during dormancy is a
delay in the fulfilment of chill requirements and consequently a delay in
the time at which perennials become receptive to warm temperatures
for spring growth (Luedeling et al., 2011). Insufficient chilling may
additionally cause uneven bud break and a reduction in flowering and/
or reproductive output of perennial fruit crops (Atkinson et al., 2013).
Inversely, the advances in spring phenology, that have dominated cli-
mate-warming responses thus far, have been explained by increasing
temperatures during ecodormancy leading to a more rapid fulfilment of
the heat requirement (Cleland et al., 2007). Blackcurrant has a rela-
tively high chilling requirement and is one of the fruit crops that is
potentially at risk in parts of Europe due to the lack of winter chilling
forecast in projections of future climatic conditions (Atkinson et al.,
2013; Jones et al., 2013). This hypothesis is supported by a recent
study, showing that seasonal asymmetric warming may specifically
delay dormancy release of high-chilling requiring cultivars (Pagter
et al., 2015).
Simultaneously with dormancy development, overwintering organs
of temperate perennials increase their freezing tolerance through cold
acclimation. Maximum freezing tolerance is reached mid-winter, while
in spring plants lose acclimated freezing tolerance by deacclimation.
Milder weather in winter and spring may decrease mid-winter freezing
tolerance and directly accelerate the deacclimation process (Ogren
et al., 1997; Taulavuori et al., 2004), increasing the risk of frost injury
in buds and stems of deciduous species. Sugars and other compatible
solutes exert cryoprotective properties and accumulate in over-
wintering organs during cold acclimation (Guy et al., 2008; Charrier
et al., 2013). Consequently, one explanation for reduced mid-winter
freezing tolerance and accelerated deacclimation at elevated tempera-
tures is the consumption of soluble carbohydrates due to increasing
respiration rates (Ögren, 1996; Taulavuori et al., 1997). During spring,
these stored carbohydrate reserves support metabolic activity of
flushing buds until new leaves develop (Cheng and Fuchigami, 2002).
Warming-induced changes in carbohydrate metabolism may therefore
also alter the availability of carbohydrates for spring growth. Consistent
with studies of other woody perennials (Ögren, 1996; Taulavuori et al.,
1997; Riikonen et al., 2013) elevated winter temperatures affect car-
bohydrate metabolism of blackcurrant stems, although without causing
a decrease in freezing tolerance (Pagter et al., 2015). It remains largely
unknown, however, whether warming alters biochemical responses
associated with phenological traits other than carbohydrate metabo-
lism.
The molecular events that regulate cold acclimation have been well
studied and numerous reviews on this topic have been published
(Welling and Palva, 2006; Fennell, 2014; Wisniewski et al., 2014; Shi
et al., 2015). In comparison, little is known about the molecular me-
chanisms underlying deacclimation (Zuther et al., 2015). CBF genes, a
sub-family of the APETALA2/ETHYLENE RESPONSE FACTOR (AP2/
ERF) transcription factors, play an integral, regulating role in both cold
acclimation and deacclimation (Wisniewski et al., 2014; Zuther et al.,
2015). Functional studies of CBF genes in woody plants indicate that
their regulation and impact on abiotic stress resistance are more com-
plex than in herbaceous plants and may include a role in the regulation
of dormancy (Wisniewski et al., 2014). Cold acclimation is associated
with induction of cold-regulated (COR) genes, which play a funda-
mental role in stress tolerance and are targets of CBF transcription
factors (Jaglo-Ottosen et al., 1998; Welling and Palva, 2006). In parti-
cular, transcripts encoding LEA proteins or the corresponding proteins
accumulate during endodormancy or in response to chilling (Welling
and Palva, 2008; Ueno et al., 2013; Falavigna et al., 2014; Shin et al.,
2015). Reconfiguration of carbohydrate metabolism in cold acclimating
plants is correlated with the expression of genes encoding enzymes in
97
Environmental and Experimental Botany 140 (2017) 96–109
the respective pathways, such as raffinose biosynthesis, consistent with
regulation by transcriptional activation (Guy et al., 2008; Maruyama
et al., 2009). A variety of genetic components that contribute to the
intricate regulation of dormancy release, bud break and flowering have
been reported (Rohde et al., 2007; Cooke et al., 2012; Falavigna et al.,
2014; Howe et al., 2015). In blackcurrant, Hedley et al. (2010) iden-
tified three genes which co-localise with previously characterized
blackcurrant bud break QTL and have probable roles in the dormancy
release and bud break processes. They encode proteins with homology
to acetyl CoA carboxylase, calmodulin-binding protein and beta-tu-
bulin. The regulation of dormancy and freezing tolerance in woody
perennials has received increasing attention in recent years. Never-
theless, there is still only limited information regarding the role of
various genes associated with overwintering or the regulation of sea-
sonal transitions in freezing tolerance and dormancy status. In addition,
molecular knowledge of the potential effects of winter warming on the
processes of deacclimation, dormancy release and bud break is com-
pletely lacking.
Blackcurrant is grown in a relatively small area, but is an econom-
ically high value crop that is increasingly recognized as a rich source of
vitamin C and anthocyanins (Vagiri et al., 2012). In Europe, where
blackcurrant is mainly grown, production is rising and there is in-
creasing interest in expanding it to countries and regions that currently
do not cultivate R. nigrum (Mitchell et al., 2011). We have previously
shown that mild winter warming may alter phenological traits in
blackcurrant, depending on genotype specific differences in chilling
requirement (Pagter et al., 2015). Here, evaluation of the effects of
asymmetric seasonal warming on phenological traits in blackcurrant
flower buds was supplemented with GC–MS based metabolite profiling
and expression analyses of endogenous genes homologous to genes
associated with overwintering or spring phenology in other plant spe-
cies. It was hypothesized that warming induced phenological changes
are associated with alterations in primary metabolism and transcript
levels of genes associated with overwintering and/or spring phenology
and that the timing of metabolic and transcriptional regulation during
deacclimation, dormancy release and bud burst varies between culti-
vars with different chilling requirements.
2. Materials and methods
2.1. Experimental set-up and plant material
The experimental area was established at the Department of Food
Science, Aarhus University in Denmark (55° 18′ N 10° 26′ E) and con-
sisted of a control plot (ambient temperature) and a warming plot
(elevated temperature), which were adjacent to each other and had
identical sizes (5 × 5.5 m). Warming was conducted from October
22nd 2014 to April 17th 2015 with 240 m of temperature-controlled
heating cable (producing a maximum of 83 W m−2) distributed be-
tween the plants at a distance of ca. 10 cm from the plants. The cables
were resting on metal holders five cm above the soil surface. The air
temperature in the plots was monitored with Pt100 temperature sensors
covered by radiation shields at 20 cm, 50 cm and 80 cm above the soil
surface. Temperature means over five-min intervals were logged
throughout the treatment period. Whenever the temperature at 20 cm
in the warmed plot was more than 2 °C higher than the corresponding
temperature in the control plot the heating cables were switched off
until the temperature difference fell below 2 °C. The soil temperature
was measured in each plot at a depth of ca. 15 cm using Tinytag Talk
temperature loggers (Gemini Data Loggers, Chichester, England). From
October onwards, after leaf abscission, both plots were covered with a
transparent polyethylene net tunnel providing some wind shelter. The
net had a shading effect of 15% and was removed during bud break on
April 17th. The net tunnel allowed the plants to be exposed to the
natural rainfall.
The experiment was carried out using two-year old vegetatively
U.B. Andersen et al.
propagated plants of Ribes nigrum L. ‘Narve Viking’ and ‘Zusha’, which
are of Norwegian (SØnsteby and Heide, 2011) and Russian origin
(Pedersen, 2008), respectively. ‘Narve Viking’ has a high chilling re-
quirement (Pagter et al., 2015) while we have previously observed that
field-grown ‘Zusha’ is released from dormancy in December indicating a
modest chilling requirement (Kjær and Pagter, unpubl. res.). The plants
were approximately 50 cm high and had 3–4 shoots per plants. Before
start of the experiment plants were grown outside. Hence, they initiated
cold acclimation under natural conditions. Before initiation of the
warming treatment 99 plants of each cultivar, grown in 3.5 L pots
containing a sandy loam, were placed in each of the experimental plots
with pots buried in the soil to avoid root frost injuries. The cultivars
were placed in alternating west-east running rows containing 18 plants
of either cultivar. Within the two cultivars, individual plants were
randomized. During the treatment period plants were exposed to nat-
ural precipitation, whereas after the removal of the polyethylene tunnel
and bud break, natural precipitation was supported by automatic irri-
gation when needed.
2.2. Freezing tolerance of buds and stems
Monthly sampling of current-year shoots for estimation of freezing
tolerance was done twice before initiation of the warming treatment in
September and October, six times during the warming treatment and
once in May after termination of the treatment. At each time, samples
were randomly collected from six plants per cultivar and treatment.
Freezing tolerance of stems was determined using the electrolyte
leakage method as described in detail by Pagter et al. (2011). For es-
timation of freezing tolerance of flower primordia two stem sections
with attached lateral buds were pruned to approximately two cm in
length, wrapped in moist paper towels to ensure ice nucleation and put
into small zipper bags. The bud samples were incubated together with
the stem samples in a pre-cooled temperature-controlled freezer and
were subjected to the same freezing profile as stem samples. After
freezing and thawing overnight on ice at 4 °C, bud samples were in-
cubated at 20 °C in darkness for seven days. Individual buds were then
excised, cut longitudinally and incubated for 2–24 h in darkness at
20 °C in a 0.5% triphenyl tetrazolium chloride (TTC) solution in 0.05 M
phosphate buffer. Following incubation in TTC, the coloration of floral
primordia was assessed under a stereo microscope. Bright and dark red
primordia were evaluated as vital, while weakly coloured, brownish or
colourless primordia were regarded as dead. Since lateral buds of
‘Narve Viking’ started to burst in March, the last evaluation of freezing
tolerance of flower primordia was performed in March.
2.3. Dormancy status
To determine dormancy status, six plants per cultivar and treatment
were moved to forcing conditions in a greenhouse (20 °C, 14 h photo-
period) once a month from October to February. Bud burst of the
terminal bud and the four uppermost lateral buds was separately ob-
served two-three times per week on one shoot per plant as described by
Pagter et al. (2015). Bud burst was recorded using a scale of 0 to 3
where 0 = no bud development, 1 = green tip visible, 2 = one or more
leaves unfolded and 3 = one or more open flowers.
2.4. Spring phenology and cropping performance
Timing of bud break in the field was recorded on one shoot of 21
plants per cultivar and treatment by recording bud development ap-
proximately once a week from March 12th to April 20th. Bud devel-
opment of the terminal bud and the four uppermost lateral buds was
recorded using the same scale as for the evaluation of dormancy status.
The total number of flowers per plant was recorded for ‘Zusha’ on April
20th and for ‘Narve Viking’ on May 7th. The total berry yield per plant
and the weight of 200 berries per plant were recorded on the same
98
Environmental and Experimental Botany 140 (2017) 96–109
plants during fruit harvest in early August. The latter was done in order
to calculate the weight of individual berries and to estimate the total
number of berries per plant.
2.5. GC–MS metabolite profiling
Polar metabolites from flower buds were analysed monthly from
September until May. At each time point, at least five of the uppermost
axillary buds from six plants per cultivar and treatment were sampled in
the morning between 8 and 9 am. Buds were collected in 2.0 mL mi-
crocentrifuge tubes and immediately flash frozen in liquid nitrogen. The
frozen samples were homogenized for five to ten min at 23 Hz in a MM
200 ball mill (Retsch, Haan, Germany) using three 3-mm tungsten
carbide balls.
Polar metabolites were extracted and processed from 80 mg ground
bud material as described previously (Dethloff et al., 2014). Gas chro-
matography coupled to electron impact ionization-time of flight-mass
spectrometry (GC/EI-TOF-MS) was performed on an Agilent 6890N24
gas chromatograph and a Pegasus III mass spectrometer, LECO, St.
Joseph, USA (Erban et al., 2007). Chromatograms were pre-processed
with ChromaTOF software (Leco, http://leco.de). Metabolites were
identified under manual supervision using the TagFinder (Luedemann
et al., 2008), the NIST08 softwares (http://chemdata.nist.gov/
dokuwiki/doku.php?id=start) and the mass spectral and retention
time index (RI) reference collection of the Golm Metabolome Database
(GMD; Kopka et al., 2005; Hummel et al., 2007). Metabolite intensities
were normalized to sample fresh weight and the internal standard 13C6-
sorbitol.
The total number of identified polar metabolites was 133, including
known and yet unknown compounds that are archived by GMD.
Metabolites that were present only in less than half of the samples were
excluded from the dataset, if their absence was not related to cultivar,
treatment or seasonal sampling time. In total, 97 metabolites were se-
lected for further analysis. A complete list of the normalized metabolite
levels can be found in Supplementary Table S1.
2.6. Gene expression analysis
Transcript levels of 32 selected genes with homology to genes en-
coding proteins previously associated with overwintering or spring
phenology were analysed using qRT-PCR. The 32 genes and their
homology to genes in other species were identified and established by
Hedley et al. (2010), and primer design was based on EST sequences
published in their study. The primer sequences and their homology to
genes in other species can be found in Supplementary Table S2. Genes
previously associated with freezing tolerance included one COR gene
(GT023739), five putative LEA/DHN genes (GT023061, GT023104,
GT024012, GT024386, GT026475), three genes encoding heat-shock
proteins (HSPs, GT024556, GT026070, GT027957) and one pathogen-
esis-related protein gene (PR protein, GT022925). Genes whose pro-
ducts are associated with dormancy release and/or bud break included
acetyl CoA carboxylase (GT022654), calmodulin-binding protein (G-
T027101), beta-tubulin (GT026930), three glutathione S-transferases
(GT023746, GT024480, GT025391) and glutathione peroxidase (G-
T023446). Two genes associated with flowering were included; early
flowering 5 (ELF5, GT022152) and abnormal floral organs (AFO, G-
T022383). Transcript abundance of genes encoding ATP-citrate syn-
thase (GT022005), NAD-dependent isocitrate dehydrogenase (G-
T025907), hexose transporter (GT022502), fructose-1,6-biphosphatase
(GT022744), beta-amylase (GT023730), galactinol synthase (G-
T026466) and trehalose-6-phosphate synthase (GT022650) were de-
termined to assess potential effects of winter warming on carbohydrate
metabolism and respiration. Two genes involved in secondary meta-
bolism were included; caffeic acid o-methyltransferase (GT023252) and
UDP-glucose: flavonoid 7-O-glucosyltransferase (GT026312). Finally, a
CBF1 homologue (GT022122), two putative ethylene response factors
U.B. Andersen et al. Environmental and Experimental Botany 140 (2017) 96–109

(ERFs, GT023545, GT023254) and a CONSTANS homologue (CO, G-
T027482) were included as representatives of transcription factors
presumably involved in regulating freezing tolerance, dormancy or
flowering.
For the RNA extraction, two biological replicate samples were
pooled resulting in three replicate samples per treatment, cultivar and
time point. Total RNA was extracted using the Plant RNeasy Mini
Extraction Kit (Qiagen, Hilden, Germany) according to the manu-
facturer’s recommendations using buffer RLC and including 10% v/v
RNA Isolation Aid (Ambion, Thermo Fisher Scientific, Wilmington, DE,
USA) and 1% v/v β-mercaptoethanol. RNA quantity and quality were
checked using Nanodrop spectrophotometry (NanoDrop, Thermo Fisher
Scientific) and gel electrophoresis. All samples were treated with the
Rapid Out DNA Removal Kit (Thermo Fisher Scientific) two or three
times, to remove genomic DNA. Quantitative PCR with rbcL intron-
specific primers (Supplementary Table S3) was used to ascertain the
absence of genomic DNA. cDNA was synthesized with SuperScript III
reverse-transcriptase (Invitrogen, Thermo Fisher Scientific) and oligo-
dT18 primers. cDNA quality was checked using forward and reverse
primers amplifying 3′ and 5′ regions of the Actin gene (supplementary
Table S3). qRT-PCR was performed in 384-well plates, using an ABI
PRISM 7900HT Fast Real-Time PCR System (Applied Biosystems, Foster
City, CA, USA). Reactions contained 2.5 μL 2 x Fast SYBR Green Master
Mix (Applied Biosystems), 2 μL primer mix (0.5 μM of each gene spe-
cific primer) and 0.5 μL diluted cDNA. To ensure accuracy, primers
were first added to each plate followed by a master mix containing the
cDNA and SYBR Green, and both steps were performed using an
Evolution P3 pipetting robot (PerkinElmer, Zaventem, Belgium). Ct-
values of the selected genes were normalized by subtracting the mean
of the two reference genes eIF4A and Actin (supplementary Table S3).
Primer design of the Actin gene and the reference genes was based on
previously published EST sequences from blackcurrant (Bruni et al.,
2012).
A first step data analysis was performed by Principal Components
Analysis (PCA) applied to the normalized Ct values and the relative
expression values. The PCA indicated that one control sample of ‘Narve
Viking’ sampled in May was an outlier. Hence, this replicate was ex-
cluded from further analysis. A complete list of all expression values
can be found in Supplementary Table S4.
comparing cultivars, treatments and seasonal harvest time points in R.
Polar metabolites were analysed in the same way as the genes. For
the statistical analysis only, metabolite levels were normalized to the
median level of all samples and log10 transformed. The statistical ana-
lyses of relative expression values and metabolite levels only included
data from November to May, as no treatment was employed in
September and October. However, data from control plants collected in
September and October are included in Figs. 4–7Figs. 4–7 to further
illustrate seasonal changes in expression values and metabolite levels.
Principal component analysis (PCA) was performed using the
pcaMethods package in R (Stacklies et al., 2007). Hierarchical clustering
of metabolite levels was performed in MultiExperiment Viewer (MeV)
v. 4.9.0 (Saeed et al., 2003) using a Euclidian distance measure and
average linkage. Spearman rank order correlations were carried out in
SAS (PROC CORR).
3. Results
3.1. Temperature conditions
The winter season 2014–2015 was mild, with average autumn and
winter temperatures being 2.6 °C and 2.3 °C higher than the average
measured between 1961 and 1990 (DMI, 2015). Spring 2015 was closer
to the climate normal, being on average 0.9 °C warmer than the normal.
In the experimental plots, the temperature at 20 cm above the soil
surface was on average 1.3 °C higher in the warming than in the control
plot (Fig. 1A). The temperature difference between the plots varied
between −0.5 and 3.3 °C, being greatest on cold days. A lower tem-
perature in the warming than in control plot was registered in less than
0.25% of the temperature recordings at 20 cm above the soil surface. In
the control plot, the air temperature did not differ at different heights,
but in the warming plot the air temperature at 50 cm and 80 cm above
the soil surface was on average 0.4 °C and 0.7 °C lower than at 20 cm,
indicating that the effect of the warming treatment was greatest closest

2.7. Statistical analyses

Freezing tolerance was estimated as LT50-values, the temperature

representing 50% relative electrolyte leakage (REL) of stems or where
50% of the buds contained dead flower primordia. For each sampling
date, treatment and cultivar, the LT50 values were determined by fitting
a curve to the REL values for all six replicates or the percentage of
survival of flower primordia to temperature using a four-parameter
sigmoid model in SAS (PROC NLIN, SAS Institute, Cary, NC, USA).
Differences between LT50 estimates were regarded as significant if the
95% confidence intervals did not overlap. Due to very varied data, it
was not possible to estimate LT50 for flower primordia of ‘Narve Viking’
in November. Also, flat curves with low maximum or high minimum
REL values prevented estimation of specific stem LT50 values in
November, December, January and March.
Bud break of flower buds was analysed using the cox proportional-
hazards model implemented in the survival package (Therneau, 2012)
in R (R Core Team, 2015). The number of days before reaching the
green tip stage was modelled.
The effects of cultivar and treatment on the number of flowers and
berries per plant, the berry yield and average berry weight were ana-
lysed using a two-way ANOVA in R. To detect differentially expressed
genes, the log2 transformed relative expression values (i.e. the Ct values
normalized to the expression of the reference genes) were analysed
using a three-way ANOVA with correction for multiple testing using the Fig. 1. Daily mean air (A) and soil (B) temperatures at 20 cm height or ∼ 15 cm depth,
Benjamini-Hochberg method (Benjamini and Hochberg, 1995) by respectively, in the control and warming plots during the experimental period.

99
U.B. Andersen et al.
to the warming cables. Accordingly, the average temperature difference
between the plots at 80 cm height was 0.2 °C, varying between −2.5
and 2.9 °C. The temperature variation between the plots was greater at
80 cm than at 20 cm height, and a higher temperature in the control
than in the warming plot was registered in 12% of the temperature
recordings at 80 cm above the soil surface. The temperature was on
average 0.2 °C higher in the control plot than outside the plots, in-
dicating that the wind protection surrounding the experimental plots
caused a slight increase in air temperature. Between late November and
early March the daily mean air temperature dropped below 0 °C on 15
and seven occasions in the control plot and the warming plot, respec-
tively (Fig. 1A), but brief events of frost occurred until April. The ab-
solute minimum temperatures were −8.9 °C and −7.4 °C in the control
and warming plots, respectively. The soil temperature at 15 cm depth
was on average 1.8 °C higher in the warming plot than in the control
plot (Fig. 1B). Soil frost did not occur in either plot.
3.2. Timing of dormancy release
In mid-October, before start of the warming treatment, buds of
‘Narve Viking’ and ‘Zusha’ were endodormant with no sign of bud de-
velopment following one month of forcing (Fig. 2). Forcing in the fol-
lowing months showed that ‘Zusha’ was released from dormancy earlier
than ‘Narve Viking’ (P = 2e−16). Hence, under forcing conditions lat-
eral buds of ‘Zusha’ began to burst in November while lateral buds of
‘Narve Viking’ remained endodormant until late in the winter. The
warming treatment had no effect on the timing of dormancy release of
lateral buds (P = 0.436). Terminal buds of both cultivars generally
broke earlier and reached a higher bud break percentage than lateral
buds, but otherwise the tendency was the same as for the lateral buds
(data not shown).
According to the chill unit accumulation of the < 7.2 °C chill units
model, which has previously been used to predict chill satisfaction and
timing of bud break of blackcurrant in the field (Rose and Cameron,
100
Environmental and Experimental Botany 140 (2017) 96–109
2009), the plants in the control plot had received 53, 494, 1173 and
1803 chilling hours, respectively, at the time of forcing in November,
December, January and February. In comparison, the plants in the
warming plot had received 31, 417, 1050 and 1673 chilling hours,
respectively.
3.3. Freezing tolerance
Freezing tolerance of flower primordia and stems changed season-
ally, but was not significantly affected by seasonal asymmetric warming
(Tables 1 and 2). In September and October, before start of the warming
treatment, freezing tolerance of flower primordia did not change much,
but LT50 values of stems decreased significantly to on average
−19.7 °C. Freezing tolerance continued to increase in late autumn and
winter and reached maximum values in January for stems and in De-
cember, January and February for flower primordia. In March, flower
primordia of ‘Narve Viking’ showed no sign of deacclimation, while
flower primordia of ‘Zusha’ were heavily damaged by all freezing test
temperatures, making it impossible to estimate LT50 for this cultivar
and implying significant deacclimation. Stems deacclimated more
slowly than flower primordia, but earlier or faster deacclimation of
‘Zusha’ than ‘Narve Viking’ was also evident for the stems. Thus, in
April stems of ‘Zusha’ were less freezing tolerant than stems of ‘Narve
Viking’, while in May stem freezing tolerance was comparably low in
both cultivars.
3.4. Bud break, flowering and berry yield
Flower buds of ‘Zusha’ started to break in early March, while bud
break in ‘Narve Viking’ occurred in late March (Fig. 3). This difference
was significant (P < 2e−16). Warming advanced lateral bud break of
‘Zusha’, while the opposite was true for ‘Narve Viking’ as indicated by a
significant interaction between cultivar and treatment (P < 0.001).
The number of flowers per plant was similar for the two cultivars,
Fig. 2. Number of lateral buds of R. nigrum ‘Narve
Viking’ (top) and ‘Zusha’ (bottom) that had attained
each of the four bud stages following 30 days of
forcing in a permissive warming environment at
different times during the winter season. Prior to
forcing plants were exposed to ambient or slightly
elevated winter temperatures. The four uppermost
lateral buds on one shoot of six plants per cultivar
and treatment were evaluated at each time point.
U.B. Andersen et al. Environmental and Experimental Botany 140 (2017) 96–109

Table 1
Freezing tolerance estimated as temperatures where 50% of the buds contained dead flower primordia (LT50) of buds of R. nigrum ‘Narve Viking’ and ‘Zusha’ exposed to ambient or
elevated temperatures during the winter season. The warming treatment was initiated October 22nd. LT50 estimates [mean ± SE (°C)] are shown for six plants tested at eight tem-
peratures. Different letters indicate significant differences between treatments, sampling dates and cultivars.

Cultivar Sep 8th Oct 20th Treatment Nov 17th Dec 16th Jan 12th Feb 9th Mar 16th

LT50 (°C) ± SE LT50 (°C) ± SE

’Narve Viking’ -11.1 ± 3.0 abc

-9.1 ± 0.1 a
Ambient temperature -24.1 ± 1.8de −29.4 ± 2.0e -25.0 ± 0.0d −30.2 ± 1.0e
Elevated temperature -32.5 ± 2.1e −27.8 ± 1.1de −32.4 ± 6.5cde −20.1 ± 7.4abcde
’Zusha’ -6.5 ± 1.0 a
-11.9 ± 0.0b Ambient temperature -19.0 ± 0.0c −28.8 ± 1.3de −25.6 ± 1.2de −26.5 ± 4.4bcde < −10
Elevated temperature -16.0 ± 0.1c -29.5 ± 0.5e −26.2 ± 0.3e −30.5 ± 4.1de < −10

Table 2
Freezing tolerance estimated as temperatures representing 50% REL (LT50) of stems of R. nigrum ‘Narve Viking’ and ‘Zusha’ exposed to ambient or elevated temperatures during the winter
season. The warming treatment was initiated 22nd of October. LT50 estimates [mean ± SE (°C)] are shown for six plants tested at eight temperatures. Different letters indicate significant
differences between treatments, sampling dates and cultivars.

Cultivar Sep 8th Oct 20th Treatment Nov 17th Dec 16th Jan 12th Feb 9th Mar 16th Apr 27th May 24th

LT50 (°C) ± SE LT50 (°C) ± SE

’Narve −8.3 ± 1.0 ab

-19.9 ± 1.6 cde
Ambient temp. – – – -32.2 ± 4.5efghi – -17.6 ± 1.1c −9.1 ± 0.8b
Viki- Elevated temp. – -17.7 ± 3.1cde -40.6 ± 2.8hi -24.4 ± 4.5defg – -18.5 ± 1.3 cd -8.7 ± 0.7 b
ng’
’Zusha’ −5.6 ± 0.5a -19.4 ± 1.7cd Ambient temp. – -19.6 ± 3.0cdef -42.7 ± 2.1i -34.5 ± 3.3ghi -25.4 ± 0.5f -7.3 ± 1.4 ab
-7.1 ± 0.5 ab
Elevated temp. – -21.3 ± 3.3cdef -42.4 ± 1.2 i -31.2 ± 3.2efgh -25.3 ± 1.2ef -7.5 ± 0.6 ab
-6.6 ± 0.5 ab

Month
4

Sept
Oct
PC3 (10.4%)
2

Nov
Dec
0

Jan
Feb
−2

Mar
Apr
May
−4

−4 −2 0 2 4 6 8
PC1 (40.9%)
Fig. 4. Score plots from principal components analysis (PCA) applied to a metabolite
profiling dataset from floral buds of R. nigrum ‘Narve Viking’ (squares) and ‘Zusha’ (cir-
Fig. 3. Average attainment of each of the four bud stages of lateral buds of R. nigrum cles) exposed to ambient (open symbols) or slightly elevated (closed symbols) tempera-
‘Narve Viking’ and ‘Zusha’ during bud break in the field. During the winter season, plants tures from late October to mid-April the following year (n = 6). Buds were sampled
were grown at ambient or slightly elevated temperatures. Values are means ± SE of the monthly. However, since the temperature treatment was initiated in late October, only
four uppermost lateral buds on n = 21 shoots. buds of plants exposed to ambient temperatures were sampled in September and October.

but elevated winter temperatures were associated with a reduction in
the total number of flowers per plant at full bloom for both cultivars
(Table 3, P = 0.020). In ‘Zusha’ only, a reduction in the number of
flowers of warmed plants led to a significantly lower berry yield per
plant (P = 0.020). Warming had no effect on the number of berries per
plant or the individual berry weight, but ‘Zusha’ produced fewer
(P < 0.001) and larger (P < 0.037) berries than ‘Narve Viking’.
3.5. GC–MS metabolite profiling
Principal component analysis (PCA) was used to identify the largest
variance components in the metabolite data (Fig. 4). Principal com-
ponent 1 (PC1) explained 40.9% of the variance and separated profiles
from plants collected in April and May from those collected in Sep-
tember to February or March, with the profile of ‘Zusha’ collected in
March located in between. PC3 explained 10.4% of the variance and
separated the metabolite profiles of the two cultivars. PC2 explained
13.5% of the variance but did not yield any meaningful separation of
cultivars, time points or treatments (not shown).
A total of 77 metabolites showed significantly changed pool sizes
depending on treatment, cultivar, seasonal time or an interaction as
revealed by three-way ANOVA. Arabinose was the only metabolite
whose pool size varied significantly between control and warmed plants
and only ascorbic acid was significantly affected by an interaction be-
tween treatment, cultivar and seasonal time. Except for methionine,
G6P and glucose, 1,6-anhydro-, beta-, pool sizes of all significantly
regulated metabolites changed according to seasonal time or an inter-
action between cultivar and seasonal time, while 40 metabolites were
significantly affected by cultivar or an interaction between cultivar and
seasonal time. Clustering of the metabolites revealed seven major pat-
terns of metabolic responses (Fig. 5, Supplementary Table S5): (a)
Metabolites whose pool sizes were high in September and October and
again in May. This cluster was small, containing only two amino acids
and one unknown compound. (b) Metabolites which were found in
relatively low amounts at most sampling time points, although their
abundances varied with seasonal time and, in some cases, between

101
U.B. Andersen et al. Environmental and Experimental Botany 140 (2017) 96–109

Fig. 5. Global overview of the dynamic changes in primary metabolism in flower buds of R. nigrum ‘Narve Viking’ (NV) and ‘Zusha’ (Z) from September to May the following year. Plants
were grown at ambient temperature (C) or slightly elevated temperatures (W) from late October to mid-April (n = 6). The metabolic profiles collected in November-May were statistically
analysed using a three-way ANOVA at FDR P < 0.05 by comparing all conditions against each other. Metabolites that showed significant differences were clustered using Euclidian
distance and average linkage. The panel shows normalized metabolite intensities as indicated by the different colors. Lists with the names of all metabolites in the different clusters can be
found in Supplementary Table S5. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

cultivars. This cluster included, among others, glucose, fructose, su-
crose and raffinose and citric and malic acid. (c) Metabolites which
increased strongly in April and May and, in some cases, in March in
‘Zusha’. Metabolites in cluster (d) also increased dramatically in
abundance in spring, but these metabolites were additionally present in
high amounts in September. Cluster (c) included seven amino acids, five
polyhydroxy acids, three acids, two phenylpropanoids, two phosphates,
one N-compound and an unknown compound, while cluster (d) only
contained three metabolites. Pool sizes of metabolites in cluster (e)
were high in October and again in March or April in ‘Zusha’ and ‘Narve
Viking’, respectively. This cluster contained G6P and two unknown
compounds. Metabolites in cluster (f) showed a distinct decrease in
abundance in April and May and, in some cases, in March in ‘Zusha’.
This cluster was dominated by six phenylpropanoids. Metabolites in
cluster (g) tended to increase in abundance in spring, but more of these
metabolites were also found in large pool sizes during late fall and
winter, and the increase was therefore not very pronounced. In addi-
tion, many of these metabolites were found in greater amounts in
‘Zusha’ than in ‘Narve Viking’. This cluster contained ten metabolites
from various major classes. In addition, five small clusters were iden-
tified, containing one or two metabolites, including gluconic acid,
maltose, xylose and three unknown compounds. Gluconic acid and
maltose were up-regulated in ‘Narve Viking’ relative to ‘Zusha’ in the
period from September to March, while xylose was up-regulated in
April and May in ‘Zusha’ and ‘Narve Viking’, respectively.
3.6. Gene expression analysis
To investigate whether effects of winter warming on blackcurrant
was visible at the transcriptional level, we determined the transcript
abundance of 32 selected genes with homology to genes associated with
overwintering or spring phenology. According to a PCA, PC1 separated

102
U.B. Andersen et al. Environmental and Experimental Botany 140 (2017) 96–109

significantly depending on seasonal time, cultivar or an interaction

4
Month between two or all three factors (Supplementary Table S6). Eight genes
3

Sept were significantly differentially expressed between plants at ambient

Oct temperature and plants at slightly elevated temperature. This group
PC2 (21.1%)
2

Nov included genes with homology to AFO (GT022383), ELF5 (GT022152),

Dec HSPs (GT024559, GT026070, GT027957), LTCOR12 (GT023739), beta-
1

Jan tubulin (GT026930) and caffeic acid O-methyltransferase (GT023252).

Feb
0

Transcript abundances of 15 additional genes were significantly af-

Mar fected by an interaction between treatment and seasonal time or be-
−2 −1

Apr tween treatment, cultivar and seasonal time.

May
−4 −2 0 2 4 GT026475) and one of the HSP homologues (GT024559) were up-
PC1 (58.9%)
Fig. 6. Score plots from principal components analysis (PCA) of the relative expression of
32 genes in floral buds of R. nigrum ‘Narve Viking’ (squares) and ‘Zusha’ (circles) exposed
to ambient (open symbols) or slightly elevated (closed symbols) temperatures from late
October to mid-April the following year. Buds were sampled monthly (n = 3). However,
since the temperature treatment was initiated in late October, only buds of plants exposed
to ambient temperatures were sampled in September and October.
samples according to seasonal time (Fig. 6) and explained 58.9% of the
total variance. The distribution of the samples followed an order from
November to December and October, January, February, March and
September, April and May. A distinct shift from left to right along PC1
occurred around March-April in ‘Narve Viking’ and around February-
March in ‘Zusha’. PC2 separated profiles of ‘Zusha’ collected in March
and profiles of both cultivars collected in September and October from
all other profiles and explained 21.1% of the variance. The two culti-
vars were distributed furthest apart in March and to a lesser extent in
February. The transcript abundance of all investigated genes varied
CBF1 (GT022122), LTCOR12 (GT023739), PR10 (GT022925), four
of the five LEA/DHN homologues (GT024386, GT024012, GT0023061,
regulated during the winter period, consistent with a putative role in
freezing tolerance (Fig. 7, Supplementary Table S7). In contrast, tran-
script abundances of the type II SK2 dehydrin (GT023104) and the
other two HSPs (GT026070, GT027957) decreased in autumn and in-
creased in spring. Bud break and flowering in April and May were as-
sociated with increased transcript abundances of calmodulin-binding
protein (GT027101), acetyl CoA carboxylase (GT022654), beta-tubulin
(GT026930), CO (GT027482), glutathione peroxidase (GT023446) and
two glutathione S-transferases (GT023746, GT025391). ‘Zusha’, which
showed earlier bud break than ‘Narve Viking’, also showed an earlier
onset to increasing transcript abundances of the genes putatively en-
coding calmodulin-binding protein, acetyl CoA carboxylase, CO, glu-
tathione peroxidase and glutathione S-transferases. The transcript
abundance of the beta-tubulin homologue (GT026930) was sig-
nificantly affected by the warming treatment, being higher in buds of
warmed than control plants in spring. Transcript abundance of the third
glutathione S-transferase (GT024480) was highest in winter and
reached its minimum level around bud break. The AFO (GT022383) and

Fig. 7. Expression of selected genes in floral buds of R. nigrum ‘Narve Viking’ (NV) and ‘Zusha’ (Z) from September to May the following year. Plants were grown at ambient temperature
(C) or slightly elevated temperatures (W) from late October to mid-April (n = 3). The top panel shows medium to highly expressed genes, while the bottom panel shows lowly expressed
genes. Both panels show relative gene expression (2−ΔCt) as indicated by the different intensity of the red color. The numerical values for 2−ΔCt are given in Supplementary Table S7. (For
interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

103
U.B. Andersen et al.
Table 3
Number of flowers and cropping performance of R. nigrum ‘Narve Viking’ and ‘Zusha’ exposed to ambient or slightly elevated temperatures during the winter season. Values given are
means ± SE of n = 21 plants. Different letters indicate significant differences between treatments and cultivars within columns.
Cultivar Treatment Flowers per plant
‘Narve Viking’ Ambient temperature 451 ± 22a
‘Narve Viking’ Elevated temperature 384 ± 29 b
‘Zusha’ Ambient temperature 491 ± 40 a
‘Zusha’ Elevated temperature 409 ± 32 b
ELF5 (GT022152) homologues were significantly differentially ex-
pressed between buds of plants at ambient and elevated temperatures.
Their transcript abundances were highest in September, low in the rest
of the fall and winter and increased transiently in spring. Spring growth
was also associated with increased transcript levels of the ATP-citrate
synthase (GT022005), NAD-dependent isocitrate dehydrogenase (G-
T025907), beta-amylase (GT023730) and fructose-1,6-biphosphatase
(GT022744). Transcripts of galactinol synthase (GT026466), trehalose-
6-phosphate synthase (GT022650) and the hexose transporter (G-
T022502) accumulated during autumn and winter, decreased in late
winter and early spring and accumulated again in bursting buds. The
two genes associated with secondary metabolism showed different ac-
cumulation patterns. Thus, caffeic acid o-methyltransferase (G-
T023252) was repressed in developing buds, while UDP-glucose: fla-
vonoid 7-O-glucosyltransferase (GT026312) was transiently up-
regulated in spring. Transcript abundances of the two ERFs (GT023545,
GT023254) were highest in September and October and then decreased.
3.7. Correlation of freezing tolerance with temperature, compatible solutes
and transcripts
LT50 of stems and buds of ‘Zusha’ and stems of ‘Narve Viking’ cor-
related with the average air temperatures since the last sampling date
(Fig. 8, data for LT50 stems not shown, numerical P-values and the
corresponding Spearman rank correlation coefficients are given in
supplementary Table S8). Correlations between levels of soluble car-
bohydrates or transcripts associated with freezing tolerance and LT50 of
buds were mostly non-significant. Only fructose and transcripts of PR10
and GolS showed medium to strong negative correlations with LT50 (i.e.
positive correlation with freezing tolerance) of buds of ‘Narve Viking’
and/or ‘Zusha’. This suggests that under natural conditions, including
both cold acclimation and deacclimation processes, regulation of sugar
levels or transcript abundances are not exclusively associated with
freezing tolerance. Few additional metabolites showed correlations
with LT50 of buds. However, these correlations were generally positive
(i.e. negative correlation with freezing tolerance), with metabolites
increasing in abundance in breaking buds.
A correlation analysis among the 11 transcripts putatively asso-
ciated with freezing tolerance, including CBF1, and sugar levels showed
significant positive correlations among CBF1, LTCOR12, LEA14, DHN2,
DHN6, HSP17.4, PR10, raffinose and ribose in both cultivars and su-
crose in ‘Zusha’. In contrast, the three genes encoding products with
homology to two 70k Da HSPs and a type II SK2 dehydrin generally
correlated negatively with raffinose, ribose and the other presumably
cold-induced genes. Levels of glucose and fructose in both cultivars, and
maltose in ‘Narve Viking’ correlated negatively with the average air
temperatures since the last sampling date. In contrast, transcript levels
of the cold related genes were generally not correlated with tempera-
ture data, i.e. the average air temperatures since the last sampling date
or the daily minimum temperature, which occurred the night prior to
sample collection (the latter is not shown in Fig. 8).
104
Environmental and Experimental Botany 140 (2017) 96–109
Berry yield (g plant−1) Berries per plant Weight per berry (g)
337 ± 16 a
349 ± 18 a
0.97 ± 0.02 a
320 ± 25 a
343 ± 29 a
0.95 ± 0.02 a
314 ± 34 a
293 ± 31 ab
1.04 ± 0.03 b
206 ± 25 b
205 ± 24 b
0.98 ± 0.03 b
4. Discussion
4.1. Seasonal asymmetric warming significantly affects spring phenology of
blackcurrant
Despite a small average temperature increase, well within the limits
of the projected future temperature increases, our results reveal that
seasonal asymmetric warming significantly affected spring phenology
of blackcurrant (Fig. 3). As expected, ‘Zusha’ had a lower chilling re-
quirement than ‘Narve Viking’, indicated by the earlier dormancy re-
lease of this cultivar (Fig. 2). According to the < 7.2 °C chill units
model the number of chill units in the two plots was on average 1111 in
January, at which time ‘Zusha’ was released from dormancy. In con-
trast, ‘Narve Viking’ was not released from dormancy until after the last
forcing in February, indicating a chilling requirement of more than
1803 chill units. This is in line with our previous observations, showing
that ‘Narve Viking’ requires more than 2000 chill units for dormancy
release (Pagter et al., 2015). Consistent with a high chilling require-
ment, winter warming significantly delayed floral bud break of ‘Narve
Viking’. Although we did not detect a delay in dormancy release of
warmed plants, this is likely a result of suboptimal chilling, increasing
the heat requirement for bud break (Harrington et al., 2010; Fu et al.,
2012; Jeong et al., 2012). The apparent lack of a delay of dormancy
release may be due to the rather coarse time scale of evaluation. Al-
ternatively, ‘Narve Viking’ may not have been fully released from
dormancy in February, when the evaluation of dormancy status was
terminated. In contrast, warming significantly advanced bud burst of
the low-chilling requiring ‘Zusha’, implying that the magnitude of this
cultivars’ response to warming was greater during the heat accumula-
tion phase in spring than during the dormancy period. Increased soil
temperature has been shown to advance bud flush in some woody
perennials (De Barba et al., 2016; Greer et al., 2006), whereas other
studies have found no effect of soil temperature on bud phenology
(Bailey and Harrington, 2006; De Barba et al., 2016). Hence, we cannot
rule out that advanced bud burst of ‘Zusha’ at elevated temperature was
driven by both the increase in air and soil temperatures.
In accordance with our hypothesis, genotypic and treatment dif-
ferences in timing of dormancy release and spring growth were partly
visible at the transcriptional level (Fig. 7). Thus, up-regulation of the
calmodulin-binding protein gene (GT027101) and acetyl CoA carbox-
ylase (GT022654), which co-localise with bud break QTL in black-
currant (Hedley et al., 2010), appeared earlier in ‘Zusha’ than in ‘Narve
Viking’, consistent with earlier bud break of the former than the latter
cultivar. Advanced spring phenology of ‘Zusha’ compared to ‘Narve
Viking’ was also reflected in earlier up-regulation of the CO homologue
(GT027482). CONSTANS is involved in controlling the photoperiodic
pathway of flower induction in both annual and perennial plants
(Böhlenius et al., 2006; Shim and Imaizumi, 2015), and in Vitis vinifera
a CONSTANS-like gene co-localizes with QTL for flowering time
(Duchêne et al., 2012). Significantly different transcript abundance of
the beta-tubulin (GT026930), AFO (GT022383) and ELF5 (GT022152)
homologues in floral buds of warmed plants compared to control plants
suggests that asymmetric warming may directly modulate the expres-
sion of genes associated with growth and development (Fig. 7). Mor-
phogenesis and the cytoskeleton have a close functional relationship,
U.B. Andersen et al. Environmental and Experimental Botany 140 (2017) 96–109

CBF CO PR1 LEA DHS DH6 DH2 HSP HSP LEA HSP Gol Tem LT50 Fru Glc Mal Raf Rib Suc
1 R12 0 5 KII 17.4 70 14 70 S p bud
s
CaCBF1 + + - + + + - + - + +
(GT022122)
LTCOR12 + + - + + + - + - + + +
(GT023739)
PR10 + + - + + + - + - - +
(GT022925)
LEA5 + + - +
(GT023061)
DHSKII - - - - - + - + + - - -
(GT023104)
DHN6 + + + - + + - + - + +
(GT024012)
DHN2 + + + - + + - + - - + +
(GT024386)
HSP17.4 + + + - + + - + - + + +
(GT024559)
HSP70 - - + + - - - - + -
(GT027957)
LEA14 + + + - + + + - - + + +
(GT026475)
HSP70 - - - + - - - + - - - -
(GT026070)
GolS - + - - -
(GT026466)
Mean temp. - - -
LT50 buds - + -
Fructose - - - +
Glucose - +
Maltose + + - + + +
Raffinose + + + - + + + - + - + +
Ribose + + - + + + - + - +
Sucrose + + - + - + - + + +
Fig. 8. Correlation analysis between freezing tolerance, average temperature since the last sampling date, transcript abundances or sugar contents in R. nigrum ‘Narve Viking’ (top panel)
or ‘Zusha’ (lower panel) exposed to ambient or slightly elevated winter temperatures. Plant material was harvested monthly from September to May. P ≤ 0.05 (yellow), P ≤ 0.01
(orange), P ≤ 0.001 (red). + and − indicate positive and negative correlations, respectively. The numerical P-values and correlation coefficients can be found in Supplementary Table
S8. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

and beta-tubulin, the basic structural unit of microtubules required for and decreased in spring, suggesting different roles for this enzyme in
cell division and cell elongation, is considered a marker for monitoring overwintering and spring growth, e.g. nitrogen storage in addition to its
dormancy in tree buds (Rohde et al., 2007; Xu et al., 2016). In Arabi- intrinsic role in stress defence (Nomura et al., 2007). Abundance of
dopsis thaliana, AFO regulates developmental processes such as shoot glutathione S-transferase protein also decreases in xylem of deaccli-
apical meristem activity and inflorescence phyllotaxy (Boter et al., mating Hydrangea paniculata (Pagter et al., 2014).
2015). Hence, high expression of AFO in blackcurrant floral buds in
September is likely associated with specification of floral primordia,
4.2. Regulation of seasonal transitions in freezing tolerance and dormancy
while a transient increase in transcript abundance in spring suggests an
involvement of AFO in flower development (Lugassi et al., 2010). The
The transcript abundance of the two ethylene response factor
expression pattern of ELF5 resembled that observed in flower buds of
homologues ERF3 (GT023254) and especially ERF6 (GT023545) was
Poncirus trifoliata, where ELF5 acts as a floral inducer (Zhang et al.,
highest in September and then decreased (Fig. 7). The timing of floral
2011). Thus, ELF5 may play a similar role in blackcurrant.
initiation and dormancy induction of blackcurrant cultivars under
Oxidative stress is an important factor controlling bud break, as
naturally decreasing photoperiodic conditions starts in early August in
evident from increased expression of genes whose products are asso-
southern Scandinavia (Sønsteby et al., 2012). Hence, upregulation of
ciated with scavenging of free radicals (Mazzitelli et al., 2007; Ueno
ERFs in early September indicates that ethylene may play a role in the
et al., 2013), such as glutathione S-transferases in blackcurrant (Hedley
activity-dormancy transition in blackcurrant. Ethylene is one of the
et al., 2010). In our study, an earlier onset of transcript abundance
major signals that act consecutively to control growth cessation, bud
increases of two glutathione S-transferases (GT023746, GT025391) and
development and dormancy induction in response to short day condi-
a glutathione peroxidase (GT023446) in ‘Zusha’ than in ‘Narve Viking’
tions in poplar (Populus tremula x Populus alba) (Ruttink et al., 2007). A
supports the assumption that bud break is coupled to an increased need
role during endodormancy induction has also been reported in Betula
for free radical removal (Fig. 7). Transcript abundance of the third in-
pendula and Prunus persica (Ruonala et al., 2006; Zhong et al., 2013).
vestigated glutathione S-transferase (GT024480) was highest in winter
Elevated winter temperatures did not increase the risk of freeze-

105
U.B. Andersen et al.
induced damage to stems and flower primordia (Tables 1 and 2), sup-
porting the suggestion that blackcurrant is resistant to mild winter
warming (Pagter et al., 2015). However, a positive correlation between
the average air temperatures since the last sampling date and LT50
values of the flower buds of ‘Zusha’ (Fig. 8) and stems of both cultivars
(Supplementary Table S8) highlight the central importance of tem-
perature in driving cold acclimation and deacclimation. It is likely that
greater temperature elevations will increase the risk of freeze-induced
damage in blackcurrant, as observed in other woody perennials (Repo
et al., 1996; Taulavuori et al., 2004). Interestingly, transcript abun-
dance of LTCOR12 was significantly affected by the warming treatment
(Fig. 7), suggesting that seasonal asymmetric warming may induce
changes in the CBF cold response pathway, although freezing tolerance
was unaffected.
The strong correlations among CBF1, COR12, DHN6, DHN2, LEA14,
HSP17.4 and PR10 transcripts (Fig. 8) indicate that expression of these
genes is coordinated during seasonal transitions in freezing tolerance,
or may even be part of the same regulatory pathway in blackcurrant.
Transcripts accumulated in autumn and winter, but generally their
abundance peaked earlier than maximum freezing tolerance and started
to decrease well before any measurable decrease in freezing tolerance
occurred, resulting in a lack of correlations between cold-related gene
expression and LT50. Hence, the current results suggests that cold-sig-
nalling pathways mediated by the CBF1 homologue, which may include
the LEA and DHN genes, and the expression of other cold-related genes
are dampened in January or February before significant deacclimation.
Future proteomic studies are needed to elucidate whether the encoded
proteins are more stable, or whether a decrease in transcript abundance
leads to a comparable reduction in protein levels. In Hydrangea pani-
culata, deacclimation was associated with a gradual decline in several
bark dehydrins (Pagter et al., 2014), indicating that in this species
freezing tolerance during deacclimation was temporally correlated with
changes in protein contents. However, it is not clear whether this was
due to transcriptional regulation or protein stability, since the corre-
sponding transcript levels were not measured. There is no indication
that CBF1 is involved in the regulation of dormancy in blackcurrant
(Wisniewski et al., 2014). Hence, the CBF1 homologue was not up-
regulated until October, suggesting that it is not involved in short-day
induced dormancy induction, which typically starts in August when
extension growth has slowed down (Sønsteby et al., 2012). In addition,
transcript abundance peaked simultaneously in the two cultivars de-
spite different chilling requirements.
The accumulation pattern of the LEA5 homologue (GT023061) was
similar to the accumulation pattern of LEA14 until bud break, where
LEA5 increased significantly (Fig. 7). Up-regulation of LEA5, the type II
SK2 dehydrin gene (GT023104) and two of the HSP genes (GT026070,
GT027957) in spring may be related to osmotic rather than freezing
stress and protection of growing tissues from dehydration (Kontunen-
Soppela and Laine, 2001).
4.3. Modulation of carbohydrate metabolism
Most of the non-phosphorylated soluble carbohydrates belonged to
cluster (b), which included metabolites found at relatively low levels at
all sampling time points (Fig. 5). However, a closer inspection of the
sugar levels showed that levels of raffinose, fructose, glucose and
maltose generally increased in autumn and decreased in spring in
agreement with biochemical measurements of fructose and raffinose
(Pagter et al., 2015). Accordingly, seasonal changes in contents of
fructose correlated with LT50 of buds in ‘Narve Viking’ (Fig. 8). In
‘Zusha’, however, accumulation patterns were more varied, suggesting
that soluble sugars are less central in seasonal transitions in freezing
tolerance in this cultivar. It is possible that alternative compatible so-
lutes are of greater importance during cold acclimation and deaccli-
mation of ‘Zusha’. Accumulation of soluble sugars in cold acclimating
buds may partly be supported by the upregulation of the hexose
106
Environmental and Experimental Botany 140 (2017) 96–109
transporter homologue (GT022502), which encodes a membrane pro-
tein shuttling monosaccharides across cellular membranes (Slewinski,
2011). Transcript abundance of the gene encoding this transporter
decreased during deacclimation, but increased again in flushing buds
(Fig. 7). Sherson et al. (2003) suggested that meristematic cells take up
free hexoses, in contrast to sucrose uptake by differentiated cells.
Therefore, increased hexose transport activity in spring could facilitate
the uptake of hexoses to satisfy the energy requirements of the flower
bud meristems.
Levels of sucrose were highly variable between replicates and in
contrast to previous observations (Pagter et al., 2015) seasonal time or
winter warming had no coherent effect on sucrose levels (Fig. 5). A
smaller effect of warming on carbohydrate metabolism than previously
observed may be due to warmer winter temperatures in this experi-
ment. Hence, in our previous experiment the warming treatment raised
the average temperature throughout the treatment period from the
freezing point to about 1 °C, whereas in the current study the tem-
perature was raised from 2.8 °C to 5.1 °C. Since the rate of cellular re-
spiration, which constitutes a major sink for soluble carbohydrates
during winter, is largely a function of temperature, the thermal con-
ditions may have profound effects on soluble carbohydrate availability
(Sperling et al., 2015; Parada et al., 2016).
Consistent with the accumulation pattern of raffinose (Fig. 5), ga-
lactinol synthase (GolS; GT026466), which catalyzes the first step in the
biosynthesis of raffinose family oligosaccharides (RFOs), was up-regu-
lated in winter (Fig. 7). In April and May, GolS transcript abundance
increased again and so did the level of raffinose, although to a lesser
extent than during winter. In buds of Quercus petrea, up-regulation of
GolS at the onset of bud burst led the authors to propose that RFOs play
a role in maintaining membrane and protein integrity in developing
bud tissues (Derory et al., 2006). Rising transcript levels of beta-amy-
lase (GT023730, Fig. 7) in April and May seem late for starch mobili-
zation to supply expanding buds with soluble carbohydrates, especially
considering the advanced developmental stage of the buds and con-
current increased transcript abundance of fructose-1,6-biphosphatase
(GT022744), indicating photosynthetic activity of flushing buds. In
Populus deltoids and P. balsamifera, high gene expression for beta-amy-
lase in growing twigs may be involved in maintaining sufficient levels
of soluble sugars for growth through starch degradation (Alluvada
et al., 2014).
Trehalose-6-phosphate synthase (TPS) is a key enzyme in the tre-
halose biosynthetic pathway. It catalyzes the biosynthesis of trehalose-
6-phosphate (T6P), which has emerged as an important signalling
metabolite (Lunn et al., 2014). Up-regulation of TPS1 (GT022650) in
November-February and again in April-May (Fig. 7) suggests that tre-
halose metabolism plays a role in both freezing tolerance and bud break
or flowering of blackcurrant. In Triticum aestivum, TPS gene expression
is induced by freezing conditions (Xie et al., 2015) and in A. thaliana
T6P and TPS1 play essential roles in plant development and the reg-
ulation of flowering (Ponnu et al., 2011). The present results suggest
that TPS plays an important role also in flowering of blackcurrant.
4.4. Bud break is associated with a dramatic remodeling of the metabolome
Asymmetric warming had minor metabolic implications, but pro-
nounced seasonal accumulation patterns of specific metabolites were
evident (Fig. 5). Especially bud break was associated with a dramatic
remodeling of the metabolome, with differential seasonal regulation of
metabolism between the early flushing ‘Zusha’ and the later flushing
‘Narve Viking’. Energy-related metabolites such as phosphoric acid, key
intermediates of glycolysis/gluconeogenesis, such as F6P and G6P, and
some TCA cycle intermediates (e.g. succinic acid and pyruvic acid)
showed increasing levels during bud development. Levels of other TCA
cycle intermediates (citric acid and malic acid) were relatively stable
throughout the experiment. ‘Zusha’, which showed earlier bud break
than ‘Narve Viking’ (Fig. 3), also showed an earlier onset of increases in
U.B. Andersen et al.
some of these metabolites (e.g. phosphoric acid, F6P, G6P and succinic
acid). Increasing respiratory activity during bud break was additionally
suggested by increased transcript levels of genes encoding ATP-citrate
synthase (GT022005) and NAD-dependent isocitrate dehydrogenase
(GT025907). Resuming growth requires activation of protein bio-
synthesis, and metabolites increasing in abundance in developing buds
included several amino acids, including valine, the aspartate-derived
amino acids methionine and threonine, and serine and glycine, which
are in addition a major source of one-carbon units (Mouillon et al.,
1999). Increasing levels of proline in developing buds are in contrast to
observations in buds of Picea abies and Abies alba (Dhuli et al., 2014)
and may seem surprising, considering the prominent role of proline in
plant freezing tolerance. However, proline also plays an important role
in plant growth and differentiation across the life cycle, including
flower development (Kavi Kishor et al., 2015). Bud break was also as-
sociated with increasing abundances of a number of polyhydroxy acids
(i.e. glyceric, erythronic, ascorbic and threonic acids). Ascorbic acid has
proposed functions in photosynthesis as an enzyme cofactor and in
control of cell growth (Smirnoff and Wheeler, 2000), while threonic
and glyceric acid are products of ascorbic acid degradation (Loewus,
1999). Myo-inositol and its derivatives arabinose and xylose increased
transiently or continuously in buds in spring, indicating increased
synthesis of cell wall material in developing buds. Floral bud develop-
ment was also associated with increasing levels of shikimic acid, the
central precursor for the biosynthesis of aromatic amino acids and
phenolic structures (Maeda and Dudareva, 2012). This agrees with our
observation that the levels of luteolin in ‘Narve Viking’, 3-caffeoyl-
quinic acid and 5-caffeoylquinic acid in both cultivars and transcript
abundance of the UDP-glucose: flavonoid 7-O-glucosyltransferase gene
(GT026312), which mediates glycosylation of flavonoids, increased
transiently in spring. In contrast, levels of several other phenolic com-
pounds (cis- and trans caffeic acid, ferulic acid, trans 4-hydroxycinnamic
acid, 4-hydroxybenzoic acid, epigallocatechin, catechin and gallic acid)
and transcript abundance of the caffeic acid O-methyltransferase
homologue (GT023252), which catalyzes multiple reactions in the
phenylpropanoid biosynthetic pathway (Ferrer et al., 2008), decreased
in developing buds (Fig. 7). Transcript abundance of caffeic acid O-
methyltransferase was significantly lower in warmed than control
plants, suggesting that some of the changes in levels of secondary me-
tabolites in response to asymmetric warming are due to transcriptional
regulation. In A. thaliana, cold acclimation leads to massive increases in
flavonol and anthocyanin content (Schulz et al., 2015) and decreasing
levels of these compounds may be associated with deacclimation (Dhuli
et al., 2014; Schulz et al., 2016). The present data indicate that loss of
freezing tolerance and floral bud development are associated with
distinct shifts in different modules of secondary metabolism. The
functional roles of these metabolites in transitions in plant freezing
tolerance and development, however, are not clear.
4.5. Seasonal asymmetric warming may reduce fruit yield
Winter warming significantly reduced the flower number of both
cultivars and in ‘Zusha’ this was accompanied by a reduction in fruit
yield (Table 3). The reduction in flower number is not attributable to
changes in initiation of flower primordia or to the warming effect on the
dormancy response, as flower primordia presumably were formed be-
fore initiation of the warming treatment (Tinklin and Schwabe, 1970)
and the reduction in flower number was similar in the two genotypes,
despite significantly different chilling requirements. Rather it may be a
result of flower bud abscission or loss of flower primordia. Hence, in
other fruit crops mild winter temperatures increase flower abscission
and necrosis of flower primordia, possibly due to inadequate resource
supply or utilization (Bonhomme et al., 2005; Atkinson et al., 2013).
When comparing flower and fruit numbers, it is evident that relatively
more flowers of ‘Narve Viking’ than ‘Zusha’ developed into fruit,
especially in the warming treatment. Possibly this cultivar-specific
107
Environmental and Experimental Botany 140 (2017) 96–109
effect is due to differences in flowering phenology. Earlier flowering in
‘Zusha’ than in ‘Narve Viking’ may be associated with a reduction in the
fruit set ratio (i.e. the ratio of fruiting flowers to flowers per plant) due
to fewer pollinators. Although blackcurrant has self-pollinating flowers,
self-pollination is not efficient and bees are the main pollinator of
blackcurrant (Fliszkiewicz et al., 2011). In our previous study, winter
warming had no effect on flower abundance (Pagter et al., 2015). The
flower numbers were, however, registered differently in the two stu-
dies: In the previous study, only flowers from the four top-most lateral
buds were recorded while in the present study the total number of
flowers per plant was recorded. The latter takes into account effects
independent of bud position along the parent shoot.
4.6. Conclusions
In conclusion, we found evidence that seasonal asymmetric
warming alters spring phenology and cropping performance of black-
currant. The former may be associated with modulation of the expres-
sion of specific genes associated with growth and development, while
the negative effect of a decline in winter chilling on fruit yield pre-
sumably is attributable to a decline in flower number. Winter warming
did not increase the risk of frost injuries to flower primordia or stems
and had only minor impact on primary metabolism of flower buds. In
addition, the data have shed light on the timing of metabolic and
transcriptional regulation during deacclimation, dormancy release and
bud burst in blackcurrant, including identification of metabolites con-
nected with spring growth, regulation of genes related to dormancy
release or floral organ development and molecular evidence for the role
of CBF and cold-induced gene homologues in seasonal transitions in
freezing tolerance. This will further support discussions about reg-
ulatory pathways of phenological events in woody perennials.
Acknowledgements
We thank Annette Steen Brandsholm, Connie Krogh Damgaard, Finn
Kristiansen, Ines Fehrle and Kaj Ole Dideriksen for outstanding tech-
nical assistance. This work was supported by the Danish Council for
Independent Research | Technology and Production Sciences [Grant
No. DFF-1335-00182 to MP] and by the Max-Planck Society.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.envexpbot.2017.06.005.
References
Alluvada, J., Fokar, M., Holaday, A.S., 2014. Variation in gene transcription and protein
for key enzymes of carbohydrate metabolism in poplar xylem with respect to growth
phase. Acta Physiol. Plant. 36, 2385–2395.
Atkinson, C.J., Brennan, R.M., Jones, H.G., 2013. Declining chilling and its impact on
temperate perennial crops. Environ. Exp. Bot. 91, 48–62.
Böhlenius, H., Huang, T., Charbonnel-Campaa, L., Brunner, A.M., Jansson, S., Strauss,
S.H., Nilsson, O., 2006. CO/FT regulatory module controls timing of flowering and
seasonal growth cessation in trees. Science 312, 1040–1043.
Bailey, J.D., Harrington, C., 2006. Temperature regulation of bud-burst phenology within
and among years in a young Douglas-fir (Pseudotsuga menziesii) plantation in western
Washington, USA. Tree Physiol. 26, 421–430.
Benjamini, Y., Hochberg, Y., 1995. Controlling the false discovery rate: a practical and
powerful approach to multiple testing. J. R. Stat. Soc. 57, 289–300.
Bonhomme, M., Rageau, R., Lacointe, A., Gendraud, M., 2005. Influences of cold depri-
vation during dormancy on carbohydrate contents of vegetative and floral primordia
and nearby structures of peach buds (Prunus persica L. Batch). Sci. Hortic.
(Amsterdam) 105, 223–240.
Boter, M., Golz, J.F., Giménez-Ibañez, S., Fernandez-Barbero, G., Franco-Zorrilla, J.M.,
Solano, R., 2015. FILAMENTOUS FLOWER is a direct target of JAZ3 and modulates
responses to jasmonate. Plant Cell 27, 1–15.
Bruni, I., De Mattia, F., Martellos, S., Galimberti, A., Savadori, P., Casiraghi, M., Nimis,
P.L., Labra, M., 2012. DNA barcoding as an effective tool in improving a digital plant
identification system: a case study for the area of Mt Valerio, Trieste (NE Italy). PLoS
One 7, e43256.
U.B. Andersen et al.
Charrier, G., Poirier, M., Bonhomme, M., Lacointe, A., Améglio, T., 2013. Frost hardiness
in walnut trees (Juglans regia L.): how to link physiology and modelling? Tree Physiol.
33, 1229–1241.
Cheng, L., Fuchigami, L.H., 2002. Growth of young apple trees in relation to reserve
nitrogen and carbohydrates. Tree Physiol. 22, 1297–1303.
Chung, H., Muraoka, H., Nakamura, M., Han, S., Muller, O., Son, Y., 2013. Experimental
warming studies on tree species and forest ecosystems: a literature review. J. Plant
Res. 126, 447–460.
Cleland, E.E., Chuine, I., Menzel, A., Mooney, H.A., Schwartz, M.D., 2007. Shifting plant
phenology in response to global change. Trends Ecol. Evol. 22, 357–365.
Cooke, J.E.K., Eriksson, M.E., Junttila, O., 2012. The dynamic nature of bud dormancy in
trees Environmental control and molecular mechanisms. Plant Cell Environ. 35,
1707–1728.
DMI, 2015. The Danish Meteorological Institute. http://www.dmi.dk.
De Barba, D., Rossi, S., Deslauriers, A., Morin, H., 2016. Effects of soil warming and
nitrogen foliar applications on bud burst of black spruce. Trees − Struct. Funct. 30,
87–97.
Derory, J., Léger, P., Garcia, V., Schaeffer, J., Hauser, M.T., Salin, F., Luschnig, C.,
Plomion, C., Glössl, J., Kremer, A., 2006. Transcriptome analysis of bud burst in
sessile oak (Quercus petraea). New Phytol. 170, 723–738.
Dethloff, F., Erban, A., Orf, I., Alpers, J., Fehrle, I., Beine-Golovchuk, O., Schmidt, S.,
Schwachtje, J., Kopka, J., 2014. Profiling methods to identify cold-regulated primary
metabolites using gas chromatography coupled to mass spectrometry. Methods Mol.
Biol. 1166, 171–197.
Dhuli, P., Rohloff, J., Strimbeck, G.R., 2014. Metabolite changes in conifer buds and
needles during forced bud break in Norway spruce (Picea abies) and European silver
fir (Abies alba). Front. Plant Sci. 5, 1–13.
Duchêne, E., Butterlin, G., Dumas, V., Merdinoglu, D., 2012. Towards the adaptation of
grapevine varieties to climate change: QTLs and candidate genes for developmental
stages. Theor. Appl. Genet. 124, 623–635.
Erban, A., Schauer, N., Fernie, A.R., Kopka, J., 2007. Nonsupervised construction and
application of mass spectral and retention time index libraries from time-of-flight gas
chromatography-mass spectrometry metabolite profiles. Methods Mol. Biol. 358,
19–38.
Falavigna, V. da S., Porto, D.D., Buffon, V., Margis-Pinheiro, M., Pasquali, G., Revers, L.F.,
2014. Differential transcriptional profiles of dormancy-related genes in apple buds.
Plant Mol. Biol. Rep. 32, 796–813.
Fennell, A., 2014. Genomics and functional genomics of winter low temperature tolerance
in temperate fruit crops. CRC Crit. Rev. Plant Sci. 33, 125–140.
Ferrer, J.-L., Austin, M.B., Stewart, C., Noel, J.P., 2008. Structure and function of enzymes
involved in the biosynthesis of phenylpropanoids. Plant Physiol. Biochem. 46,
356–370.
Fitter, A.H., Fitter, R.S.R., 2002. Rapid changes in flowering time in British plants. Science
296, 1689–1691.
Fliszkiewicz, M., Giejdasz, K., Wilkaniec, Z., 2011. The importance of male red mason bee
(Osmia rufa L.) and male bufftailed bumblebee (Bombus terrestris L.) pollination in
blackcurrant (Ribes nigrum L.). J. Hortic. Sci. Biotechnol. 86, 457–460.
Fu, Y.H., Campioli, M., Deckmyn, G., Janssens, I.A., 2012. The impact of winter and
spring temperatures on temperate tree budburst dates: results from an experimental
climate manipulation. PLoS One 7, e47324.
Greer, D.H., Wünsche, J.N., Norling, C.L., Wiggins, H.N., 2006. Root-zone temperatures
affect phenology of bud break, flower cluster development, shoot extension growth
and gas exchange of Braeburn (Malus domestica) apple trees. Tree Physiol. 26,
105–111.
Guy, C., Kaplan, F., Kopka, J., Selbig, J., Hincha, D.K., 2008. Metabolomics of tempera-
ture stress. Physiol. Plant. 132, 220–235.
Harrington, C.A., Gould, P.J., StClair, J.B., 2010. Modeling the effects of winter en-
vironment on dormancy release of Douglas-fir. For. Ecol. Manage. 259, 798–808.
Hedley, P.E., Russell, J.R., Jorgensen, L., Gordon, S., Morris, J.A., Hackett, C.A., Cardle,
L., Brennan, R., 2010. Candidate genes associated with bud dormancy release in
blackcurrant (Ribes nigrum L.). BMC Plant Biol. 10, 202.
Howe, G.T., Horvath, D.P., Dharmawardhana, P., Priest, H.D., Mockler, T.C., Strauss,
S.H., 2015. Extensive transcriptome changes during natural onset and release of
vegetative bud dormancy in Populus. Front. Plant Sci. 6, 1–28.
Hummel, J., Selbig, J., Walther, D., Kopka, J., 2007. The golm metabolome database: a
database for GC–MS based metabolite profiling. Top. Curr. Genet. 18, 75–95.
Ipcc, Press, C.U, 2013. Climate change 2013: the physical science basis. Clim. Change
2013, 1535.
Jaglo-Ottosen, K.R., Gilmour, S.J., Zarka, D.G., Schabenberger, O., Thomashow, M.F.,
1998. Arabidopsis CBF1 overexpression induces COR genes and enhances freezing
tolerance. Science 280, 104–106.
Jeong, S.J., Medvigy, D., Shevliakova, E., Malyshev, S., 2012. Uncertainties in terrestrial
carbon budgets related to spring phenology. J. Geophys. Res. Biogeosci. 117,
G01030.
Jones, H.G., Hillis, R.M., Gordon, S.L., Brennan, R.M., 2013. An approach to the de-
termination of winter chill requirements for different Ribes cultivars. Plant Biol. 15,
18–27.
Kavi Kishor, P.B., Hima Kumari, P., Sunita, M.S.L., Sreenivasulu, N., 2015. Role of proline
in cell wall synthesis and plant development and its implications in plant ontogeny.
Front. Plant Sci. 6, 544.
Kontunen-Soppela, S., Laine, K., 2001. Seasonal fluctuation of dehydrins is related to
osmotic status in Scots pine needles. Trees-Struct. Funct. 15, 425–430.
Kopka, J., Schauer, N., Krueger, S., Birkemeyer, C., Usadel, B., Bergmüller, E., Dörmann,
P., Weckwerth, W., Gibon, Y., Stitt, M., Willmitzer, L., Fernie, A.R., Steinhauser, D.,
2005. GMD@CSB.DB: the golm metabolome database. Bioinformatics 21,
1635–1638.
108
Environmental and Experimental Botany 140 (2017) 96–109
Ladwig, L.M., Ratajczak, Z.R., Ocheltree, T.W., Hafich, K.A., Churchill, A.C., Frey, S.J.K.,
Fuss, C.B., Kazanski, C.E., Mu??oz, J.D., Petrie, M.D., Reinmann, A.B., Smith, J.G.,
2016. Beyond arctic and alpine: the influence of winter climate on temperate eco-
systems. Ecology 97, 372–382.
Liu, Y., Mu, J., Niklas, K.J., Li, G., Sun, S., 2012. Global warming reduces plant re-
productive output for temperate multi-inflorescence species on the Tibetan plateau.
New Phytol. 195, 427–436.
Loewus, F.A., 1999. Biosynthesis and metabolism of ascorbic acid in plants and of analogs
of ascorbic acid in fungi. Phytochemistry 52, 193–210.
Luedeling, E., Girvetz, E.H., Semenov, M.A., Brown, P.H., 2011. Climate change affects
winter chill for temperate fruit and nut trees. PLoS One 6, e20155.
Luedemann, A., Strassburg, K., Erban, A., Kopka, J., 2008. TagFinder for the quantitative
analysis of gas chromatography–mass spectrometry (GC–MS)-based metabolite pro-
filing experiments. Bioinformatics 24, 732–737.
Lugassi, N., Nakayama, N., Bochnik, R., Zik, M., 2010. A novel allele of FILAMENTOUS
FLOWER reveals new insights on the link between inflorescence and floral meristem
organization and flower morphogenesis. BMC Plant Biol. 10, 131.
Lunn, J.E., Delorge, I., Figueroa, C.M., Van Dijck, P., Stitt, M., 2014. Trehalose metabo-
lism in plants. Plant J. 79, 544–567.
Maeda, H., Dudareva, N., 2012. The shikimate pathway and aromatic amino acid bio-
synthesis in plants. Annu. Rev. Plant Biol. 63, 73–105.
Maruyama, K., Takeda, M., Kidokoro, S., Yamada, K., Sakuma, Y., Urano, K., Fujita, M.,
Yoshiwara, K., Matsukura, S., Morishita, Y., Sasaki, R., Suzuki, H., Saito, K., Shibata,
D., Shinozaki, K., Yamaguchi-Shinozaki, K., 2009. Metabolic pathways involved in
cold acclimation identified by integrated analysis of metabolites and transcripts
regulated by DREB1A and DREB2A. Plant Physiol. 150, 1972–1980.
Mazzitelli, L., Hancock, R.D., Haupt, S., Walker, P.G., Pont, S.D.A., McNicol, J., Cardle, L.,
Morris, J., Viola, R., Brennan, R., Hedley, P.E., Taylor, M.A., 2007. Co-ordinated gene
expression during phases of dormancy release in raspberry (Rubus idaeus L.) buds. J.
Exp. Bot. 58, 1035–1045.
Mitchell, C., Brennan, R.M., Cross, J.V., Johnson, S.N., 2011. Arthropod pests of currant
and gooseberry crops in the U.K. Their biology, management and future prospects.
Agric. For. Entomol. 13, 221–237.
Mouillon, J.M., Aubert, S., Bourguignon, J., Gout, E., Douce, R., Rébeillé, F., 1999.
Glycine and serine catabolism in non-photosynthetic higher plant cells: their role in
C1 metabolism. Plant J. 20, 197–205.
Nomura, K., Ikegami, A., Koide, A., Yagi, F., 2007. Glutathione transferase, but not ag-
glutinin, is a dormancy-related protein in Castanea crenata trees. Plant Physiol.
Biochem. 45, 15–23 (d).
Ögren, E., 1996. Dehardening in Vaccinium myrtillus during a mild winter: a cause for
winter dieback? Funct. Ecol. 10, 724–732.
Ogren, E., Nilsson, T., Sundblad, L.-G., 1997. Relationship between respiratory depletion
of sugars and loss of cold hardiness in coniferous seedlings over-wintering at raised
temperatures: indications of different sensitivities of spruce and pine. Plant Cell
Environ. 20, 247–253.
Pagter, M., Lefèvre, I., Arora, R., Hausman, J.F., 2011. Quantitative and qualitative
changes in carbohydrates associated with spring deacclimation in contrasting
Hydrangea species. Environ. Exp. Bot. 72, 358–367.
Pagter, M., Sergeant, K., Møller, S.M., Bertram, H.C., Renaut, J., 2014. Changes in the
proteome and water state in bark and xylem of Hydrangea paniculata during loss of
freezing tolerance. Environ. Exp. Bot. 106, 99–111.
Pagter, M., Andersen, U.B., Andersen, L., 2015. Winter warming delays dormancy release,
advances budburst, alters carbohydrate metabolism and reduces yield in a temperate
shrub. AoB Plants 7 plv024–plv024.
Parada, F., Noriega, X., Dantas, D., Bressan-Smith, R., Pérez, F.J., 2016. Differences in
respiration between dormant and non-dormant buds suggest the involvement of ABA
in the development of endodormancy in grapevines. J. Plant Physiol. 201, 71–78.
Parmesan, C., 2006. Ecological and evolutionary responses to recent climate change.
Annu. Rev. Ecol. Evol. Syst. 37, 637–669.
Pedersen, H.L., 2008. Juice quality and yield capacity of black currant cultivars in
Denmark. Acta Hortic. 511–516.
Ponnu, J., Wahl, V., Schmid, M., 2011. Trehalose-6-phosphate: connecting plant meta-
bolism and development. Front. Plant Sci. 2, 70.
R Core Team (2015). R: A language and environment for statistical computing. R
Foundation for Statistical Computing, Vienna
Repo, T., Hänninen, H., Kellomäki, S., 1996. The effects of long-term elevation of air
temperature and CO2 on the frost hardiness of Scots pine. Plant Cell Environ. 19,
209–216.
Riikonen, J., Kontunen-Soppela, S., Vapaavuori, E., Tervahauta, A., Tuomainen, M.,
Oksanen, E., 2013. Carbohydrate concentrations and freezing stress resistance of
silver birch buds grown under elevated temperature and ozone. Tree Physiol. 33,
311–319.
Rohde, A., Bhalerao, R.P., 2007. Plant dormancy in the perennial context. Trends Plant
Sci. 12, 217–223.
Rohde, A., Ruttink, T., Hostyn, V., Sterck, L., Van Driessche, K., Boerjan, W., 2007. Gene
expression during the induction, maintenance, and release of dormancy in apical
buds of poplar. J. Exp. Bot. 58, 4047–4060.
Rose, G.A., Cameron, R.W., 2009. Chill unit models for blackcurrant (Ribes nigrum L.)
cultivars “Ben Gairn”: “Ben Hope” and “Ben Tirran”. Sci. Hortic. (Amsterdam). 122,
654–657.
Ruonala, R., Rinne, P.L.H., Baghour, M., Moritz, T., Tuominen, H., Kangasjärvi, J., 2006.
Transitions in the functioning of the shoot apical meristem in birch (Betula pendula)
involve ethylene. Plant J. 46, 628–640.
Ruttink, T., Arend, M., Morreel, K., Storme, V., Rombauts, S., Fromm, J., Bhalerao, R.P.,
Boerjan, W., Rohde, A., 2007. A molecular timetable for apical bud formation and
dormancy induction in poplar. Plant Cell 19, 2370–2390.
U.B. Andersen et al.
SØnsteby, A., Heide, O.M., 2011. Elevated autumn temperature promotes growth cessa-
tion and flower formation in black currant cultivars (Ribes nigrum L.). J. Hortic. Sci.
Biotechnol. 86, 120–127.
Sønsteby, A., Opstad, N., Heide, O.M., 2012. Effects of summer temperature on growth
and flowering in six black currant cultivars (Ribes nigrum L.). J. Hortic. Sci.
Biotechnol. 87, 157–164.
Saeed, A.I., Sharov, V., White, J., Li, J., Liang, W., Bhagabati, N., Braisted, J., Klapa, M.,
Currier, T., Thiagarajan, M., Sturn, A., Snuffin, M., Rezantsev, A., Popov, D., Ryltsov,
A., Kostukovich, E., Borisovsky, I., Liu, Z., Vinsavich, A., Trush, V., Quackenbush, J.,
2003. TM4: A free, open-source system for microarray data management and ana-
lysis. Biotechniques 34, 374–378.
Schulz, E., Tohge, T., Zuther, E., Fernie, A.R., Hincha, D.K., 2015. Natural variation in
flavonol and anthocyanin metabolism during cold acclimation in Arabidopsis thaliana
accessions. Plant Cell Environ. 38, 1658–1672.
Schulz, E., Tohge, T., Zuther, E., Fernie, A.R., Hincha, D.K., 2016. Flavonoids are de-
terminants of freezing tolerance and cold acclimation in Arabidopsis thaliana. Sci.
Rep. 6, 34027.
Sherson, S.M., Alford, H.L., Forbes, S.M., Wallace, G., Smith, S.M., 2003. Roles of cell-wall
invertases and monosaccharide transporters in the growth and development of
Arabidopsis. J. Exp. Bot. 525–531.
Shi, Y., Ding, Y., Yang, S., 2015. Cold signal transduction and its interplay with phyto-
hormones during cold acclimation. Plant Cell Physiol. 56, 7–15.
Shim, J.S., Imaizumi, T., 2015. Circadian clock and photoperiodic response in Arabidopsis:
From seasonal flowering to redox homeostasis. Biochemistry 54, 157–170.
Shin, H., Oh, Y., Kim, D., 2015. Differences in cold hardiness, carbohydrates, dehydrins
and related gene expressions under an experimental deacclimation and reacclimation
in Prunus persica. Physiol. Plant. 154, 485–499.
Slewinski, T.L., 2011. Diverse functional roles of monosaccharide transporters and their
homologs in vascular plants: a physiological perspective. Mol. Plant. 4, 641–6662.
Smirnoff, N., Wheeler, G.L., 2000. Ascorbic acid in plants: biosynthesis and function. Crit.
Rev. Biochem. Mol. Biol. 35, 291–314.
Sperling, O., Earles, J.M., Secchi, F., Godfrey, J., Zwieniecki, M.A., 2015. Frost induces
respiration and accelerates carbon depletion in trees. PLoS One 10, e0144124.
Stacklies, W., Redestig, H., Scholz, M., Walther, D., Selbig, J., 2007. pcaMethods − a
bioconductor package providing PCA methods for incomplete data. Bioinformatics
23, 1164–1167.
Taulavuori, K., Laine, K., Taulavuori, E., Pakonen, T., Saari, E., 1997. Accelerated de-
hardening in bilberry (Vaccinium myrtillus L.) induced by a small elevation in air
temperature. Environ. Pollut. 98, 91–95.
109
Environmental and Experimental Botany 140 (2017) 96–109
Taulavuori, K.M.J., Taulavuori, E.B., Skre, O., Nilsen, J., Igeland, B., Laine, K.M., 2004.
Dehardening of mountain birch (Betula pubescens ssp. czerepanovii) ecotypes at ele-
vated winter temperatures. New Phytol. 162, 427–436.
Therneau, T., 2012. A Package for Survival Analysis in S. R package version, Survival
(Lond).
Tinklin, I.G., Schwabe, W.W., 1970. Lateral bud dormancy in the blackcurrant Ribes ni-
grum. Ann. Bot. 34, 691–706.
Ueno, S., Klopp, C., Leplé, J.C., Derory, J., Noirot, C., Léger, V., Prince, E., Kremer, A.,
Plomion, C., Le Provost, G., 2013. Transcriptional profiling of bud dormancy in-
duction and release in oak by next-generation sequencing. BMC Genomics 14, 236.
Vagiri, M., Ekholm, A., Andersson, S.C., Johansson, E., Rumpunen, K., 2012. An opti-
mized method for analysis of phenolic compounds in buds, leaves, and fruits of black
currant (Ribes nigrum l.). J. Agric. Food Chem. 60, 10501–10510.
Welling, A., Palva, E.T., 2006. Molecular control of cold acclimation in trees. Physiol.
Plant. 127, 167–181.
Welling, A., Palva, E.T., 2008. Involvement of CBF transcription factors in winter
hardiness in birch. Plant Physiol. 147, 1199–1211.
Wisniewski, M., Nassuth, A., Teulières, C., Marque, C., Rowland, J., Cao, P.B., Brown, A.,
2014. Genomics of cold hardiness in woody plants. CRC Crit. Rev. Plant Sci. 33,
92–124.
Xia, J., Chen, J., Piao, S., Ciais, P., Luo, Y., Wan, S., 2014. Terrestrial carbon cycle affected
by non-uniform climate warming. Nat. Geosci. 7, 173–180.
Xie, D.W., Wang, X.N., Fu, L.S., Sun, J., Zheng, W., Li, Z.F., 2015. Identification of the
trehalose-6-phosphate synthase gene family in winter wheat and expression analysis
under conditions of freezing stress. J. Genet. 94, 55–65.
Xu, H., Cao, D., Chen, Y., Wei, D., Wang, Y., Stevenson, R.A., Zhu, Y., Lin, J., 2016. Gene
expression and proteomic analysis of shoot apical meristem transition from dormancy
to activation in Cunninghamia lanceolata (Lamb.) Hook. Sci. Rep. 6, 19938.
Zhang, J.Z., Ai, X.Y., Sun, L.M., Zhang, D.L., Guo, W.W., Deng, X.X., Hu, C.G., 2011.
Molecular cloning and functional characterization of genes associated with flowering
in citrus using an early-flowering trifoliate orange (Poncirus trifoliata L. Raf.) mutant.
Plant Mol. Biol. 76, 187–204.
Zhong, W., Gao, Z., Zhuang, W., Shi, T., Zhang, Z., Ni, Z., 2013. Genome-wide expression
profiles of seasonal bud dormancy at four critical stages in Japanese apricot. Plant
Mol. Biol. 83, 247–264.
Zuther, E., Juszczak, I., Ping Lee, Y., Baier, M., Hincha, D.K., 2015. Time-dependent
deacclimation after cold acclimation in Arabidopsis thaliana accessions. Sci. Rep. 5,
12199.