autophagy is responsible for the accumulation of proteogenic ......2020/03/25 · dipeptide levels...

Dipeptide levels are dependent on autophagy

1 Thirumalaikumar et al.

Autophagy is responsible for the accumulation of proteogenic dipeptides in 1

response to heat stress in Arabidopsis thaliana 2

3

Venkatesh P. Thirumalaikumar1, Mateusz Wagner1,3, Salma Balazadeh1,2, 4

Aleksandra Skirycz1* 5 6 1 Max Planck Institute of Molecular Plant Physiology, 14476 Potsdam, Germany 7 2 University of Leiden, 2333 Leiden, The Netherlands 8 3 University of Wroclaw, 50-383, Wroclaw, Poland 9

10 * Corresponding author; [email protected] 11

12

ORCID IDs: 13

Salma Balazadeh: 0000-0002-5789-4071 14 Venkatesh P. Thirumalaikumar: 0000-0002-5009-1460 15 Aleksandra Skirycz: 0000-0002-7627-7925 16 17

Running title: Dipeptide levels are dependent on autophagy 18 19 20 Keywords: autophagy, dipeptides, heat stress 21

22

Abbreviations: LC-MS, liquid chromatography mass spectrometry; GAPDH – 23

glyceraldehyde-3-phosphate dehydrogenase 24

25

26

27

28

29

30

31

32

33

34

35

36

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 25, 2020. ; https://doi.org/10.1101/2020.03.25.004978doi: bioRxiv preprint

https://doi.org/10.1101/2020.03.25.004978



Abstract 1

Proteogenic dipeptides are intermediates of proteolysis as well as an emerging class 2

of small-molecule regulators with diverse and often dipeptide-specific functions. 3

Herein, prompted by differential accumulation of dipeptides in a high-density 4

Arabidopsis thaliana time-course stress experiment, we decided to pursue an identity 5

of the proteolytic pathway responsible for the buildup of dipeptides under heat 6

conditions. By querying dipeptide accumulation versus available transcript data, 7

autophagy emerged as a top hit. To examine whether autophagy indeed contributes 8

to the accumulation of dipeptides measured in response to heat stress, we 9

characterized the loss-of-function mutants of crucial autophagy proteins to test 10

whether interfering with autophagy would affect dipeptide accumulation in response 11

to the heat treatment. This was indeed the case. This work implicates the 12

involvement of autophagy in the accumulation of proteogenic dipeptides in response 13

to heat stress in Arabidopsis. 14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30


https://doi.org/10.1101/2020.03.25.004978



Introduction 1

Chemically speaking, dipeptides are a pair of amino acids joined by a single peptide 2

bond. Considering only the combination of the 20 proteogenic amino acids, 3

discussed here, the total number of dipeptides is 400. In the past, dipeptides have 4

been exclusively considered as short-lived intermediates on their way to specific 5

amino acid degradation pathways following further proteolysis. However, this view 6

has been challenged by recent studies reporting regulatory roles for various 7

proteogenic dipeptides across different organisms. 8

A compelling example comes from the regulation of the yeast (Saccharomyces 9

cerevisiae) N-end rule pathway. UBR1 (Ubiquitin Protein Ligase E3 Component N-10

Recognin 1) is an ubiquitin ligase that has three ligand binding sites, of which two are 11

specific for dipeptides. The consecutive binding of a dipeptide with the basic N-12

terminal residue (Arg, Lys, or His) to site-1 of UBR1 and of a dipeptide with the bulky 13

hydrophobic N-terminal residue to site-2 (Trp, Phe, Tyr, Leu, or Ile) enables binding 14

and ubiquitylation of the CUP9 protein, a transcriptional repressor of peptide import 15

[1]. Other regulatory dipeptides include Tyr-Arg, also known as kyotorphin, a 16

neurotransmitter, and a potent analgesic [2]. Interestingly, kyotorphin is produced in 17

cells via active synthesis mediated by the enzyme kyotorphin synthase [3]. A 18

different Tyr-containing dipeptide, Tyr-Leu, was demonstrated to display anxiolytic 19

(i.e., stress-reducing) activity in mice, its potency being comparable to that of valium 20

[4]. Notably, both the retro-sequence peptide Leu-Tyr and the mixture of tyrosine and 21

leucine were inactive, demonstrating that even chemically similar dipeptides can 22

have specific bioactivities. Moreover, the lack of anxiolytic activity reported for the 23

amino acids supported the notion that the dipeptide function would be independent of 24

the degradation to the constituent amino acids. In other words, when it comes to 25

function, dipeptides are more than a sum of the constituent amino acids [4]. 26

Recently, to identify potential regulatory small molecules on account of their presence 27

in protein complexes, we developed a simple approach based on co-fractionation 28

that exploits a significant size difference between protein�metabolite complexes and 29

free metabolites [5] [6]. Remarkably, among the small molecules co-fractionating 30

with proteins and protein complexes isolated from a native Arabidopsis thaliana 31

lysate, dipeptides represented a large and diverse group. Specifically, of the 237 32

detected proteogenic dipeptides, 106 were classified as protein-bound; dipeptide 33


https://doi.org/10.1101/2020.03.25.004978



elution profiles span the whole protein separation range, indicating the presence of 1

multiple protein-binding partners. By focusing on one selected dipeptide, namely, 2

Tyr-Asp, we could narrow down the list of Tyr-Asp interactors to the glycolytic 3

enzyme glyceraldehyde-3-phosphate dehydrogenase (GAPDH). By testing multiple 4

dipeptides and multiple GAPDHs, we demonstrated that the binding is specific to Tyr-5

Asp for both plant and mammalian GAPDHs [6]. In addition to its enzymatic function, 6

GAPDH plays additional roles in non-metabolic processes such as the control of 7

gene expression and actin-bundling [7]. However, the exact biological function of the 8

Tyr-Asp–GAPDH interaction remains unclear; specific binding of the Tyr-Asp to 9

GAPDH provides supporting evidence to the previously reported regulatory roles of 10

the proteogenic dipeptides. 11

Intrigued by their presence in the Arabidopsis protein complexes, we decided to 12

pursue the functions of the diverse proteogenic dipeptides. One immediate and 13

essential question concerns dipeptides' biogenesis. Proteolysis is considered the 14

main source of the proteogenic dipeptides. Plant genomes encode multiple 15

proteases with assigned exopeptidase activity. Exopeptidases would cleave amino 16

acids but also dipeptides from either the C- (carboxypeptidases) or N- 17

(aminopeptidases) termini of the protein substrates. Notably, known dipeptidyl 18

exopeptidases are characterized by cleavage specificity, which could result in 19

specific accumulation patterns. For instance, animal dipeptidyl peptidase-4 serine 20

exopeptidase would cleave X-proline or X-alanine dipeptides from the N-terminus of 21

peptides exclusively [8]. Dipeptides are subsequently cleaved to single amino acids 22

by enzymes with dipeptidase activity. Similar to exopeptidases, dipeptidases also 23

display substrate specificity [9]. 24

Herein, prompted by the differential accumulation of proteogenic dipeptides in a high-25

density Arabidopsis time-course stress experiment [10,11], which we re-annotated 26

using a dipeptide library, we decided to pursue an identity of the proteolytic pathway 27

responsible for the increase of dipeptides measured under heat conditions. Co-28

expression analysis identified autophagy as a top hit, on par with previous reports 29

demonstrating the importance of autophagy for plant response to stress [12]. During 30

autophagy, cytoplasmic components including proteins are captured into a double-31

membraned structure, namely, the autophagosome, and delivered to the vacuole for 32

recycling. The plant vacuolar proteome comprises multiple proteases, which 33


https://doi.org/10.1101/2020.03.25.004978



subsequently cleave protein cargos to amino acids, which are thereafter transported 1

back to the cytosol by amino acid transporters [13]. Notably, in the context of our 2

findings, the vacuolar proteome is also characterized by the presence of dipeptidyl 3

exopeptidases and dipeptide transporters [13]. In Arabidopsis, the role of autophagy 4

has been studied extensively, benefiting from the multiple loss-of-function mutants as 5

well as small-molecule inhibitors, which interfere with autophagosome formation, 6

protein cargo recruitment, or vacuolar function [14]. Following accepted strategies, 7

we used four autophagy mutants, to test whether disrupting autophagy would affect 8

dipeptide accumulation in response to the heat treatment. Indeed, this was the case. 9

In summary, we demonstrate that autophagy is involved in the accumulation of 10

proteogenic dipeptides in response to heat stress. 11

12

Results 13

The levels of dipeptides’ response to varying light and temperature conditions 14

To gain insight into dipeptide biogenesis and function, we investigated dipeptide 15

accumulation in the previously published Arabidopsis stress time-course experiment 16

[10]. The experiment comprised eight different environmental conditions varying in 17

temperature and light, and 22 time-points. All samples were subjected to 18

transcriptomics [10] and untargeted metabolomics analysis [11]. Herein, we retrieved 19

an existing metabolomics dataset [11] to annotate dipeptides by matching metabolic 20

features, previously assigned as unknown, with a dipeptide reference compound 21

library. 22

In total, we annotated 54 dipeptides, which were clustered into six distinctive groups 23

(Figure 1; Supplemental Table S1). Dipeptides in clusters 1 to 3 accumulated 24

under high light conditions. Cluster 4 was mainly composed (16 of 21) of aspartic 25

acid or glutamic acid-containing dipeptides (referred to later as acidic dipeptides) that 26

responded rapidly to multiple stress conditions, with the levels increasing under heat 27

and dark conditions but decreasing in the cold treatment. Dipeptides in clusters 5 28

and 6 were characterized by the presence of branch-chain amino acids (valine, 29

leucine, isoleucine) and accumulated in all tested conditions except for high light, but 30

the increase was moderate and measured 6 h or longer after the stress onset. 31


https://doi.org/10.1101/2020.03.25.004978



Based on the clearly defined accumulation pattern, we decided to focus on the 1

dipeptides of cluster 4. 2

In summary, dipeptide levels respond to changes in environmental conditions. The 3

exact response patterns could be linked to dipeptide amino acid composition. 4

5

Level of acidic dipeptides correlates with autophagy-related transcripts across 6

the stress time-course 7

To look for putative protein candidates responsible for the accumulation of acidic 8

dipeptides, we correlated dipeptide levels (average of all dipeptides in cluster 4, see 9

above and Figure 1) with the available transcript data from the same stress time-10

course [10]. We selected the 100 top correlated transcripts (referred to as the 11

´positive list´) and looked at the functionality of the encoded proteins using the 12

STRING database [15]. Remarkably, the positive list was enriched with transcripts 13

encoding proteins associated with autophagy and proteolysis (Figure 2A-B; 14

Supplementary Table S2). This included autophagy-associated proteins 8 (ATG8C, 15

ATG8H, and ATG8F), 12 (ATG12A), 13 (ATG13) and 18 (ATG18G), multiple 16

ubiquitin-protein ligases, cathepsin B-like cysteine protease (AtCathB2), serine 17

carboxypeptidase S28, cysteine-type peptidase RD21, and aspartyl aminopeptidase 18

(AT5G60160). Interestingly, three of the four proteases from the positive list are 19

located in the vacuole [13], which is also the final destination of proteins captured by 20

autophagosomes. Moreover, AtCathB2 has an experimentally tested dipeptidyl 21

peptidase activity [16]. 22

In summary, we showed that acidic dipeptides accumulate under conditions of 23

darkness and heat stress (see Figure 1), both of which are known to induce 24

autophagy and proteolysis, and on par dipeptide levels are tightly correlated with the 25

levels of transcripts encoding associated autophagy proteins, ubiquitin-protein 26

ligases, and vacuolar proteases including dipeptidyl peptidases. 27

28

29

30


https://doi.org/10.1101/2020.03.25.004978



Dipeptide accumulation during recovery from heat stress depends on 1

autophagy 2

To examine whether the accumulation of dipeptide characteristics for dark and heat 3

stress conditions (see above) depends on autophagy, we next employed a genetic 4

approach. To this end, two different single (atg18 and nbr1-2) and two double 5

(atg4(4a/4b), nbr1-2/atg5) mutant lines defective in core and selective autophagy 6

were grown on media plates, and 5-day-old seedlings were subjected to a two-step 7

heat treatment established by [17] (see Materials and Methods). Samples were 8

harvested two days after the heat treatment (within the recovery phase). Our 9

experimental design was based on recent findings, which demonstrated a peak of 10

autophagy at 2-3 days following heat stress treatment [17]. Collected seedlings were 11

used for untargeted LC-MS-based metabolomics analysis, and a total of 365 12

metabolic features comprising dipeptides, primary and secondary metabolites, and 13

lipids were putatively annotated to a compound using an in-house reference 14

compound library [18]. 15

Statistical analysis of the 57 measured dipeptides revealed striking differences 16

between wild-type and mutant lines of Arabidopsis. While under control conditions, 17

dipeptides were overall more abundant in all the four mutant lines when compared to 18

the wild type; the opposite was observed in the heat stress recovery samples. That 19

is because heat induced a significant increase in the dipeptide levels in the wild type, 20

but no such increase was measured in the autophagy mutants. In contrast to the 21

stress time-course experiment, 52 of the measured dipeptides responded in a similar 22

fashion (Figure 3 and Supplemental Figure S1). There were, however, five 23

exceptions: Asp-Thr, Glu-Glu, Glu-Gln, Ala-Thr, and Val-Lys (Supplemental Figure 24

S1 and Figure S2). Whilst levels of both Asp-Thr and Val-Lys were significantly 25

(Student’s t-test p-value <0.05) decreased by stress in the wild-type plants, Glu-Glu 26

was not affected, and the levels of Glu-Gln and Ala-Thr significantly increased in both 27

wild-type and mutant plants (Supplemental Figure S2). 28

In conclusion, dipeptide accumulation measured during heat stress recovery 29

depends on autophagy. 30

31

32


https://doi.org/10.1101/2020.03.25.004978



Metabolite profiling of autophagy-deficient mutants reveals significant 1

metabolome alterations 2

The canonical role of autophagy is the recycling of cellular constituents, including 3

nutrients. Thus, not surprisingly, autophagy-deficient mutants are characterized by 4

significant alterations in the metabolome under both control and stress conditions 5

[12,19,20]. Changes in the accumulation of amino acids, sugars, lipids, and 6

secondary metabolites have been reported and were shown to be strongly 7

dependent on the plant's developmental stage and growth conditions. For instance, 8

etiolated seedlings of the autophagy-deficient mutant atg5 are characterized by 9

reduced levels of amino acids and glucosinolates, while the levels of triacylglycerols, 10

fatty acids, and sugars increased [12]. By contrast, mature rosette leaves of the atg5 11

mutant were shown to accumulate more amino acids but fewer sugars and 12

anthocyanins under both control and nitrate-limiting conditions [19]. 13

Herein, we present primary and secondary metabolites and lipid data for Arabidopsis 14

seedlings grown under control conditions and subjected to heat stress, followed by a 15

recovery (Supplemental Table S3). Notably, all of the four tested autophagy 16

mutants were characterized by significant alterations in their metabolome, already 17

under control conditions (Figure 4, Supplemental Figure S3). Amino acids and 18

flavonols were overall lower in the mutants, irrespective of the treatment, whereas 19

phospholipids, lysophospholipids, nucleotides, and cofactors NADH, NADP+, and 20

FAD+ were higher, but only under control conditions. While heat stress induced 21

nucleotide accumulation in the wild-type plants, no such accumulation was measured 22

in the autophagy mutants, consistent with the involvement of autophagy in nucleic 23

acid degradation. Another interesting observation relates to the difference in the 24

accumulation of triacylglycerols (TAGs) in nbr1 plants in comparison to the wild type 25

and the other autophagy mutants. While TAGs 42-50 (the number relates to the total 26

carbon length of fatty acid chains) were overall lower in all mutants in comparison to 27

the wild type, the much more abundant TAGs 50-60 were significantly decreased in 28

the nbr1 mutant only. In contrast to the atg4, atg18, and atg5 mutants, nbr1 is 29

specifically defective in selective autophagy (Supplemental Figure S4). The 30

differential accumulation of TAGs in nbr1 plants is therefore an exciting finding for a 31

follow-up analysis. 32


https://doi.org/10.1101/2020.03.25.004978



In summary, the results presented demonstrate the essential role of autophagy for 1

the regulation of plant metabolism. 2

3

Discussion 4

5

Dipeptide accumulation depends on autophagy 6

The prime role of autophagy is the recycling of macromolecules. Proteins captured 7

by autophagosomes are delivered to the vacuoles, where they are cleaved to amino 8

acids, which in turn get transported back to the cytosol by amino acid transporters 9

situated in a vacuolar membrane. Herein, we demonstrate that autophagy, in 10

addition to amino acid recycling, is also responsible for the increased accumulation of 11

dipeptides in recovery from heat stress (Figures 2-3). In support of our results, 12

vacuolar peptidases with dipeptidyl peptidase activity, e.g., AtCathB2; At1g02305 13

and serine carboxypeptidase S28, At5g60360 [13,16], as well as vacuolar dipeptide 14

transporters, e.g., PTR2; AT2G02040 [13,21], have been reported. 15

Dipeptides can be further cleaved to single amino acids, and as such, represent the 16

intermediates of proteolysis. However, proteogenic dipeptides have also been 17

reported to function as small-molecule regulators. In addition to the already given 18

examples (see Introduction), dipeptides were shown, for instance, to activate the 19

p38MAPK-Smad3 signaling pathway in the rare chronic myelogenous leukemia 20

(CML) stem cell population, and in line pharmacological inhibition of dipeptide uptake 21

inhibited CML stem cell activity [22]. Dipeptides were also reproducibly measured in 22

the Arabidopsis root exudates, downstream of MPK3/6 signaling, suggesting a role in 23

plant-microbe and plant-plant communication [23]. An inhibitory effect of five different 24

dipeptides on maize root growth was reported previously, supporting such a role [24], 25

while phenylalanine-containing dipeptides accumulate in wheat upon treatment with 26

deoxynivalenol, the major toxin and virulence factor of the fungal pathogen Fusarium 27

graminearum [25]. 28

Interestingly, in the context of our findings, a single dipeptide Tyr-Ala, but neither 29

tyrosine nor alanine, was shown to significantly enhance the lifespan and healthspan 30

of the model nematode Caenorhabditis elegans [26]. Tyr-Ala activity was associated 31

with the upregulation of stress-related proteins, ROS (reactive oxygen species) 32

scavenging enzymes, and chaperones. In agreement, Tyr-Ala effects were 33


https://doi.org/10.1101/2020.03.25.004978



particularly pronounced under heat and oxidative stress conditions, reminiscent of the 1

dipeptide accumulation reported in our study. Moreover, we previously demonstrated 2

that Tyr-Asp binds to a glycolytic enzyme GAPDHs [6], and inhibition of GAPDH 3

activity shifts glycolytic flux towards the pentose phosphate pathway and production 4

of NADPH, the primary source of reducing power for biosynthesis and detoxification 5

of ROS during stress [27]. Moreover, dipeptides were shown to promote protein 6

folding, which is important under stress conditions [28]. Based on the presented 7

results and on literature knowledge, we speculate that the dipeptide accumulation 8

observed under heat conditions, and related to autophagy, serves to alleviate 9

oxidative stress and protein damage. 10

11

Note on the specificity of dipeptide accumulation under different environmental 12

regimes 13

An important question about dipeptide biogenesis is specificity. Analysis of the stress 14

time-course experiment revealed that dipeptide levels are stress-responsive, but 15

what’s more intriguing is that not all the dipeptides respond in the same manner 16

(Figure 1). Amino acid composition, such as the presence of an acidic or branch-17

chain amino acid, defined groups of dipeptides sharing similar stress behavior, but 18

the different groups behaved differently. On par with our results, a recent 19

metabolomics analysis of different root cell types revealed preferential accumulation 20

of Leu and Ile containing dipeptides in the endodermis and epidermis [29]. By 21

contrast, heat treatment of the Arabidopsis seedlings (Figure 3, referred to as 22

seedling experiment) resulted in the bulk accumulation of dipeptides, with no 23

apparent specificity. Of the 57 dipeptides, 52 were characterized by similar 24

accumulation patterns. The five remaining dipeptides could not be easily clustered 25

based on either their amino acid composition or response pattern. 26

However, both the time-course and seedling experiments exploited heat stress; the 27

regimes were very different. While the seedling experiment is representative of a 28

severe stress regime resulting from a combination of in vitro growth conditions, young 29

plants, and severe heat treatment (37 °C and 44 °C), the time-course experiment 30

was performed using soil-grown, mature plants subjected to prolonged but mild heat 31

conditions (32 °C), which do not threaten plant survival. Moreover, in the seedling 32

experiment, rather than looking at heat stress per se, samples were harvested in the 33

recovery-phase after the heat stress. However, it was demonstrated that heat 34


https://doi.org/10.1101/2020.03.25.004978



activates autophagy and increases the number of autophagosomes and autophagic 1

bodies during the recovery phase [17]. Recycling of the stress-associated proteins, 2

such as heat shock proteins, is considered an essential component of the stress-3

resetting mechanism. Bulk accumulation patterns that are characteristic for the 4

seedling experiment could thus be explained by the rapid turnover of multiple stress 5

proteins without the preference for particular dipeptide motifs. By contrast, prolonged 6

and mild heat stress would favor more selective recycling, leading to the specificity in 7

dipeptide accumulation. This is a hypothesis that needs to be addressed in future 8

studies. 9

10

Materials and Methods 11

Dipeptide and transcript data: time-course stress experiment 12

Time-course experiment [10]: Briefly, mature Arabidopsis plants grown in soil at 21 13

°C and in standard light (150 µE m-2 s-1) were either kept at this condition or 14

transferred into seven different environments varying in temperature and/or light 15

intensity. More specifically, the seven conditions were as follows; low light (75 µE m-2 16

s-1), darkness, high light (300 µE m-2 s-1), heat (32 °C, 150 µE m-2 s-1), cold (4° C, 85 17

µE m-2 s-1), cold and darkness, and heat and darkness. Samples were harvested at 18

22 different time points ranging from 5 minutes to approximately 21 hours after the 19

transfer. Raw metabolite LC-MS data from the time-course stress experiment were 20

retrieved from [11]. We used the retrieved dataset to annotate dipeptides by 21

matching metabolic features, previously assigned as unknown, with a dipeptide 22

reference compound library. Data were subjected to day-normalization and sample-23

median-normalization. Dipeptide annotation was made, allowing 0.15 RT and 5 ppm 24

deviation from the dipeptide reference compound library comprising 331 dipeptides. 25

Transcript data used for co-expression analysis were retrieved from [10]. 26

Plant materials and growth conditions 27

Seeds of A. thaliana ecotype Col-0 and the mutant or transgenic lines nbr1-2 28

[30,31], nbr1-2/atg5 [30,31], (atg4(4a/4b) [32], and atg18a-2 [33] were surface-29

sterilized and sown in Petri dishes containing Murashige-Skoog (MS) agar medium 30

supplemented with 1% (w/v) sucrose. They were stratified at 4 °C in darkness for 31


https://doi.org/10.1101/2020.03.25.004978



2 days, and seedlings were grown in growth chambers providing 16 h light/8 h dark 1

cycles at 22 °C. 2

3

Heat stress experiment 4

Heat stress (also referred to as priming) was applied in two steps [17]. Five-day-5

old seedlings were subjected to an incubation (with an incubator) at 37 °C for 1.5 6

h, then at 22 °C for a 1.5�h recovery period, followed by 45�min of heat stress at 7

44� °C (in a hot water bath). After the treatment, the seedlings were transferred to 8

normal growth conditions (16 h light/8 h dark cycles) at 22 °C for 2 days (recovery 9

phase), during which samples (whole seedlings) were harvested for analyses. 10

11

Metabolite extraction 12

The extraction protocol was adapted and modified from [18]. Proteins and 13

metabolites were extracted using a methyl-tert-butyl ether (MTBE)/methanol/water 14

solvent system, which separates molecules into pellets (proteins), organics (lipids), 15

and an aqueous phase (primary and secondary metabolites). Equal volumes of the 16

polar and lipid fractions were dried in a centrifugal evaporator and stored at –80 °C 17

until further processing. 18

LC-MS metabolomics 19

The dried aqueous phase was measured using ultra-performance liquid 20

chromatography coupled to an Exactive mass spectrometer (Thermo Fisher 21

Scientific) in positive and negative ionization modes, as described in [18]. A 2-μl 22

sample (the dried aqueous fraction was resuspended in 200 μl of UHPLC-grade 23

water) was loaded per injection. Similarly, the dried organic phase was measured 24

using ultra-performance liquid chromatography coupled to an Exactive mass 25

spectrometer (Thermo Fisher Scientific) in positive and negative ionization modes, as 26

described in [18]. A 2-μl sample (the dried organic fraction was resuspended in 200 27

μl of UPLC-grade 7:3 acetonitrile:isopropanol) was loaded per injection. 28

Data pre-processing: LC-MS metabolite data 29

The LC-MS data were processed using Expressionist Refiner MS 11.0 (Genedata 30

AG, Basel, Switzerland). Settings were as follows: chromatogram alignment (RT 31


https://doi.org/10.1101/2020.03.25.004978



search interval, 0.5 min), peak detection (summation window, 0.09 min; minimum 1

peak size, 0.03 min; gap/peak ratio, 50%; smoothing window, 5 points; center 2

computation by intensity-weighted method with threshold at 70%; boundary 3

determination using inflection points), isotope clustering (RT tolerance at 0.015 min, 4

m/z tolerance 5 ppm, allowed charges 1–5), filtering for a single peak not assigned to 5

an isotope cluster, adduct detection, and cluster grouping (RT tolerance 0.05 min, 6

m/z tolerance 5 ppm, maximum intensity of side adduct 100000%). All metabolite 7

clusters were matched to in-house libraries of reference compounds, allowing a 10-8

ppm mass and dynamic retention time deviation (maximum 0.1 min). All annotations 9

are putative. 10

Data analysis 11

Raw metabolites’ intensities were normalized to the median of chromatogram 12

intensity in order to correct for variation in sample extraction and measurement. Prior 13

statistical analysis data were subjected to log2 transformation. P-value was 14

calculated using unpaired Student’s t-test with two-tailed distribution; n=4 of 15

biological replicates. Heat stress experiment was replicated with individual plates 16

grown in one experiment. The data are presented in Supplemental Tables S3. 17

Figure Legends 18

Figure 1. Heat map of dipeptide accumulation in the stress time series 19

experiment. Data are expressed as fold change calculated using paired control 20

treatment (21 °C and 150 μmol m−2s−1). Blue stands for down and yellow for 21

upregulation. The figures 5 to 1280 refer to the first and last sampling time points 22

given in minutes. Data were clustered using the Self Organizing Map (SOM) method 23

(standard settings) implemented in the Multiple Experiment Viewer (MeV) data 24

analysis application. Numbers 1 to 6 refer to clusters 1 to 6. For the complete data, 25

see Supplemental Table S1. 26

Figure 2. Transcripts co-related with acidic dipeptides in the stress time-27

course experiment. (A) Proteins encoded by the 100 top positively co-related 28

transcripts were queried against the STRING protein-protein interaction database 29

[15] and visualized using Cytoscape [34]. The edges represent medium confidence 30


https://doi.org/10.1101/2020.03.25.004978



interactions (SCORE>0.4) based on experimental, database, and literature evidence. 1

Red shading indicates proteins associated with autophagy and green shading with 2

protein ubiquitination. Unconnected nodes are not displayed. (B) Accumulation of 3

the acidic dipeptides (average of all dipeptides in cluster 4) and two representative 4

transcripts from the positive list of the time-course stress experiment. Data are 5

expressed as fold changes calculated in comparison to the control treatment (paired 6

control). For the complete data, see Supplemental Table S2. 7

Figure 3. Dipeptide accumulation measured during heat stress recovery 8

depends on the autophagy. Volcano plot representation of the log2 fold changes 9

(axis x) in dipeptide levels in the function of –log10 of p-value (axis y). P-value was 10

calculated using unpaired Student’s t-test with two-tailed distribution assuming 11

unequal variance; n=4 of biological replicates. (A) Between wild-type (WT) and 12

autophagy-deficient mutants grown under control conditions. (B) Between wild-type 13

(WT) and autophagy-deficient mutants measured two days into recovery from heat 14

treatment. (C) Between heat and control samples for wild-type (WT) and autophagy-15

deficient mutants. (D) The median of all dipeptides was used to calculate the 16

average and standard deviation (SD) across mutants and treatments. Note that data 17

for individual dipeptides used to calculate the median were expressed as log2 of 18

normalized LC-MS intensities. (A-C) Horizontal dash line indicates p-value of 0.05. 19

For the complete data, see Supplemental Table S3. 20

Figure 4. Metabolomic alterations in autophagy-deficient mutants. The heat 21

map represents the significant (P-value < 0.05) log2 fold changes of the selected 22

metabolites (for the full dataset, see Supplemental Table S3). P-value was 23

calculated using unpaired Student’s t-test with two-tailed distribution assuming 24

unequal variance; n=4 biological replicates. 25

26

Supplemental figure legends 27

Supplemental Figure S1. Heat map of log2 fold change of dipeptides measured in 28

heat recovery in comparison to control conditions. The data for all samples are 29

given. Clustering was performed using ClustVis tool (https://biit.cs.ut.ee/clustvis/) 30

and default settings. Please see Supplemental Table S3. 31


https://doi.org/10.1101/2020.03.25.004978



Supplemental Figure S2. The median of the fold change of all dipeptides measured 1

in the heat versus control conditions, separately in the wild type and the mutants, was 2

used to calculate Pearson correlation with all the dipeptides individually. For five of 3

the measured dipeptides (indicated by an orange shading), the Pearson correlation 4

was below 0.5, indicating a difference in response pattern. 5

Supplemental Figure S3. PCA plot prepared using all annotated metabolites in wild 6

type (WT) and mutants grown under control conditions using ClustVis tool 7

(https://biit.cs.ut.ee/clustvis/) and default settings. Please see Supplemental Table 8

S3. 9

Supplemental Figure S4. TAG accumulation in the wild type and in the nbr1 10

mutant measured under control conditions. Data are expressed as mean of the 11

median normalized intensity. P-value was calculated using unpaired Student’s t-test 12

with two-tailed distribution assuming unequal variance; n=4 of biological replicates. 13

Asterisks indicate significance (P-value < 0.05). 14

15

Supplemental Tables 16

Supplementary Table S1. Dipeptide data for Figure 1. The fold change in 17

comparison to the paired control is given. D, darkness; 4, cold; 30, heat; LL, low 18

light; HL, high light. The time is given in minutes. For example, 4D5: cold and dark 19

stress, five minutes after the stress onset. 20

Supplementary Table S2. Transcript data for Figure 2. The fold change in 21

comparison to the paired control is given. D, darkness; 4, cold; 30, heat; LL, low 22

light; HL, high light. The time is given in minutes. For example, 4D5: cold and dark 23

stress, five minutes after the stress onset. Pearson correlation (column D) was 24

calculated using the average response of dipeptides from cluster 4 (refer 25

Supplementary Table S1). 26

Supplementary Table S3. Normalized and log2 transformed metabolite data for 27

Figure 3 and Figure 4. Raw metabolite intensities were normalized to the median of 28

chromatogram intensity in order to correct for variation in sample extraction and 29

measurement. Prior statistical analysis data were subjected to log2 transformation. 30

31


https://doi.org/10.1101/2020.03.25.004978



1

Acknowledgements 2

We thank Anne Michaelis for running the metabolomics samples and Prof. Alisdair 3

R. Fernie for his valuable scientific input. Salma Balazadeh thanks the Deutsche 4

Forschungsgemeinschaft for the funding (CRC 973). 5

6

Author Contributions 7

A.S. and V.P.K. conceived the experiments and wrote the manuscript. V.P.K. 8

performed experiments within the DFG-funded research project of S.B. M.W. 9

performed metabolic extractions. A.S. analyzed the data. 10

11

12

Conflicts of Interest 13

The authors declare no conflict of interest. 14

References

1. Du, F.Y.; Navarro-Garcia, F.; Xia, Z.X.; Tasaki, T.; Varshavsky, A. Pairs of 15

dipeptides synergistically activate the binding of substrate by ubiquitin ligase 16 through dissociation of its autoinhibitory domain. Proceedings of the National 17 Academy of Sciences of the United States of America 2002, 99, 14110-14115, 18 doi:DOI 10.1073/pnas.172527399. 19

2. Takagi, H.; Shiomi, H.; Ueda, H.; Amano, H. Morphine-Like Analgesia by a 20 New Dipeptide, L-Tyrosyl-L-Arginine (Kyotorphin) and Its Analog. European 21 Journal of Pharmacology 1979, 55, 109-111, doi:Doi 10.1016/0014-22 2999(79)90154-7. 23

3. Ueda, H.; Yoshihara, Y.; Fukushima, N.; Shiomi, H.; Nakamura, A.; Takagi, H. 24 Kyotorphin (Tyrosine-Arginine) Synthetase in Rat-Brain Synaptosomes. 25 Journal of Biological Chemistry 1987, 262, 8165-8173. 26

4. Mizushige, T.; Kanegawa, N.; Yamada, A.; Ota, A.; Kanamoto, R.; Ohinata, K. 27 Aromatic amino acid-leucine dipeptides exhibit anxiolytic-like activity in young 28 mice. Neuroscience Letters 2013, 543, 126-129, 29 doi:10.1016/j.neulet.2013.03.043. 30

5. Veyel, D.; Kierszniowska, S.; Kosmacz, M.; Sokolowska, E.M.; Michaelis, A.; 31 Luzarowski, M.; Szlachetko, J.; Willmitzer, L.; Skirycz, A. System-wide 32 detection of protein-small molecule complexes suggests extensive metabolite 33 regulation in plants. Scientific Reports 2017, 7, 42387, doi:10.1038/srep42387. 34

6. Veyel, D.; Sokolowska, E.M.; Moreno, J.C.; Kierszniowska, S.; Cichon, J.; 35 Wojciechowska, I.; Luzarowski, M.; Kosmacz, M.; Szlachetko, J.; Gorka, M., et 36 al. PROMIS, global analysis of PROtein-Metabolite Interactions using Size 37 separation in Arabidopsis thaliana. The Journal of Biological Chemistry 2018, 38 10.1074/jbc.RA118.003351, doi:10.1074/jbc.RA118.003351. 39


https://doi.org/10.1101/2020.03.25.004978



7. Zaffagnini, M.; Fermani, S.; Costa, A.; Lemaire, S.D.; Trost, P. Plant 1 cytoplasmic GAPDH: redox post-translational modifications and moonlighting 2 properties. Frontiers in Plant Science 2013, 4, 450, 3 doi:10.3389/fpls.2013.00450. 4

8. Kenny, A.J.; Booth, A.G.; George, S.G.; Ingram, J.; Kershaw, D.; Wood, E.J.; 5 Young, A.R. Dipeptidyl peptidase IV, a kidney brush-border serine peptidase. 6 The Biochemical Journal 1976, 157, 169-182, doi:10.1042/bj1570169. 7

9. Kumar, S.; Kaur, A.; Chattopadhyay, B.; Bachhawat, A.K. Defining the 8 cytosolic pathway of glutathione degradation in Arabidopsis thaliana: role of 9 the ChaC/GCG family of gamma-glutamyl cyclotransferases as glutathione-10 degrading enzymes and AtLAP1 as the Cys-Gly peptidase. The Biochemical 11 Journal 2015, 468, 73-85, doi:10.1042/BJ20141154. 12

10. Caldana, C.; Degenkolbe, T.; Cuadros-Inostroza, A.; Klie, S.; Sulpice, R.; 13 Leisse, A.; Steinhauser, D.; Fernie, A.R.; Willmitzer, L.; Hannah, M.A. High-14 density kinetic analysis of the metabolomic and transcriptomic response of 15 Arabidopsis to eight environmental conditions. Plant J 2011, 67, 869-884, 16 doi:10.1111/j.1365-313X.2011.04640.x. 17

11. Wu, S.; Tohge, T.; Cuadros-Inostroza, A.; Tong, H.; Tenenboim, H.; Kooke, 18 R.; Meret, M.; Keurentjes, J.B.; Nikoloski, Z.; Fernie, A.R., et al. Mapping the 19 Arabidopsis Metabolic Landscape by Untargeted Metabolomics at Different 20 Environmental Conditions. Mol Plant 2018, 11, 118-134, 21 doi:10.1016/j.molp.2017.08.012. 22

12. Avin-Wittenberg, T.; Bajdzienko, K.; Wittenberg, G.; Alseekh, S.; Tohge, T.; 23 Bock, R.; Giavalisco, P.; Fernie, A.R. Global analysis of the role of autophagy 24 in cellular metabolism and energy homeostasis in Arabidopsis seedlings under 25 carbon starvation. The Plant Cell 2015, 27, 306-322, 26 doi:10.1105/tpc.114.134205. 27

13. Carter, C.; Pan, S.; Zouhar, J.; Avila, E.L.; Girke, T.; Raikhel, N.V. The 28 vegetative vacuole proteome of Arabidopsis thaliana reveals predicted and 29 unexpected proteins. The Plant Cell 2004, 16, 3285-3303, 30 doi:10.1105/tpc.104.027078. 31

14. Marion, J.; Le Bars, R.; Besse, L.; Batoko, H.; Satiat-Jeunemaitre, B. 32 Multiscale and Multimodal Approaches to Study Autophagy in Model Plants. 33 Cells 2018, 7, doi:10.3390/cells7010005. 34

15. von Mering, C.; Huynen, M.; Jaeggi, D.; Schmidt, S.; Bork, P.; Snel, B. 35 STRING: a database of predicted functional associations between proteins. 36 Nucleic Acids Res 2003, 31, 258-261. 37

16. Porodko, A.; Cirnski, A.; Petrov, D.; Raab, T.; Paireder, M.; Mayer, B.; 38 Maresch, D.; Nika, L.; Biniossek, M.L.; Gallois, P., et al. The two cathepsin B-39 like proteases of Arabidopsis thaliana are closely related enzymes with 40 discrete endopeptidase and carboxydipeptidase activities. Biological 41 Chemistry 2018, 399, 1223-1235, doi:10.1515/hsz-2018-0186. 42

17. Sedaghatmehr, M.; Thirumalaikumar, V.P.; Kamranfar, I.; Marmagne, A.; 43 Masclaux-Daubresse, C.; Balazadeh, S. A regulatory role of autophagy for 44 resetting the memory of heat stress in plants. Plant, Cell & Environment 2019, 45 42, 1054-1064, doi:10.1111/pce.13426. 46

18. Giavalisco, P.; Li, Y.; Matthes, A.; Eckhardt, A.; Hubberten, H.M.; Hesse, H.; 47 Segu, S.; Hummel, J.; Kohl, K.; Willmitzer, L. Elemental formula annotation of 48 polar and lipophilic metabolites using (13) C, (15) N and (34) S isotope 49 labelling, in combination with high-resolution mass spectrometry. The Plant 50


https://doi.org/10.1101/2020.03.25.004978



Journal : for cell and molecular biology 2011, 68, 364-376, doi:10.1111/j.1365-1 313X.2011.04682.x. 2

19. Masclaux-Daubresse, C.; Clement, G.; Anne, P.; Routaboul, J.M.; Guiboileau, 3 A.; Soulay, F.; Shirasu, K.; Yoshimoto, K. Stitching together the Multiple 4 Dimensions of Autophagy Using Metabolomics and Transcriptomics Reveals 5 Impacts on Metabolism, Development, and Plant Responses to the 6 Environment in Arabidopsis. The Plant Cell 2014, 26, 1857-1877, 7 doi:10.1105/tpc.114.124677. 8

20. McLoughlin, F.; Augustine, R.C.; Marshall, R.S.; Li, F.; Kirkpatrick, L.D.; 9 Otegui, M.S.; Vierstra, R.D. Maize multi-omics reveal roles for autophagic 10 recycling in proteome remodelling and lipid turnover. Nature Plants 2018, 4, 11 1056-1070, doi:10.1038/s41477-018-0299-2. 12

21. Chiang, C.S.; Stacey, G.; Tsay, Y.F. Mechanisms and functional properties of 13 two peptide transporters, AtPTR2 and fPTR2. The Journal of Biological 14 Chemistry 2004, 279, 30150-30157, doi:10.1074/jbc.M405192200. 15

22. Naka, K.; Jomen, Y.; Ishihara, K.; Kim, J.; Ishimoto, T.; Bae, E.J.; Mohney, 16 R.P.; Stirdivant, S.M.; Oshima, H.; Oshima, M., et al. Dipeptide species 17 regulate p38MAPK-Smad3 signalling to maintain chronic myelogenous 18 leukaemia stem cells. Nature Communications 2015, 6, 8039, 19 doi:10.1038/ncomms9039. 20

23. Strehmel, N.; Hoehenwarter, W.; Monchgesang, S.; Majovsky, P.; Kruger, S.; 21 Scheel, D.; Lee, J. Stress-Related Mitogen-Activated Protein Kinases 22 Stimulate the Accumulation of Small Molecules and Proteins in Arabidopsis 23 thaliana Root Exudates. Front Plant Sci 2017, 8, doi:Artn 1292 24

10.3389/Fpls.2017.01292. 25 24. Liu, D.L.Y.; Christians, N.E. Isolation and Identification of Root-Inhibiting 26

Compounds from Corn Gluten Hydrolysate. J Plant Growth Regul 1994, 13, 27 227-230, doi:Doi 10.1007/Bf00226041. 28

25. Doppler, M.; Kluger, B.; Bueschl, C.; Steiner, B.; Buerstmayr, H.; Lemmens, 29 M.; Krska, R.; Adam, G.; Schuhmacher, R. Stable Isotope-Assisted Plant 30 Metabolomics: Investigation of Phenylalanine-Related Metabolic Response in 31 Wheat Upon Treatment With the Fusarium Virulence Factor Deoxynivalenol. 32 Frontiers in Plant Science 2019, 10, doi:Artn 1137 33

10.3389/Fpls.2019.01137. 34 26. Zhang, Z.; Zhao, Y.; Wang, X.; Lin, R.; Zhang, Y.; Ma, H.; Guo, Y.; Xu, L.; 35

Zhao, B. The novel dipeptide Tyr-Ala (TA) significantly enhances the lifespan 36 and healthspan of Caenorhabditis elegans. Food Funct 2016, 7, 1975-1984, 37 doi:10.1039/c5fo01302j. 38

27. Dick, T.P.; Ralser, M. Metabolic Remodeling in Times of Stress: Who Shoots 39 Faster than His Shadow? Molecular Cell 2015, 59, 519-521, 40 doi:10.1016/j.molcel.2015.08.002. 41

28. Saini, S.K.; Ostermeir, K.; Ramnarayan, V.R.; Schuster, H.; Zacharias, M.; 42 Springer, S. Dipeptides promote folding and peptide binding of MHC class I 43 molecules. Proceedings of the National Academy of Sciences of the United 44 States of America 2013, 110, 15383-15388, doi:10.1073/pnas.1308672110. 45

29. Moussaieff, A.; Rogachev, I.; Brodsky, L.; Malitsky, S.; Toal, T.W.; Belcher, H.; 46 Yativ, M.; Brady, S.M.; Benfey, P.N.; Aharoni, A. High-resolution metabolic 47 mapping of cell types in plant roots. Proceedings of the National Academy of 48 Sciences of the United States of America 2013, 110, E1232-E1241, 49 doi:10.1073/pnas.1302019110. 50


https://doi.org/10.1101/2020.03.25.004978



30. Hafren, A.; Hofius, D. NBR1-mediated antiviral xenophagy in plant immunity. 1 Autophagy 2017, 13, 2000-2001, doi:10.1080/15548627.2017.1339005. 2

31. Hafren, A.; Macia, J.L.; Love, A.J.; Milner, J.J.; Drucker, M.; Hofius, D. 3 Selective autophagy limits cauliflower mosaic virus infection by NBR1-4 mediated targeting of viral capsid protein and particles. Proceedings of the 5 National Academy of Sciences of the United States of America 2017, 114, 6 E2026-E2035, doi:10.1073/pnas.1610687114. 7

32. Chung, T.; Phillips, A.R.; Vierstra, R.D. ATG8 lipidation and ATG8-mediated 8 autophagy in Arabidopsis require ATG12 expressed from the differentially 9 controlled ATG12A AND ATG12B loci. The Plant Journal : for cell and 10 molecular biology 2010, 62, 483-493, doi:10.1111/j.1365-313X.2010.04166.x. 11

33. Shibata, M.; Oikawa, K.; Yoshimoto, K.; Kondo, M.; Mano, S.; Yamada, K.; 12 Hayashi, M.; Sakamoto, W.; Ohsumi, Y.; Nishimura, M. Highly oxidized 13 peroxisomes are selectively degraded via autophagy in Arabidopsis. The Plant 14 Cell 2013, 25, 4967-4983, doi:10.1105/tpc.113.116947. 15

34. Shannon, P.; Markiel, A.; Ozier, O.; Baliga, N.S.; Wang, J.T.; Ramage, D.; 16 Amin, N.; Schwikowski, B.; Ideker, T. Cytoscape: a software environment for 17 integrated models of biomolecular interaction networks. Genome Research 18 2003, 13, 2498-2504, doi:10.1101/gr.1239303. 19

20

21

22

23

24


https://doi.org/10.1101/2020.03.25.004978


https://doi.org/10.1101/2020.03.25.004978

autophagy is responsible for the accumulation of proteogenic ......2020/03/25 · dipeptide levels...

Documents