Noncell‐autonomous HSC70.1 chaperone displays homeostatic feedback regulation by binding its own mRNA

Abstract

The HSC70/HSP70 family of heat shock proteins are evolutionarily conserved chaperones involved in protein folding, protein transport, and RNA binding. Arabidopsis HSC70 chaperones are thought to act as housekeeping chaperones and as such are involved in many growth‐related pathways. Whether Arabidopsis HSC70 binds RNA and whether this interaction is functional has remained an open question.
We provide evidence that the HSC70.1 chaperone binds its own mRNA via its C‐terminal short variable region (SVR) and inhibits its own translation.
The SVR encoding mRNA region is necessary for HSC70.1 transcript mobility to distant tissues and that HSC70.1 transcript and not protein mobility is required to rescue root growth and flowering time of hsc70 mutants. We propose that this negative protein‐transcript feedback loop may establish an on‐demand chaperone pool that allows for a rapid response to stress.
In summary, our data suggest that the Arabidopsis HSC70.1 chaperone can form a complex with its own transcript to regulate its translation and that both protein and transcript can act in a noncell‐autonomous manner, potentially maintaining chaperone homeostasis between tissues.

Full text

Noncell-autonomous HSC70.1 chaperone displays homeostatic
feedback regulation by binding its own mRNA
Lei Yang
1
, Yuan Zhou
1
, Shuangfeng Wang
1
, Ying Xu
1
, Steffen Ostendorp
2
, Melissa Tomkins
3
,
Julia Kehr
2
, Richard J. Morris
3
and Friedrich Kragler
1
1
Max-Planck-Institute of Molecular Plant Physiology, Wissenschaftspark Golm, Am M€uhlenberg 1, 14476 Golm, Germany;
2
Institute for Plant Science and Microbiology, Universit€at
Hamburg, Ohnhorststr. 18, 22609 Hamburg, Germany;
3
Computational and Systems Biology, John Innes Centre, Norwich, NR4 7UH, UK
Author for correspondence:
Friedrich Kragler
Email: kragler@mpimp-golm.mpg.de
Received: 4 November 2022
Accepted: 8 December 2022
New Phytologist (2023) 237: 2404–2421
doi: 10.1111/nph.18703
Key words: HSC70 chaperone, intercellular
transport, long-distance transport, mRNA
translation, plant growth, RNA transport,
translational feedback regulation.
Summary
The HSC70/HSP70 family of heat shock proteins are evolutionarily conserved chaperones
involved in protein folding, protein transport, and RNA binding. Arabidopsis HSC70 chaper-
ones are thought to act as housekeeping chaperones and as such are involved in many
growth-related pathways. Whether Arabidopsis HSC70 binds RNA and whether this interac-
tion is functional has remained an open question.
We provide evidence that the HSC70.1 chaperone binds its own mRNA via its C-terminal
short variable region (SVR) and inhibits its own translation.
The SVR encoding mRNA region is necessary for HSC70.1 transcript mobility to distant tis-
sues and that HSC70.1 transcript and not protein mobility is required to rescue root growth
and flowering time of hsc70 mutants. We propose that this negative protein-transcript feed-
back loop may establish an on-demand chaperone pool that allows for a rapid response to
stress.
In summary, our data suggest that the Arabidopsis HSC70.1 chaperone can form a complex
with its own transcript to regulate its translation and that both protein and transcript can act
in a noncell-autonomous manner, potentially maintaining chaperone homeostasis between
tissues.
Introduction
HSC70/HSP70 heat shock proteins are evolutionarily conserved
housekeeping chaperones that assist folding and formation of
functional structural domains of client proteins. By this means,
HSC70 chaperones contribute to environmental and genomic
stress tolerance (Mayer & Bukau, 2005;No€el et al., 2007). In
addition, HSC70s are often facilitating protein import into cellu-
lar compartments. Correspondingly, some plant viruses produce
HSC70-related chaperones that interact with viral RNA-protein
(RNP) complexes and intercellular channels to enable viral trans-
port to neighboring cells (Gilbertson & Lucas, 1996; Lazarowitz
& Beachy, 1999; Medina et al., 1999; Alzhanova et al., 2001),
suggesting a potential role for HSC70-related chaperones in
intercellular transport of macromolecules. Yeast and human
HSC70 chaperones are reported to bind to 30UTRs AU-rich
motifs present in distinct mRNAs either via their conserved
ATPase- and substrate-binding domains or their C-terminal
domain (Zimmer et al., 2001; Kishor et al., 2013,2017).
Although HSC70 family members bind to mRNAs in animals,
the role of chaperone-RNA interaction –beside that it might sta-
bilize interacting transcripts –remains elusive because RNA-
HSC70 interaction does not change chaperone activity (Malter,
1989; Chen & Shyu, 1995; Henics et al., 1999; Guhaniyogi &
Brewer, 2001; Kishor et al., 2017).
HSC70s are abundant in phloem exudates (Aoki
et al., 2002; Giavalisco et al., 2006) and in plasmodesmal cell
fractions (Aoki et al., 2002) and are part of phloem-specific
RNP complexes consisting of PHLOEM PROTEIN 16
(PP16), RNA-BINDING PROTEIN 50 (RBB50; related to
pyrimidine tract binding proteins), elF-5A elongation factor,
and TRANSLATIONALLY CONTROLLED TUMOR PRO-
TEIN1 (TCTP1) (Ham et al., 2009; Saplaoura & Kra-
gler, 2016). This complex attaches to 30UTR polypyrimidine
track binding RNA motifs (UUCUCUCUCUU) and poten-
tially aids mRNA long-distance transport (Ham et al., 2009;
Yang et al., 2019). In line, the individual components of this
RNP move between cells in single-cell microinjection assays
(Xoconostle-Cazares et al., 1999; Aoki et al., 2002; Ham
et al., 2009). A detailed analysis of mobile pumpkin HSC70s
revealed a C-terminal short variable region (SVR) domain,
enabling them to move to neighboring pumpkin cells (Aoki
et al., 2002). However, their function in phloem-mediated
RNA transport and a direct interaction with mobile transcripts
remains to be established. In line with the detected intercellular
mobility of pumpkin HSC70 protein, Drosophila HSC70 is
secreted via exosomes that are transferred to stressed recipient
cells where they elevate the protein-folding environment
(Takeuchi et al., 2015). By this means, HSC70 activity home-
ostasis is maintained at the organismal level without local gene
2404 New Phytologist (2023) 237: 2404–2421 Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
www.newphytologist.com
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and
distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
Research

expression changes. Again, whether plant HSC70 has a similar
function remains to be shown.
In Arabidopsis, cucumber, watermelon, grapevine, Nicotiana
benthamiana, soybean, and tomato several hundred mRNAs were
reported to move over graft junctions to distant body parts
(Notaguchi et al., 2015; Thieme et al., 2015; Yang et al., 2015;
Walther & Kragler, 2016; Z. Zhang et al., 2016; Xia et al., 2018;
Li et al., 2021). Among the set of conserved graft-mobile mRNAs
are HSC70 transcripts. Arabidopsis thaliana HSC70.1 transcript
was shown to move shoot-to-root in juvenile plants depending
on secondary ribonucleic m
5
C modifications (Yang et al., 2019),
suggesting that HSC70.1 transcript transport may depend on dis-
tinct RNA structures (W. Zhang et al., 2016; Kehr & Kra-
gler, 2018; Yang et al., 2019).
To characterize a potential HSC70-RNA interaction, we identi-
fied A. thaliana HSC70.1 interacting mRNAs, show that the
HSC70.1 C-terminal SVR motif binds to RNA, and provide evi-
dence for HSC70.1 protein binding its own mRNA to negatively
regulate its own translation. We established that HSC70.1 transcript
transport over graft junctions depends on distinct RNA and not
protein motif(s) and provide evidence for that HSC70.1 mRNA
mobility –and not HSC70.1 protein mobility –is required for nor-
mal plant growth. We propose that HSC70.1 transcript transport
maintains chaperone activity homeostasis between tissues.
Materials and Methods
Plant material and growth
Arabidopsis thaliana (L.) Heynh. Columbia-0 and respective trans-
genics and mutants were used in the grafting and complementation
experiments. Seeds were sterilized with 70% ethanol, 1% Tween
20 for 5 min, and 4% sodium hypochlorite for 5 min, washed five
times with sterile water and resuspend in 0.15% selected agar prior
transfer to plates. For root growth measurements, seeds were germi-
nated on vertical half-strength Murashige & Skoog medium plates
(½MS salts, 1% sucrose, and 1% micro agar (Duchefa Biochemie,
Haarlem, the Netherlands)) in controlled growth chambers (Perci-
val, Perry, IA, USA) under long-day light conditions (16 h : 8 h,
light : dark; 22°C:19°C, day : night; light intensity: 170 lmol
mol
1
m
2
s
1
). For rosette leaves growth and flowering time
measurement, seeds were germinated directly on soil and trans-
ferred 10 d after germination (DAG) to single pots and grown
under long-day light conditions (16 h : 8 h, light : dark; 22°C:
19°C, day : night; light intensity: 170 lmol mol
1
m
2
s
1
;rela-
tive humidity: 60%). The hsc70.1 (SALK_135531) and hsp70.4
(SALK_088253) T-DNA insertion mutants were obtained from
the Arabidopsis Biological Resource Center at Ohio State Univer-
sity (Alonso et al., 2003). hsc70.1 hsp70.4 mutant was obtained
from the Su Laboratory, Fudan University, China (Leng
et al., 2017).
Arabidopsis thaliana hypocotyl grafting
Arabidopsis thaliana seeds were vertically grown on plates
(½MS salts, 1% sucrose, and 1% micro agar (Duchefa
Biochemie)) in short-day light conditions (8 h : 16 h,
light : dark; 22°C:19°C, day : night; light intensity:
170 lmol mol
1
m
2
s
1
). Seedlings (7–8 DAG) with c.4cm
long hypocotyls were cut in the upper half of the hypocotyl
using a sterile razor blade, and silicon microtubes with
0.3 mm internal diameter were used to stabilize the graft junc-
tion. Grafted plants were transferred to new plates and verti-
cally grown (½MS salts, 1% sucrose, and 1% micro agar
(Duchefa Biochemie)) in short-day light conditions. After
grafting (6–7 d), adventitious roots appearing on the upper
hypocotyl junction were removed every day. Ten days after
grafting, successfully grafted plants were analyzed by confocal
laser scanning microscopy. Fourteen days after grafting, plants
were transferred to liquid culture (½MS salts, 1% sucrose),
prior RNA isolation at 30 d after grafting or transfer to soil
for flower time measurements.
Expression constructs
35S::YFP-HSC70 fusion constructs were produced by PCR ampli-
fying HSC70.1 full-length CDS from cDNA templates (FK1527/
FK1528) and cloning it into the pENTR4 vector (Addgene,
Watertown, MA, USA) using Bam HI and EagI sites. The
sequenced HSC70.1 CDS was transferred into pEarlygate104 bin-
ary vector (Earley et al., 2006) by LR recombinase-GW reaction.
HSC70.1 DSVR (FK1527/FK1529) and SVR (FK1530/FK1528)
fusion constructs were produced using the same strategy.
HSC70.1M and HSC70.1M DSVR codon usage modified con-
structs originated from a synthesized DNA PCR template to
replace the 30end of wild-type (WT) HSC70.1 30CDS to produce
HSC70.1M (modified CDS base 1185–1956) and HSC70.1M
DSVR (modified CDS base 1185–1812; for sequences see Sup-
porting Information Dataset S1)byusingBgIII and EagI restric-
tion sites in pENTR4-HSC70.1. Both pENTR4-HSC70.1M
constructs were transferred into pEarlygate104 binary vector by
LR recombinase reaction. The binary vectors were introduced into
Agrobacterium strain AGL1 and transformed into A. thaliana
plants using the double floral dip method (Davis et al., 2009).
RNA isolation
Grafted scion and roots were separately harvested, transferred
into 2-ml tubes containing metal beats and frozen in liquid nitro-
gen. After grinding, the broken tissue was supplemented with
0.75 ml Trizol LS Reagent (Invitrogen, Thermo Fisher, Hennigs-
dorf, Germany), homogenized (20 s vortexing), and incubated
for 5 min at RT. After adding 0.2 ml of chloroform per 0.75 ml
of TRIzol
TM
LS Reagent, samples were vortexed and incubated
for 5 min at RT and centrifuged for 15 min (12 000 gat 4°C).
The upper aqueous phase was transferred to 1.5-ml tubes and
supplemented with 1 volume isopropanol and 1/10 volume 3 M
NaOAc (pH 5.2), inverted three times, and incubated at 20°C
overnight. After centrifugation for 30 min at 4°C and the RNA
pellets were washed two times with 1 ml ice-cold 80% ethanol,
one time with 99% ethanol, dissolved in 10–25 ll DEPC H
2
O,
incubated for 5 min at 65°C, and stored at 80°C.
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
New
Phytologist Research 2405

Reverse transcription reactions and RT-PCR
Reverse transcription reaction was performed using a reverse tran-
scription kit (Promega/Invitrogen, Thermo Fisher). Total RNA
(c. 1.5 lg) and an oligo (dT) mixture were denatured for 10 min
at 70°C and annealed for 5 min at 37°C, submitted to RT reac-
tion (42°C, 90 min), and deactivated (72°C, 10 min). RT-PCR
was conducted under standard conditions after testing cDNA
samples’ quality using ACTIN2 (FK424/FK425) primers (28
cycles). For RT-PCR detection of mobile mRNA, 45 PCR cycles
(for presence) or 50 PCR cycles (for absence) specific PCR pri-
mers were applied (Dataset S2).
Quantitative RT-PCR assay
Transcript levels were measured by qRT-PCR SYBR Green
method using an ABI System Sequence Detector (Applied
Biosystems 7900HT fast Real-time PCR, /Invitrogen, Thermo
Fisher). For all assays, three technical replicates were performed.
The thermal cycling parameters were as follows: Step 1: 1 cycle,
2 min, 50°C; Step 2: 1 cycle, 10 min, 95°C; Step 3: 40 cycles,
15 s, 95°C, 1 min, 60°C. Dissociation step: 15 s, 95°C; 15 s,
60°C; 15 s, 95°C. Used specific PCR primers, see Dataset S2.
Microscopy
YFP-HSC70.1 fusion proteins in hypocotyl-grafted A. thaliana
plants (10 d after grafting) were detected using confocal laser
scanning microscope (CLSM; Leica SP5 or SP8, Leica Microsys-
tems, Wetzlar, Germany). For low-abundant signals, a HyD
hybrid detector (Leica SP8) was used. Z-stack images were assem-
bled and processed using the FIJI software package as described
previously (Yang et al., 2019).
Protein isolation and immunoblot detection
Ten-day-old plant lines (each replicate c.300–360 plants)
were harvested, and ground samples were transferred to 1.5-
ml tubes, incubated with 0.2 ml extraction buffer (50 mM
Tris–HClpH8.0,2.5mMEDTA,150mMNaCl,10%glyc-
erol, 0.5 mM PMSF, 1 Tablet/10 ml EDTA free Protease
inhibitor cocktail; Roche, Mannheim, Germany), vortexed
vigorously, incubated on ice for 10 min, centrifuged 10 min
(12 000 gat 4°C), transferred to a new 1.5-ml RNase-free
tube, 29SDS loading buffer was added prior incubating the
samples at 95°C (5 min). Proteins were separated on 10%
SDS-PAGE gels and transferred to nitrocellulose membranes.
After blocking (5% w/v dry-milk TBS-T solution) and incu-
bation with 1 : 10 000 anti-GFP from rat (Roche) antibody,
washing, and incubation with 1 : 20 000 anti-rat IgG Horse-
radish Peroxidase (HRP) secondary antibodies (Promega,
Invitrogen, Thermo Fisher). Chemiluminescent (ECL
TM
Prime
Western Blotting Detection Reagent; GE Healthcare,
Chicago, IL, USA) signal was detected using the ChemiDoc
MP Imaging system (Bio-Rad). Coomassie blue S-stained
RuBisCO large subunit served as loading control.
RNA immunoprecipitation
Transgenic seeds were vertically grown on plates under short-day
conditions for 10 d and then subjected to irradiation with
254 nm UV light at a dose of 500 mJ cm
2
(UVP CL-1000 UV
crosslinker; Analytik Jena, Jena, Germany) on ice and frozen in
liquid nitrogen after crosslinking. Ground samples were supple-
mented with 1 ml lysis buffer (150 mM NaCl, 50 mM Tris–
HCL pH 7.5, 5 mM MgCl
2
, 10% glycerol, 1% NP-40 (IGE-
PAL), 0.5 mM DTT, 1 mM PMSF, and 1% plant PIC (Gold-
Bio, St Louis, MO, USA)), incubated for 30 min on ice, and
gently mixed for lysis. After centrifugation (15 min, 12 000 gat
4°C), 1 ml supernatant was transferred to new tubes with 10 ll
RNase inhibitor (Promega, Invitrogen, Thermo Fisher) and
25 ll beads (equilibrated GFP-Trap
®
M beads; ChromoTek,
Planegg, Germany) and incubated (gentle end-over-end mixing)
2 h at 4°C. Magnetic beads were collected (inverted six times)
and washed three times with 1 ml ice-cold wash buffer (150 mM
NaCl, 50 mM Tris–HCL pH 7.5, 5 mM MgCl
2
, 0.5 mM
DTT). RNA was isolated using the Trizol protocol (see RNA Iso-
lation) and submitted to RT-PCR (see Reverse transcription reac-
tions and RT-PCR).
In vitro transcription
YFP and YFP-HSC70.1 cDNA transcription templates were pro-
duced by PCR using T7 YFP FP/T7 YFP RP and T7 YFP-
HSC70.1 FP/T7 YFP-HSC70.1 RP primers from
pEarlyGate104-HSC70.1 template. YFP-HSC70.1 and YFP
RNA were produced (100 lg) using T7 RNA transcription kit
(P1320; Promega) according to the manufacturer’s instructions.
The RNAs were analyzed by 1% agarose gel electrophoresis to
calculate the concentration and confirm the absence of degrada-
tion. For microscale thermophoresis, HSC70.1 and BAG1
protein-encoding sequences were amplified from cDNAs with an
additional T7 promotor sequence at the 50ends. Ten picomoles
of DNA was used as template per 100 ll RNA synthesis reaction.
In vitro RNA synthesis was performed as described for 3 h at
37°C (Cazenave & Uhlenbeck, 1994) and subsequently purified
using the RNA clean & concentrator-25 kit (Zymo Research,
Freiburg, Germany).
RNA-binding quantification by microscale thermophoresis
RNA-binding quantification of YFP-HSC70.1 and YFP protein
was performed with cell lysates from 14-d-old transgenic plants.
Extracts from YFP-HSC70.1 and YFP-expressing lines were
made freshly as described previously with minor modifications
(Chen et al., 2017). In brief, plant material was ground in liquid
nitrogen using mortar and pestle. Two hundred microlitres of
29MST buffer (100 mM HEPES pH 7.5, 300 mM NaCl,
20 mM MgCl
2
, 0.2% (v/v) NP-40, 2 mM PMSF, 2 mM AEBSF
(Applichem, Darmstadt, Germany), 0.5 U ll
1
RiboLock
(Thermo-Fisher)) was added per 100 mg of ground plant mate-
rial and incubated on ice for 5 min. Cell lysates were centrifuged
twice at 13 000 gfor 10 min at 4°C and diluted 1 : 1 with water.
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
Research
New
Phytologist
2406

Dilutions were made to achieve fluorescence reads between 200
and 1600 counts. For RNA-binding quantification, serial dilu-
tions of target RNAs were made and assays were performed
according to the manufacturer’s instructions. Samples were mea-
sured in standard capillaries on a Monolith NT.115 (NanoTem-
per GmbH, Munich, Germany) and analyzed using MO.
Affinity Analysis software and GRAPHPAD PRISM 5. Binding was
regarded as true when a signal-to-noise ratio and response ampli-
tude larger than 5 was achieved.
Translation assays
YFP-HSC70.1 protein was enriched from protein extract
obtained from transgenic 35S::YFP-HSC70.1 seedlings using
anti-GFP/YFP AB beads following the RNA immunoprecipita-
tion experiment (see description above) without crosslinking
treatment. YFP-HSC70.1 beads were resuspended in 100 ll BSA
(0.5 lgll
1
) solution. In vitro translation assays were performed
at 25°C for 120 min using Wheat Germ Expression (WGE) kit
(L4380; Promega), and detection of produced protein was per-
formed using the FluoroTect
TM
Green
Lys
tRNA (L5001; Pro-
mega) labeling system. Note that no or barely detectable YFP-
HSC70.1 translation was observed using the standard WGE reac-
tion protocol provided by the manufacturer. After optimization
of YFP-HSC70.1 RNA template, potassium acetate, and magne-
sium acetate concentrations, we were able to detect YFP-
HSC70.1 protein translation. The optimized in vitro translation
reaction with a final volume of 50 ll contained 25 ll wheat germ
extract, 4 ll complete amino acid mixture, 2 lgYFP-HSC70.1
(4 ll) or YFP RNA (1.5 ll), 5 ll potassium acetate (1 M), 1 ll
magnesium acetate (50 mM), 1 ll RNasin ribonuclease inhibitor
(40 U ll
1
), 2 ll FluoroTect
TM
Green
Lys
tRNA, and increasing
YFP-HSC70.1 protein (0.07, 0.14, 0.28, 0.56, or 1.12 lg bound
on AB beads resuspended in 0.5 lgll
1
BSA). The translation
reactions were terminated by placing on ice, and 10 ll samples
were used for SDS-PAGE and western blot/fluorescence detec-
tion using a Typhoon FLA 7000 imaging system (GE Health-
care) with a 532 nm excitation following the
FluoroTect
TM
Green
Lys
in vitro Translation Labelling System
instructions provided by the supplier (L5001; Promega).
In vivo HSP70 inhibition
YFP-HSC70.1, YFP-HSC70 DSVR, and YFP transgenic plants
(14 d after germination on ½MS plates supplemented with 1%
sucrose; n>20) were transferred to 5 ml liquid ½MS medium
with 1% sucrose and incubated for 15, 30, and 60 min with
150 lM HSP70 inhibitor VER-155008 (SML0271; Sigma;
resuspended in DMSO) or 0.1% DMSO (mock treatment) at
20°C. After 15, 30, and 60 min, protein samples (150 mg) from
three independent replicates for each construct and time point
were collected for western blot (1 : 10 000 anti-GFP from rat,
Roche; see Protein isolation and immunoblot detection) and
RNA samples (150 mg) for qRT-PCR (YFP-specific primers; see
methods above). The total protein loaded for western blot assays
was adjusted according to Coomassie staining and YFP fusion
protein detection was performed by chemiluminescent (ECL
TM
Prime Western Blotting Detection Reagent; GE Healthcare) and
was quantified by chemiluminescent emission detected by a Bio-
Rad ChemiDoc MP Imaging system (Bio-Rad). The IMAGE LAB
software (v.5.2.1) Quantity Tools function was used to calculate
the relative (nonsaturated) density of protein bands.
Biolistic bombardment
Leaves of 36-d-old N. benthamiana and Nicotiana sylvestris plants
were used for biolistic bombardment of DNA-coated 1 lm Gold
Microcarriers (Bio-Rad) with a Biolistic
®
PDS-1000/He Particle
Delivery System (Bio-Rad) as described previously (Winter
et al., 2007). To prepare gold microcarriers, 60 mg of gold parti-
cles was washed three times with 70% ultra-pure ethanol and
resuspended in 1 ml ultra-pure ddH
2
O. Twenty-five microlitres
of gold microcarriers (60 mg ml
1
) was mixed with binary plas-
mid DNA (c. 7 nmol; YFP-HSC70.1,YFP-HSC70.1 DSVR,
YFP-HSC70.1 SVR,orYFP), 25 ll 2.5 M CaCl
2
, vortexed for
2 min, mixed with 20 ll 0.1 M spermidine, submitted to washing
steps, and used for biolistic bombardment using the following
parameters: Helium pressure 1100 psi (7584.2 kPa); Vacuum of
27–28.0 Hg and examined under a CSLM SP8 (Leica) 36–48 h
after bombardment for the presence of YFP fluorescence in
neighboring cells as described previously (Winter et al., 2007).
RNAseq analysis of RNA immunoprecipitation
Total RNA from three replicated RNA immunoprecipitation
(RIP) samples from erGFP and HSC70.1-GFP transgenics was
submitted to Illumina cDNA library production and sequenced
(50 million reads paired-end RNAseq; BGI, Shenzhen, China).
The RNAseq datasets were analyzed against all annotated
A. thaliana CDS (Araport 11 release) using the CLC GENOMICS
WORKBENCH v.21 software (Qiagen) using default settings except
that a similarity fraction of 0.9 was set. Enrichment of log
2
>1
and FDR P-value ≤0.05 in the bulk HSC70.1 RIP sample vs the
bulk erGFP sample RNAseq reads were considered as significant
as presented in Table S1.
Computational simulation of HSC70 feedback loop
We first set up equations to describe the in vitro translation
experiments (Fig. 2c,d, see later). To represent the experimental
conditions, we assumed a constant amount of HSC70.1 mRNA,
ribosomal proteins, and associated machinery. Protein degrada-
tion was not included. A baseline translation rate for mRNA in
the absence of HSC70.1 protein was set. To mimic the enhance-
ment of ribosomal activity in the presence of chaperones, we
allowed for the translation rate to increase with HSC70.1 protein
concentration following a Hill equation. Binding of HSC70.1
mRNA to its own protein (Fig. 1d) was assumed to block its
translation. Different amounts of HSC70.1 protein were added
to the system, and the amount of translated HSC70.1 mRNA
was computed by solving the set of ordinary differential equa-
tions derived from the described interactions. After ensuring the
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
New
Phytologist Research 2407

model was able to qualitatively reproduce the experimental obser-
vations shown in Fig. 2(c,d,f,g), we then extended the equations
to simulate the in vivo system. Protein degradation was included
and was set to be the same for all proteins. Cases with and with-
out HSC70.1 binding to its own mRNA were analyzed. For the
system without HSC70.1 binding to its own mRNA, the transla-
tion rate was adjusted such that the steady-state HSC70.1 protein
levels were the same as for the case with mRNA binding.
HSC70.1 mRNA was assumed to be translated only when free,
that is, not bound to HSC70.1 protein. A stress situation was
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
Research
New
Phytologist
2408

simulated by a rapid production (at t=200) of misfolded protein
(at a steady-state concentration determined by its production and
degradation rate). The binding constant of HSC70.1 to its own
mRNA was assumed to be higher than the binding constant of
HSC70.1 to misfolded protein. We computed the dynamics of
HSC70.1 binding to misfolded proteins and assumed that this
catalyzes their refolding. The systems of ordinary differential
equations (Table S2) were solved numerically with the LSODA
method of the DESOLVE package (Soetaert et al., 2010) in R. The
reaction scheme is depicted in the information, and all equations
and parameters are listed in Table S2.
Results
HSC70.1 binds to its own and other HSC70 transcripts via
the C-terminal SVR
Previous studies indicate that human HSC70 chaperones bind
to 30UTR AU-rich elements of transcripts via their conserved
ATPase and substrate binding domains. This interaction seems
to stabilize transcripts present in the animal vasculature (Henics
et al., 1999; Kishor et al., 2013,2017). Interestingly, an analysis
of the mammalian RNA interactome indicated the HSC70 C-
terminal sequence as a potential RNA-binding region (Hentze
et al., 2018). However, to date, there are no reports on plant
HSC70s interacting with mRNAs, and if so, whether their C-
terminal SVR sequence is involved. To address these questions,
we produced transgenic lines expressing YFP fusions of full-
length HSC70.1 (YFP-HSC70.1), deletion variants lacking
either the HSC70.1 C-terminal SVR domain (YFP-HSC70.1
DSVR) or the HSC70 substrate and ATPase domain (YFP-
SVR), structurally modified HSC70.1 transcripts with alternative
codon usage (YFP-HSC70.1M,YFP-HSC70.1M DSVR), and a
transcript fusion with a stop codon after YFP (YFP(s)::
HSC70.1) (Fig. 1a; Dataset S1). These HSC70.1 fusions covered
also five stable HSC70.1 RNA structures predicted by co-
transcriptional folding (Co-Fold; Proctor & Meyer, 2013) and
minimal free energy RNA folding (MFE) algorithms (Zuker &
Stiegler, 1981; Lorenz et al., 2011; Figs 1a,S1,S2). We con-
firmed that these various YFP-HSC70.1 fusion constructs were
stably expressed in A. thaliana and that the introduced changes
did not change their cytosolic and nuclear protein localization
(Figs S3,S4).
To identify RNAs binding to HSC70.1, we performed RNA-
protein immunoprecipitation (RIP) assays on extracts from 35S::
HSC70.1-GFP and 35S::erGFP (control) transgenic plants using
anti-YFP/GFP-trap beads. RIP RNA samples (three biological
replicates per line) were submitted to RNAseq (paired-end reads)
and analyzed for significant transcript enrichment (log
2
≥1,
P≤0.05 FDR) using the A. thaliana Araport 11 cDNA database.
According to this analysis, 13 RNAs were found to be signifi-
cantly enriched in the HSC70.1-GFP RIP samples compared
with erGFP control samples (Fig. S5). Notably, nine of the 13
enriched transcripts encode RNA-binding proteins and four of
the 13 transcripts had been identified as being potentially graft-
mobile (Fig. S5; Thieme et al., 2015). The most significant
(FDR P-value 6.10E-4) enriched HSC70.1 interacting RNA
sequence was the 50region of the BCL-2-associated athanogene 1
(BAG1) transcript (Fig. S5). BAG1 is a highly conserved HSC70
co-chaperone known to bind to and regulate HSP70 chaperone
activity (Lee et al., 2016). Also, the HSC70.1 transcript was
found to be significantly enriched (FDR P-value <0.05), point-
ing toward a potential autoregulatory function based on
HSC70.1 abundance and mRNA translation.
To substantiate this finding, we performed RIP assays on sam-
ples from 35S::YFP-HSC70.1,35S::YFP-HSC70.1 DSVR,35S::
YFP-SVR and 35S::YFP-HSC70.1M transgenic plants (Fig. 1b–
d). These assays, combined with RT-PCR assays to detect the
presence of HSC70.1, indicated that HSC70.1 mRNA binds to
YFP-HSC70.1 and YFP-SVR but not to YFP-HSC70.1 DSVR
(Fig. 1d), suggesting that the SVR region binds the HSC70.1
transcript. Notably, the endogenously produced HSC70.1 tran-
script (Fig. 1d) could be detected in RT-PCR assays performed
on YFP-HSC70.1M RIP samples using a set of specific PCR pri-
mers discriminating between HSC70.1M and WT HSC70.1.
This suggests that HSC70.1 protein interacts also with other
HSC70.1 mRNAs from which it has not been translated. Also,
BAG1 RNA interaction with HSC70.1 protein was confirmed
with YFP-HSC70.1 RIP samples (Fig. 1d).
Microscale thermophoresis (MST) assays with cell extracts
from transgenic plants expressing YFP-HSC70.1 or YFP (con-
trol) confirmed that HSC70.1 binds to the discovered RNAs.
Fig. 1 YFP-HSC70.1 fusion constructs and HSC70.1 RNA immunoprecipitation (RIP) assays. (a) Schematic drawing and folding structures of HSC70.1
mRNA YFP fusion and deletion constructs used in the study. Left panel: YFP-HSC70.1 (CDS 1–1953); YFP-HSC70.1M with altered codon usage (wild-type
(WT) CDS +modified CDS 1192–1953); YFP-HSC70.1 DSVR (CDS 1–1812); YFP-HSC70.1M DSVR (WT CDS 1–1191 +modified CDS 1192–1812); YFP-
SVR (CDS 1813–1953), and YFP(s)::HSC70.1 (WT CDS 1–1953 preceded by a stop codon). SVR, short variable region. Right panel: Predicted folding of
each HSC70.1 construct according to co-transcriptional folding (Co-Fold; Proctor & Meyer, 2013). Red arrows: five stable regions consistent with RNA
folding predicted according to minimal free energy (MFE; Supporting Information Figs S1,S2). (b) Schematic drawing indicating PCR primers used to detect
HSC70.1,HSC70.1DSVR,HSC70.1M, and BAG1 transcripts in Arabidopsis thaliana RIP samples. HSC70.1 FP1/RP1: specific for WT (endogenous)
HSC70.1 and YFP-HSC70.1 does not amplify YFP-HSC70.1DSVR and YFP-HSC70.1M.HSC70.1 FP2/RP2: specific for WT HSC70.1,YFP-HSC70.1, and
YFP-HSC70.1DSVR.HSC70.1M FP3/RP3: specific for YFP-HSC70.1M.BAG1 FP1/RP1: Specifically match to BAG1 30region. (c) Control PCR with FP2/
RP2 and FP3/RP3 primer pairs showing that they discriminate between HSC70.1 and HSC70.1M sequences. (d) RT-PCR assays on RNA from YFP-
HSC70.1,YFP-HSC70.1 DSVR,YFP-SVR, and YFP-HSC70.1M input and RIP samples. Lanes 1, 2, and 3 indicate three biological replicates. Left and middle
panel: RT-PCR assays to detect HSC70.1,HSC70.1DSVR,SVR, and HSC70.1M. Right panel: control RT-PCRs with ACTIN2-specific primers. Bottom right
panel: RT-PCR assays to detect BAG1 RNA from YFP-HSC70.1 RIP samples. Black arrows, presence of HSC70.1; orange arrows, presence of endogenous
(WT) HSC70.1; red arrows, presence of HSC70.1M and BAG1 transcript in RIP samples; arrowheads, no transcript detected. INPUT, DNaseI-treated cell
extracts used for RIP assays. H
2
O, PCR contamination control with water instead of cDNA.
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
New
Phytologist Research 2409

These extracts were supplemented with in vitro produced tagged
HSC70.1 and BAG1 mRNA, and thermophoretic mobility were
measured (Fig. 2a,b). These assays established that HSC70.1 pro-
tein binds to HSC70.1 and BAG1 mRNA in the nanomolar
range.
HSC70.1 protein binding its mRNA negatively regulates its
own translation
Previous studies with Drosophila cells provided evidence that
HSP70 protein production might be controlled transcriptionally
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
Research
New
Phytologist
2410

and post-transcriptionally by regulation of both HSP70 mRNA
synthesis and mRNA destabilization and that this may depend
on HSC70 abundance (DiDomenico et al., 1982). In line with
this prediction, we found that Arabidopsis HSC70.1 binds to its
own transcript.
To test whether HSC70 directly regulates its own translation,
we asked whether increasing concentrations of YFP-HSC70.1
protein impact translation of YFP-HSC70.1 transcript in in vitro
assays. Note that YFP-HSC70.1 translation could only be
detected in the presence of high amounts of bivalent Mg
2+
,
known to alter and stabilize RNA structures (for details, see the
Materials and Methods section). To assay the HSC70.1 effect on
its own translation, we enriched YFP-HSC70.1 protein from
transgenic seedlings using specific YFP AB beads and added these
in increasing concentrations to the translation system (Figs 2c,d,
S6a,b). We detected very low or no translation of YFP-HSC70.1
transcript in the presence of >0.5 lg of YFP-HSC70.1. This was
in stark contrast to the control with YFP transcript translation
where no inhibition was observed in the presence of YFP-
HSC70.1 protein (Figs 2c,d,S6b).
Negative or balancing feedback is known to provide a means
of maintaining a desired level (Maxwell, 1868), but the impli-
cations of this for HSC70.1 were not clear. To explore how
the negative feedback regulation of HSC70.1 might impact its
ability to act as a chaperone responding to stress, we built a
computational model of the system based on the above obser-
vations. There are some key assumptions built into the model,
some of which require further validation. (1) We assumed that
the HSC70.1 binds to its own transcript with a defined equi-
librium binding constant, (2) HSC70.1 transcript cannot be
translated when bound to HSC70.1 protein, (3) HSC70.1 pro-
tein has a lower equilibrium binding constant to misfolded
proteins than to its own transcript, that is, in the presence of
misfolded protein, a shift from complexes between HSC70.1
protein and HSC70.1 transcript to complexes between
HSC70.1 protein and misfolded proteins will occur. Thus,
when no or low concentrations of misfolded proteins are pre-
sent, HSC70.1 will be at a defined level, determined by the
binding constant to its own transcript, and most HSC70.1
mRNA will be in this bound, nontranslatable form. We com-
puted what happens when the amount of misfolded protein is
increased (Figs 2e,S6c). Using the above assumptions, the
model predicts that with a translation-regulating feedback loop
(i.e. including the above assumptions of HSC70.1 binding to
HSC70.1 transcript and thereby preventing further translation),
the response time for achieving the same chaperone availability
is shorter relative to an equivalent system without feedback
(i.e. not accounting for HSC70.1 binding to HSC70.1 tran-
script in the model; Fig. 2e). This reduction in response time
as a consequence of negative feedback is consistent with other
biological systems (Rosenfeld et al., 2002).
To validate these predictions, we conducted an in vivo experi-
ment in which we measured the abundance of constitutively
expressed YFP (control), YFP-HSC70.1, and YFP-HSC70.1
DSVR (control, not binding to its own transcript) in 14-d-old
seedlings after applying the HSP70s specific inhibitor VER-
155008 (Fig. 2f,g). VER-155008 is an adenosine analog target-
ing specifically the HSC70s ATPase binding domain impairing
HSC70 chaperone function (Williamson et al., 2009; Schlecht
et al., 2013; Merret et al., 2015). We expected that the inhibitor
limits HSC70 substrate turnover, leading to an acute chaperone
demand. Here, the notion is that YFP-HSC70.1 bound to its
own transcript maintains a reservoir of both HSC70.1 mRNA
and protein, which can be made available on demand. Such a
negative feedback system based on chaperone-mRNA-binding
should show increased translation once mRNA is released,
whereas the two YFP-HSC70.1 DSVR and YFP proteins not
binding to their own transcripts should show a considerably
delayed change in translation because HSC70.1 must first be
transcribed. In other words, in the presence of VER-155008, an
acute response would be that more HSC70 protein is produced
independently of gene transcription. As anticipated, only YFP-
Fig. 2 HSC70.1 RNA-binding capacity measured by microscale thermophoresis and autoregulatory feedback regulation of translation. (a) Microscale ther-
mophoresis experiment. In vitro synthesized BAG1 and HSC70.1 RNAs ranging between 1 lM and 30 pM were titrated against cell lysates from transgenic
Arabidopsis thaliana plants expressing YFP-HSC70.1 fusion protein. Binding curves were evaluated and plotted as fraction bound against increasing RNA
concentrations. Differences in binding are visible as a shift of the sigmoidal binding curve along the x-axis. Error bars, SE. (b) EC50 estimation for both
curves. The E
C50
of BAG1 toward HSC70.1 is c. 9 times higher than toward HSC70.1 indicating a significantly higher affinity of HSC70.1 protein toward
HSC70.1 compared with BAG1. Significance was calculated using a two-tailed Student’s t-test (n=3; P<0.05). Error bars, SE. (c) Inhibition of YFP-
HSC70.1 vs YFP translation in the presence of increasing concentrations of YFP-HSC70.1 protein. YFP-HSC70.1 was extracted from 10-d-old transgenic
plants and added to the in vitro wheat germ expression assay translating either YFP (control) or YFP-HSC70.1. Anti-GFP/YFP antibody was used to detect
YFP-HSC70.1 protein added, and Green
lys
was used to detect newly synthesized YFP-HSC70.1 and YFP. (d) Relative ratios of in vitro translated YFP-
HSC70.1 and YFP protein (see also Supporting Information Fig. S6b). (e) Model of the predicted effect of HSC70.1 inhibiting its own translation on the
refolding of misfolded client proteins. A stress event is assumed to occur (at t=200 s) that gives rise to a sudden increase in misfolded protein. How quickly
this amount of misfolded protein decreases shows how well the HSC70.1 system performs. The model predicts no (blue), slow (red), or a fast (green)
refolding of client proteins due to sudden (acute) chaperone demands with no feedback or with translational feedback of HSC70.1 on its own translation
(for details, see the Results section, Fig. S6c; Table S3). (f) Western blot (WB) assays on 14-d-old YFP, YFP-HSC70.1, and YFP-HSC70.1 DSVR transgenic
wild-type (WT) plants used for HSC70.1 inhibitor (VER-155008) treatment (0, 15, 30, and 60 min). YFP, YFP-HSC70.1, and YFP-HSC70.1 DSVR fusion
proteins were detected by YFP antibody. (g) Line plot representing average density of bands measured on western blots relative to mock control (n=3
independent experiments). Significance was calculated using Student’s t-test (two tails); P-values indicated by lowercase letters: a, b <0.001. (h) qRT-PCR
assays on samples from 14-d-old YFP,YFP-HSC70.1, and YFP-HSC70.1 DSVR transgenic WT plants treated with HSC70.1 inhibitor VER-155008 for 0, 15,
30, and 60 min. Compare with nontreated samples, all treated samples show no significant differences in YFP and YFP-HSC70.1 transcript levels (P-
values >0.05; n=3 biological replicates; 4 technical replicates). Y-axis: relative transcript levels of YFP and YFP-HSC70.1 normalized to UBQ10. Error bars,
SE. Significance was calculated using Student’s t-test (two tails).
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
New
Phytologist Research 2411

HSC70.1, and not YFP or YFP-HSC70.1 DSVR, protein levels
increased significantly by c. 1.59after 15 and 30 min of VER-
155008 incubation (Fig. 2f,g; Table S1). Notably, YFP-
HSC70.1 protein levels returned to normal after 1 h showing no
significant differences in preincubation levels. Also, transcription
from the constitutive 35S promoter was not significantly changed
by VER-155008 treatment as indicated by qRT-PCR assays
(Fig. 2h). Thus, the detected specific and significant increase
in YFP-HSC70.1 protein in such a short period was due to an
increased YFP-HSC70.1 translation and not due to a change in
gene expression of the 35S promoter constructs.
Our observations imply that A. thaliana HSC70.1 regulates its
own translation by interacting with its mRNA. This constitutes a
negative regulatory feedback loop between HSC70.1 transcript
translation and cellular HSC70.1 protein demands and consti-
tutes a stable/homeostatic system that can respond rapidly to
fluctuating environmental conditions (Figs 2e,S6c).
HSC70.1 transcript mobility depends on the SVR RNA-
binding motif
Arabidopsis thaliana HSC70.1 transcript lacking its endogenous
50and 30UTR sequences fused to GFP moves from aboveground
tissue (shoot, scion) to roots (rootstock) in hypocotyl-grafted
plants (Yang et al., 2019). This suggests that if there is a specific
sequence motif that promotes mobility it does not reside in the 50
or 30UTR. To determine whether transport of HSC70.1 tran-
script depends on the protein-coding mRNA sequence, we first
grafted the YFP-HSC70.1 transgenic lines (TG) with WT (Col-
0) and assayed the presence of HSC70.1 fusion constructs in
heterologous tissues (Figs 3a,S7). Consistent with previous
results (Yang et al., 2019), CLSM and RT-PCR assays revealed
that both full-length HSC70.1 fusion protein and HSC70.1
fusion transcript produced in transgenic shoots could be detected
in WT roots (Fig. 3a,b).
We next tested the mobility of YFP-HSC70.1M, which has a
changed codon usage in the 30half of the HSC70.1 protein-
coding sequence. HSC70.1M produces WT HSC70 protein but
shows changes in the predicted folding of four out of the five pre-
dicted stable RNA folding structures (Figs 1a,S1; Dataset S1). In
grafted plants, YFP-HSC70.1M protein produced in scions was
abundantly detected in WT roots (Fig. 3a); however, YFP-
HSC70.1M transcript was not detectable by RT-PCR (Fig. 3b).
This suggests that YFP-HSC70.1 transcript and YFP-HSC70.1
protein can be independently transported and that mobility of
the YFP-HSC70.1 transcript depends on the protein-coding
HSC70.1 RNA sequence (Figs 1a,S1,S2).
To address whether the HSC70.1 transcript sequence encoding
the SVR region is necessary for mobility, we examined the trans-
port of YFP-HSC70.1 constructs lacking the SVR RNA sequence
(YFP-HSC70.1 DSVR and YFP-HSC70.1M DSVR; lacking the 30
SVR region stretching from base +1813 to +1953). Neither SVR
deletion transcripts were detected in grafted WT roots in RT-
PCR assays (Fig. 3b), indicating that this mRNA region is
enabling HSC70.1 mobility. We next asked whether the SVR
encoding RNA sequence is sufficient to mediate mobility. To this
end, we grafted YFP-SVR expressing transgenic shoots with WT
roots (Fig. 3a). Here, the transcript was detected in grafted WT
roots, suggesting that the fused SVR encoding sequence-mediated
YFP mRNA transport (Fig. 3b).
We also asked whether graft mobility of the YFP-HSC70.1
transcript depends on HSC70.1 translation by introducing a stop
codon between the YFP and HSC70.1 RNA fusion (YFP(s)::
HSC70.1). This transcript contains the SVR-encoding RNA
sequence necessary for HSC70.1 RNA mobility but is not trans-
lated (Figs 1a,S4). Notably, this nontranslated HSC70.1 RNA
sequence was not sufficient to mediate transport of YFP RNA
(Fig. 3b). The YFP protein is similar high expressed from the
nonmobile YFP(s)::HSC70.1 fusion transcript as seen in 35S::
YFP transgenic plants (Fig. S4), indicating that it is a functional
and stable expressed fusion transcript. Thus, the observed lack of
YFP(s)::HSC70.1 mobility suggests that protein translation of
HSC70.1 transcript is required to mediate its transport. Here,
one should also note that the YFP protein produced by the non-
mobile YFP(s)::HSC70.1 transcriptional fusion is small enough
to diffuse over graft junctions via the phloem; hence, YFP is
detected in WT roots grafted onto YFP(s)::HSC70.1 scions
(Fig. 3a).
Arabidopsis HSC70.1 protein transport is independent of its
SVR motif
In contrast to the reported role of the pumpkin CmHSC70.1
SVR motif in mediating intercellular mobility in microinjection
assays (Aoki et al., 2002), the A. thaliana HSC70.1 SVR protein
sequence does not seem to be necessary for HSC70.1 shoot-to-
root transport in grafted plants (Fig. 3a). Previous work indicated
that free YFP protein (and not YFP mRNA) is delivered via the
phloem to the root in grafted plants (Yang et al., 2019) and that
proteins >60 kDa cannot move freely between cells and via the
phloem to grafted roots (Paultre et al., 2016). In line, compared
with YFP, which is unloaded into the nonvascular root cells at
the root tip, the distribution of the various YFP-HSC70.1 fusions
appeared to be restricted to the root vasculature (Figs 3a,S8a).
All YFP-HSC70.1 protein fusion variants were found in the
grafted root vasculature independent of the SVR region and of
their sizes, which ranged from >100 kDa (YFP-HSC70.1) to
36 kDa (YFP-SVR; Figs S3,S8a). In particular, the large YFP
fusion proteins produced by YFP-HSC70.1 DSVR,YFP-
HSC70.1M DSVR,andYFP-HSC70.1M transgenic plants, whose
transcripts were not detected in the grafted WT roots (Fig. 3b),
appeared as full-length protein fusions in WT roots (Fig. S8b).
Thus, it seems that the HSC70.1 SVR domain does not play a
crucial role in providing long-distance transport of HSC70.1
protein in grafted A. thaliana plants.
To address whether the A. thaliana HSC70.1 SVR motif plays
a similar role in local cell-to-cell transport as seen with pumpkin
HSC70.1 (Aoki et al., 2002), we performed transient single-cell
expression assays after biolistic gold particle bombardment of
N. benthamiana and N. sylvestris leaves. Previous work revealed
that free GFP with a size of c. 27 kDa diffuses via plasmodesmata
to neighboring epidermal cells (Oparka et al., 1999).
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
Research
New
Phytologist
2412

Analogously, we used free YFP protein with a size of c. 31 kDa
(with C-terminal linker) expressed from the binary vector used to
produce the YFP-HSC70.1 fusion construct as a reference for
local cell-to-cell transport. In control assays with YFP (c.
31 kDa), 19% (8 out of 42) of bombarded cells showed YFP flu-
orescence in one neighboring cell and 9.5% (4 out of 42) in more
than one neighboring cell 36 h after bombardment. In contrast
to YFP alone, mobility of the HSC70.1 fusion constructs was
higher. YFP-HSC70.1 (c. 102 kDa) appeared in 27.3% (27 out
of 99), YFP-HSC70.1 DSVR (c. 98 kDa) in 37.3% (21 out of
68), and YFP-SVR (c. 36 kDa) in 30.9% (40 out of 107) in one
neighboring cell. YFP-HSC70.1 appeared 7.1% (7 out of 99),
YFP-HSC70.1 DSVR 16.2% (11 out of 68), and YFP-SVR 28%
(30 out of 107) of the time in more than one neighboring cell,
respectively (Fig. 4a,b). Also, YFP fusion protein mobility
seemed to be independent of cell size (Fig. 4c). However, cell-to-
cell mobility of the HSC70.1 fusion proteins was significantly
higher than that observed with free diffusing YFP and the overall
mobility of the YFP-SVR construct was significantly higher than
YFP alone, and the YFP-HSC70.1 and YFP-HSC70.1 DSVR
fusions (Fig. 4b). This is in contrast to previous reports on pump-
kin HSC70 mobility assays based on microinjection, suggesting
that the SVR region is necessary for HSC70 intercellular trans-
port in pumpkin (Aoki et al., 2002). The HSC70.1 SVR domain
–although its presence significantly increased YFP mobility com-
pared with YFP alone (Fig. 4b)–seems not to be essential for
Fig. 3 Mobility of the YFP-HSC70.1 transcript fusion variants in grafted Arabidopsis thaliana. (a) YFP fluorescence detected by confocal local scanning
microscopy (CLSM) in leaves, graft junctions, and primary roots of grafted YFP fusion transgenic/wild-type (WT; Col-0) plants. Blue color, auto-
fluorescent; green color, YFP fluorescence. Three independent lines (each n>30 plants) were used for each graft combination showing a similar YFP signal
distribution. Bars, 200 lm. (b) RT-PCR detection of YFP fusion constructs in root and shoot samples from grafted plants. Grafted plant material was pooled
(n>6) and tested for transcript presence in shoots and roots (45 PCR cycles). Note that transcript absence was additionally confirmed by 50 PCR cycles.
Red arrows, presence; red arrowheads, absence of YFP RNA fusion constructs in grafted WT (Col-0) tissue. Bar, BASTA transcript expressed by the trans-
genic plants serving as a RT-PCR contamination control; ACTIN, positive RT-PCR control with ACTIN2-specific PCR primers.
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
New
Phytologist Research 2413

local HSC70.1 cell-to-cell movement in single-cell expression
assays.
Mobile YFP-HSC70.1 transcript complements the root
growth of hsc70.1 hsp70.4 mutants
We noticed that 10 out of 18 A. thaliana HSC70 family members
were found to be graft-mobile (Thieme et al., 2015; Table S2),
indicating that there may be a benefit of HSC70s transcripts act-
ing noncell-autonomously. No phenotypes were reported for
hsc70 single mutants. However, hsc70.1 hsp70.4 double and
hsp70.2 hsp70.4 hsp70.5 triple mutants develop strong pheno-
types with curly/round leaves, twisted petioles, thin stems, early
flowering, and short siliques (Leng et al., 2017). Thus, we asked
whether graft-mobile HSC70.1 protein or transcript is necessary
to complement the reported hsc70.1 hsp70.4 double-mutant phe-
notype. To this end, we produced independent hsc70.1 hsp70.4
double-mutant lines (n>59) expressing mobile YFP-HSC70.1
and nonmobile YFP-HSC70.1M transcripts (Fig. S9). Note that,
in contrast to transgenic YFP-HSC70.1 and YFP-HSC70.1M
WT lines, the transgenic YFP-HSC70.1 and YFP-HSC70.1M
hsc70.1 hsp70.4 double mutants showed a high degree of patchy
Fig. 4 Single-cell expression and intercellular movement of YFP-HSP70.1 fusions. (a) Representative confocal local scanning microscopy (CLSM) images of
Nicotiana benthamiana epidermal cells bombarded with 35S promoter constructs expressing YFP-HSC70.1, YFP-HSC70.1DSVR, YFP-SVR, or YFP. Green,
YFP fluorescence; blue, chloroplastic auto-fluorescence; white arrows, neighboring cells with YFP fluorescence. Numbers indicate the fraction of analyzed
YFP-expressing cells that show a YFP signal in neighboring cells. Bars, 100 lm. (b) Intercellular movement ratio in % of all expressing cells and fraction
showing movement to one or two and more (≥2) neighboring cells. Significance was calculated using Student’s t-test (two tails); P-values indicated by low-
ercase letters: a, b <0.01; b, c <0.05; a, c <0.001. (c) Size distribution of cells with detected incidents of cell-to-cell mobility. Note that the maximum diam-
eter was measured and that individual points indicate cells in which the YFP construct moved. Significance was calculated using Student’s t-test (two tails).
No significant difference was detected (P-values a >0.19).
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
Research
New
Phytologist
2414

YFP fluorescence indicative for gene silencing (Dalmay
et al., 2000; Fig. S3a–ecompared with Fig. S9a–e). Relatively
high YFP-HSC70.1 silencing was also apparent in hsc70.1 single
mutants and may be explained by strong upregulation of other
HSC70 family members in hsc70 mutants (Taipale et al., 2014;
Leng et al., 2017) as detected with induced endogenous
HSC70.4 expression in the hsc70.1 mutant (Fig. S9g). Despite
this, we were able to select relatively stable YFP-HSC70.1 (n=2)
and YFP-HSC70.1M (n=3) transgenic lines, allowing us to ask
whether these fusion constructs can complement the hsc70.1
hsp70.4 phenotype. Expression of the constructs in transgenic
lines was confirmed by western blot and quantitative RT-PCR
assays. All five transgenic lines produced similar levels of YFP
fusion protein (Fig. S9f), YFP fusion transcript levels (Fig. S9g),
and showed a similar intracellular distribution of the fusion pro-
tein (Fig. S9a–e). Also, no significant difference in germination
time and rate between the transgenic and WT lines was noticed
in the used lines (Fig. S10).
We next measured primary root growth of the selected plant
lines (Fig. 5a,b). This revealed that WT, hsc70.1, and hsp70.4
single mutants have significantly longer primary roots than
hsc70.1 hsp70.4 double mutants. The two independent YFP-
HSC70.1 #1 (hsc70.1 hsp70.4) and YFP-HSC70.1 #2 (hsc70.1
hsp70.4) transgenic lines showed similar primary root length as
WT plants. By contrast, the three independent hsc70.1 hsp70.4
transgenic plant lines expressing the nonmobile YFP-HSC70.1M
(line #6, #9, and #11) transcript showed no significant root
growth differences compared with hsc70.1 hsp70.4 double
mutants. To confirm that the YFP-HSC70.1 transcript is mobile
despite the lack of endogenous HSC70.1 and HSC70.4 chaper-
one activity, we grafted YFP-HSC70.1 #1 (hsc70.1 hsp70.4)
scions with hsc70.1 hsp70.4 roots. RT-PCR assays on root RNA
samples from grafted juvenile plants (Fig. 5c) indicated that the
mobility of YFP-HSC70.1 transcript was independent of endoge-
nous HSC70.1 activity. Notably, although the YFP-HSC70.1
protein produced by the nonmobile YFP-HSC70.1M transcript
was still transported to WT roots (Figs 3a,S8), it failed to rescue
primary root growth (Fig. 5). Given that the transcript levels
between the two fusion constructs are not significantly different
(Fig. S9g), this finding suggests that only graft-mobile HSC70.1
transcript –and not graft-mobile HSC70.1 protein, is needed to
restore WT growth of a hsc70.1 hsp70.4 mutant.
Fig. 5 Mobile HSC70.1 transcript rescues hsc70.1 hsp70.4 root growth.
(a) Representative pictures of analyzed Arabidopsis thaliana wild-type
(WT; Col-0), hsc70.1,hsp70.4,hsc70.1 hsp70.4,YFP-HSC70.1 #1
(hsc70.1 hsp70.4), YFP-HSC70.1 #2 (hsc70.1 hsp70.4), YFP-HSC70.1M
#6 (hsc70.1 hsp70.4), YFP-HSC70.1M #9 (hsc70.1 hsp70.4), and YFP-
HSC70.1M-1 #11 (hsc70.1 hsp70.4) plants 14 d after germination.
(b) Quantitative data of measured primary root length of WT and indi-
cated mutant plants. Box plot graph: boxes denote variation between
datasets and means; error bar, SE; black dots, measurements out of SE
range. n, number of analyzed plants. Significance was evaluated using
one-way ANOVA (a=0.05) followed by multiple comparisons of means
using Tukey’s HSD test at the 0.05 significance level. P-value indicated by
lowercase letters: 0.006 <a, b <0.018; b, c <1.53E-06; a, c <5.24E-12.
ab, no significant differences (P-value >0.24). (c) RT-PCR detection of
YFP-HSC70.1 transcript in root and shoot samples from YFP-HSC70.1 #1
(hsc70.1 hsp70.4)/hsc70.1 hsp70.4 grafted plants. Grafted plant material
was pooled (n>6) and tested for transcript presence in shoots and roots
(45 PCR cycles). Note that transcript absence was additionally confirmed
by 50 PCR cycles. Red arrows, presence of YFP HSC70.1 RNA in grafted
root tissue. Bar, BASTA transcript expressed by the transgenic plants serv-
ing as a RT-PCR contamination control; ACTIN2, positive RT-PCR control
with ACTIN2-specific PCR primers.
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
New
Phytologist Research 2415

Mobile HSC70.1 transcript rescues early flowering of
hsc70.1 hsp70.4 double mutants
In our phenotypic analysis of hsc70.1 hsp70.4 double mutants,
we noticed an early flowering (bolting) phenotype under long-
day growth conditions. Thus, we compared the flowering time of
WT, hsc70.1 and hsp70.4 single mutants, hsc70.1 hsp70.4 double
mutants, and transgenic YFP-HSC70.1 and YFP-HSC70.1M
hsc70.1 hsp70.4 double mutants (Fig. 6). This comparison
revealed that WT, hsc70.1 and hsp70.4 single mutants, and YFP-
HSC70.1 (hsc70.1 hsp70.4) transgenics bolted 32–36 d after ger-
mination (DAG), whereas hsc70.1 hsp70.4 double mutants and
YFP-HSC70.1M (hsc70.1 hsp70.4) transgenics bolted signifi-
cantly earlier between 27–32 DAG and 27–34 DAG, respectively
(Fig. 6b). To substantiate this finding, we also counted the num-
ber of rosette leaves present on flowering plants. At the time of
bolting the WT, hsc70.1 and hsp70.4 single mutants had 15–20
rosette leaves, which was significantly more than detected on
hsc70.1 hsp70.4 mutant plants with 13–17 rosette leaves
(Fig. 6c). YFP-HSC70.1 expressing hsc70.1 hsp70.4 mutants
formed 14–21 rosette leaves, which was similar to WT and
hsc70.1 and hsp70.4 single mutants. By contrast, YFP-
HSC70.1M (hsc70.1 hsp70.4) transgenic plants showed 11–18
rosette leaves, which was not significantly different compared
with hsc70.1 hsp70.4 double mutants (Fig. 6c). These results
indicate that the shoot-to-root mobile HSC70.1 protein encoded
by the nonmobile HSC70.1M mRNA was not sufficient to rescue
the root growth and early flowering phenotype of hsc70.1
hsp70.4 double mutants (Figs 5, 6).
We next expressed YFP(s)::HSC70.1 fusion transcript in
mutant plants to examine whether nontranslated HSC70.1 RNA
would complement an hsc70 mutant phenotype. This fusion
transcript does not produce an HSC70.1 protein (Figs 1a,S4),
and the transcript is not mobile in grafted plants (Fig. 3b). To
this end, we analyzed the phenotypes of hsc70.1 hsp70.4 double-
mutant lines expressing the YFP(s)::HSC70.1 fusion (Fig. S4)
and compared these to WT (Col-0) transgenic (n>13) and
hsc70.1 hsp70.4 mutants. Three independent, stable YFP(s)::
HSC70.1 (hsc70.1 hsp70.4) transgenic lines were selected, and
their root growth and flowering time were measured (Figs S11,
S12). None of the three YFP(s)::HSC70.1 transgenic lines
showed significant root growth and flowering time differences
compared with hsc70.1 hsp70.4 double-mutant plants. Notably,
the nontranslatable HSC70.1 mRNA fused to YFP seems to be
stable, showing similar high YFP fluorescence and protein levels
compared with YFP-HSC70.1 expressing lines (Figs S3,S8).
Thus, in contrast to mobile YFP-HSC70.1 transcript, expression
of nontranslatable and nonmobile HSC70.1 fusion transcript did
not complement the mutant phenotypes.
We finally asked whether phenotype complementation can be
achieved by the endogenous HSC70.1 transcript independent of
a ubiquitous 35S promoter-driven HSC70.1 fusion construct.
We grafted hsp70.4 single mutants expressing WT HSC70.1 with
hsc70.1 hsp70.4 double mutants and evaluated the flowering phe-
notype of grafted hsc70.1 hsp70.4 scions (Fig. 7). Here, homo-
grafted hsp70.4 and homografted hsc70.1 hsp70.4 plants served as
Fig. 6 Mobile HSC70.1 transcript rescues hsc70.1 hsp70.4 early flowering
phenotype. (a) Representative pictures of analyzed Arabidopsis thaliana
wild-type (WT; Col-0), hsc70.1,hsp70.4,hsc70.1 hsp70.4,HSC70.1 #1
(hsc70.1 hsp70.4), HSC70.1 #2 (hsc70.1 hsp70.4), HSC70.1M #6
(hsc70.1 hsp70.4), HSC70.1M #9 (hsc70.1 hsp70.4), and HSC70.1M-1
#11 (hsc70.1 hsp70.4) plants 30 d after germination. (b) Age of plants at
bolding of WT and indicated mutant plants. (c) Numbers of rosette leaves
at the time of bolding of WT and indicated mutant plants. Box plot graph:
boxes indicate variation between datasets and means; error bars, SE;
black dots, measurements out of range SE. n, number of plants ana-
lyzed; significance was evaluated using one-way ANOVA (a=0.05) fol-
lowed by multiple comparisons of means using Tukey’s HSD test at the
0.05 significance level. P-value indicated by lowercase letters: a, b <0.005;
0.035 <b, c <0.99; a, c <7.82E-05. ab and bc, no significant differences
(P-value >0.9).
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
Research
New
Phytologist
2416

a control for the flowering time phenotype. Homografted hsc70.1
hsp70.4/hsc70.1 hsp70.4 plants started to bolt after 27–30 d
whereas homografted hsp70.4/hsp70.4 plants started to bolt sig-
nificantly later after 31–34 d. In contrast to homografted hsc70.1
hsp70.4 (scion)/hsc70.1 hsp70.4 (root) plants, hsc70.1 hsp70.4
(scion)/hsp70.4 (root) heterografted mutants started to bolt signif-
icantly later after 31–34 d (Fig. 7a,c), suggesting that hsp70.4
rootstocks expressing the endogenous HSC70.1 could suppress
the earlier hsc70.1 hsp70.4 scion flowering phenotype. Notably,
the complementation of the early flowering phenotype correlated
with endogenous HSC70.1 mRNA movement from hsp70.4
rootstock to hsc70.1 hsp70.4 scion in flowering plants, as revealed
by RT-PCR assays (Fig. 7b). In summary, the complementation
and grafting assays suggest that both, HSC70.1 transcript mobil-
ity followed by HSC70.1 protein expression, rescues the growth
phenotypes of hsc70.1 hsp70.4 mutants. Furthermore, the find-
ings support the notion that mobility of HSC70.1 transcript –
although it is expressed in all tissues in WT plants –is necessary
for proper plant growth.
Discussion
HSC70s are housekeeping chaperone and as such involved in
many important pathways related to plant growth
(Vierling, 1991; Sung et al., 2001; Cazale et al., 2009; Clement
et al., 2011; Leng et al., 2017). HSC70.1 is expressed at equally
high levels in all plant cell types and c.2–3 times higher in divid-
ing and endoreduplicating cells (Apelt et al., 2022). Although
HSC70.1 transcription was reported to be induced by severe heat
stress, it does not appear to be changed by a first mild and a later
applied severe heat stress (Olas et al., 2021). In line, there is a low
correlation between gene expression and protein abundance of
HSC70 family members, suggesting that a post-transcriptional
control mechanism is in place regulating HSP70s presence (Berka
et al., 2022). In animals, while a direct interaction between
HSC70 proteins and HSC70 transcripts was not established,
HSC70 expression was postulated to be regulated on the transla-
tional level (DiDomenico et al., 1982). Although HSC70s are
described as bona fide RNA-binding proteins proposed to stabi-
lize binding transcripts (Kishor et al., 2013,2017), the relevance
of the reported HSC70 RNA-binding capacity has remained
unclear. Our thermophoretic RNA-binding studies and transla-
tional assays with plant-produced HSC70.1 protein samples
(Fig. 2) show that HSC70.1 binds to its own transcript with high
affinity and that HSC70.1 inhibits specifically its own transla-
tion. In line, specific inhibition of HSC70 chaperone activity in
plants results in an increase in HSC70.1 translation after 15 min
independent of transcriptional changes (Fig. 2). Taken together,
Fig. 7 Grafted Arabidopsis thaliana hsp70.4
mutants producing wild-type (WT) HSC70.1
transcript complement early hsc70.1 hsp70.4
flowering. (a) Representative pictures of
root/shoot grafted hsc70.1 hsp70.4 mutant
plants 29 d after grafting. (b) RT-PCR assays
confirming the presence of mobile HSC70.1
transcript produced in hsp70.4 mutant tissue
in heterologous hsc70.1 hsp70.4 (DKO) root
and shoot tissue (arrowheads). (c) Time of
flowering (bolding) in days after grafting.
Box plots: boxes indicate the variation
between datasets and means; 16 grafted
plants were analyzed for each graft
combination; error bars, SE; black dots,
measurements out of range SE.
Significance was evaluated using one-way
ANOVA (a=0.05) followed by multiple
comparisons of means using Tukey’s HSD
test at the 0.05 significance level. P-value
indicated by lowercase letters: a, b <6.67E-
16.
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
New
Phytologist Research 2417

this points to a negative feedback regulation of HSC70.1 transla-
tion (Fig. 2e). In such, a feedback regulation high demand for
chaperone activity would dissociate HSC70.1 from its own tran-
script, freeing up HSC70.1 protein and making mRNA available
for translation.
Given the redundancy with >18 HSC70 chaperones expressed
in Arabidopsis, their highly conserved sequence and the house-
keeping function, and their suggested role of chaperones in com-
pensating random somatic mutations (Queitsch et al., 2002), it
might be surprising that 10 of these HSC70-related transcripts
were identified as graft-mobile (Thieme et al., 2015). Supporting
the notion that mRNA long-distance transport has a function,
rescue of early flowering and WT-like root growth was provided
by mobile HSC70, and not by nonmobile transcript versions
(Figs 5, 6). Considering the high cellular chaperone activity
required and the low protein transport rates detected in grafted
plants, it is not surprising that mobile HSC70.1 protein alone
was not sufficient to compensate for the loss of HSC70 function.
Another question was whether HSC70.1 transcript mobility
depends on the capacity of HSC70.1 protein to bind its RNA.
We could demonstrate that HSC70.1 RNA transport over graft
junctions is mediated by a defined RNA sequence motif that
coincides with the SVR encoding region mediating HSC70 pro-
tein binding to its RNA. Thus, it seems that both features, the
capacity of the SVR motif to bind to RNA (Fig. 1d) and the pres-
ence SVR encoding sequence, enable HSC70.1 transcript trans-
port (Fig. 3). However, although the SVR motif significantly
enhances intercellular protein transport (Fig. 4), its presence is
not essential for long-distance transport in grafted plants (Fig. 3).
Notably, the YFP-expressing fusion transcript fused to nontrans-
lated HSC70.1 RNA was not graft-mobile (Fig. 1c), which sug-
gests that the SVR RNA sequence has to be translatable to be
recognized as an RNA transport motif. Whether this is a conse-
quence of the translation process itself or induced changes in
RNA structures during translation remains unclear. Here, one
could speculate that –when not interacting with client proteins –
HSC70.1 protein binds to HSC70.1 mRNAs at ribosomes dur-
ing or soon after translation.
In summary, we established that the SVR motif enhances inter-
cellular protein transport and mediates HSC70.1 binding to its
Fig. 8 Speculative model of HSC70 functioning as a noncell-autonomous mobile chaperone regulating its own translation. HSC70 can block its own transla-
tion by binding its mRNA (negative feedback). In nonstressed cells (with regular chaperone activity demands), this will result in a balanced pool of com-
plexes of HSC70 bound to its own mRNA and of translated HSC70 protein available for folding of client proteins. HSC70 transcript and protein, possibly as
HSC70 transcript-protein complexes, can move between cells and to distant tissues. The function of this movement remains to be determined, but poten-
tially plays a role in maintaining homeostatic chaperone levels between tissues/cells over longer time periods. Under stress conditions, the HSC70
transcript-protein complexes may dissociate, freeing up both active HSC70 protein and relieving the auto-inhibitory activity on its own translation, thus
allowing for a rapid (minutes) response to sudden increases in chaperone demand. After longer time periods (hours), HSC70 gene expression can be
induced in a stressed tissue/cell producing more of HSC70 protein and RNA. As HSC70 transcript and protein can move, they would be delivered to non-
stressed tissues/cells. In recipient cells, imported additional HSC70 protein might interfere with translation and lower chaperone activity in distant tissues
resulting in coordinated decreased growth in, for example, a locally stressed plant. In parallel, recipient cells/tissues might be able to adopt faster to antici-
pated stress by additionally received HSC70 mRNA. Both features, HSC70 protein –transcript auto-inhibitory activity and transcript/protein transport
between cells, may enable a multicellular organism to establish chaperone homeostasis between, for example, stressed and nonstressed tissues over long
time periods providing additional robustness under stress conditions to ensure coordinated growth within and between tissues.
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
Research
New
Phytologist
2418

own and other transcripts and that the SVR encoding RNA
sequence is necessary for HSC70.1 transcript transport from shoot
to root and vice versa. Furthermore, the HSC70.1 chaperone
potentially negatively regulates its own translation. One could
speculate that these features have the potential to create a noncell
autonomous homeostatic chaperone system allowing cells/tissues
to adopt to acute local stresses within minutes (Fig. 8)asindicated
by the application of chaperone inhibitor resulting in a transla-
tional change independent of transcription (Fig. 2). This has the
potential to balance chaperone activity over longer time periods
between distant tissues similar to that found with secreted chaper-
ones in Drosophila (Takeuchi et al., 2015). We hypothesize that
the autoregulatory feedback regulation of HSC70 translation acts
locally within minutes whereas the slow long-distance transport of
HSC70 RNA is required for coordinated growth of distant tissues
when parts of a plant are locally exposed to stresses over extended
time periods and under normal conditions. This model would be
consistent with our finding that mobile HSC70.1 transcript, and
not mobile HSC70.1 protein, is the critical determinant for nor-
mal growth (Figs 5, 7) and that when produced in roots can com-
plement the mutant early flower phenotype when grafted on roots
expressing WT HSC70.1 (Fig. 7). This model is also based on the
observation that transported mRNA is translated in receiving cells
(W. Zhang et al., 2016).
Given that HSC70 chaperones are also noncell-autonomous
proteins in animals (De Maio, 2014) acting in distant cells
(Takeuchi et al., 2015), we propose that both intercellular trans-
port of HSC70s and negative feedback regulation of its own
translation is an evolutionarily conserved feature potentially con-
stituting a relatively simple way to coordinate and balance chap-
erone activity within and between tissues.
Acknowledgements
We thank Saurabh Gupta (MPI-MPP, Golm, Germany) for his
advice to analyze the deep sequencing data and Dana Schindelasch
(MPI-MPP, Golm, Germany) for her excellent technical support;
Wei Su (Fu Dan University, Shanghai, China) for providing
hsc70.1 hsp70.4 double-mutant seeds and the anonymous reviewers
of the initial version of the manuscript for their helpful comments
and corrections. This work was supported by MPI-MPP Internal
Funds to FK; LY, YZ, and YX were supported by a Chinese Scho-
larship Council (CSC) PhD Stipend; this article is part of a project
that has received funding from the European Research Council
(ERC) under the European Union’s Horizon 2020 research and
innovation program (Grant agreement no. 810131). Open Access
funding enabled and organized by Projekt DEAL.
Competing interests
None declared.
Author contributions
LY and FK conceived the project and suggested experiments; LY
produced the transgenic plants, performed the grafting
experiments, analyzed the growth and mobility data, and with
YZ performed bombardment experiments; SW with LY per-
formed the in vitro and in vivo expression experiments. YX per-
formed RNA IP experiments. FK analyzed the RIP RNAseq
data; RJM developed the model for the simulation of feedback
regulation; MT analyzed the HSC70.1 RNA sequences; SO with
JK suggested and SO performed the thermophoretic RNA inter-
action assays. FK assisted by LY wrote the manuscript with con-
tributions and input from all authors. All authors read the
manuscript text and expressed their consent to submit the final
version of the manuscript.
ORCID
Julia Kehr https://orcid.org/0000-0003-3617-9981
Friedrich Kragler https://orcid.org/0000-0001-5308-2976
Richard J. Morris https://orcid.org/0000-0003-3080-2613
Steffen Ostendorp https://orcid.org/0000-0001-5931-0511
Melissa Tomkins https://orcid.org/0000-0002-1593-5023
Shuangfeng Wang https://orcid.org/0000-0003-3463-0361
Ying Xu https://orcid.org/0000-0001-5989-9047
Lei Yang https://orcid.org/0000-0001-7969-5981
Yuan Zhou https://orcid.org/0000-0003-1574-7601
Data availability
All study data are included in the article text and/or figures and/
or Supporting Information. The RIP RNAseq data are available
for download at the European Nucleotide Archive (ENA) under
the accession no. PRJEB44090 (https://www.ebi.ac.uk/ena).
References
Alonso JM, Stepanova AN, Leisse TJ, Kim CJ, Chen H, Shinn P, Stevenson
DK, Zimmerman J, Barajas P, Cheuk R et al. 2003. Genome-wide insertional
mutagenesis of Arabidopsis thaliana.Science 301: 653–657.
Alzhanova DV, Napuli AJ, Creamer R, Dolja VV. 2001. Cell-to-cell movement
and assembly of a plant closterovirus: roles for the capsid proteins and Hsp70
homolog. EMBO Journal 20: 6997–7007.
Aoki K, Kragler F, Xoconostle-Cazares B, Lucas WJ. 2002. A subclass of plant
heat shock cognate 70 chaperones carries a motif that facilitates trafficking
through plasmodesmata. Proceedings of the National Academy of Sciences, USA
99: 16342–16347.
Apelt F, Mavrothalassiti E, Gupta S, Machin F, Olas JJ, Annunziata MG,
Schindelasch D, Kragler F. 2022. Shoot and root single cell sequencing reveals
tissue- and daytime-specific transcriptome profiles. Plant Physiology 188:861–878.
Berka M, Kopecka R, Berkova V, Brzobohaty B, Cerny M. 2022. Regulation of
heat shock proteins 70 and their role in plant immunity. Journal of
Experimental Botany 73: 1894–1909.
Cazale AC, Clement M, Chiarenza S, Roncato MA, Pochon N, Creff A, Marin
E, Leonhardt N, Noel LD. 2009. Altered expression of cytosolic/nuclear
HSC70-1 molecular chaperone affects development and abiotic stress tolerance
in Arabidopsis thaliana.Journal of Experimental Botany 60: 2653–2664.
Cazenave C, Uhlenbeck OC. 1994. RNA template-directed RNA synthesis by
T7 RNA polymerase. Proceedings of the National Academy of Sciences, USA 91:
6972–6976.
Chen CY, Shyu AB. 1995. AU-rich elements: characterization and importance in
mRNA degradation. Trends in Biochemical Sciences 20: 465–470.
Chen S-T, He N-Y, Chen J-H, Guo F-Q. 2017. Identification of core subunits
of photosystem II as action sites of HSP21, which is activated by the GUN5-
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
New
Phytologist Research 2419

mediated retrograde pathway in Arabidopsis. The Plant Journal 89:
1106–1118.
Clement M, Leonhardt N, Droillard MJ, Reiter I, Montillet JL, Genty B,
Lauriere C, Nussaume L, Noel LD. 2011. The cytosolic/nuclear HSC70 and
HSP90 molecular chaperones are important for stomatal closure and modulate
abscisic acid-dependent physiological responses in Arabidopsis. Plant Physiology
156: 1481–1492.
Dalmay T, Hamilton A, Rudd S, Angell S, Baulcombe DC. 2000. An RNA-
dependent RNA polymerase gene in Arabidopsis is required for
posttranscriptional gene silencing mediated by a transgene but not by a virus.
Cell 101: 543–553.
Davis AM, Hall A, Millar AJ, Darrah C, Davis SJ. 2009. Protocol: streamlined
sub-protocols for floral-dip transformation and selection of transformants in
Arabidopsis thaliana.Plant Methods 5:3.
De Maio A. 2014. Extracellular Hsp70: export and function. Current Protein &
Peptide Science 15: 225–231.
DiDomenico BJ, Bugaisky GE, Lindquist S. 1982. The heat shock response is
self-regulated at both the transcriptional and posttranscriptional levels. Cell 31:
593–603.
Earley KW, Haag JR, Pontes O, Opper K, Juehne T, Song K, Pikaard CS.
2006. Gateway-compatible vectors for plant functional genomics and
proteomics. The Plant Journal 45: 616–629.
Giavalisco P, Kapitza K, Kolasa A, Buhtz A, Kehr J. 2006. Towards the
proteome of Brassica napus phloem sap. Proteomics 6: 896–909.
Gilbertson RL, Lucas WJ. 1996. How do viruses traffic on the ‘vascular
highway’? Trends in Plant Science 1: 250–251.
Guhaniyogi J, Brewer G. 2001. Regulation of mRNA stability in mammalian
cells. Gene 265:11–23.
Ham BK, Brandom JL, Xoconostle-Cazares B, Ringgold V, Lough TJ, Lucas
WJ. 2009. A polypyrimidine tract binding protein, pumpkin RBP50, forms
the basis of a phloem-mobile ribonucleoprotein complex. Plant Cell 21:
197–215.
Henics T, Nagy E, Oh HJ, Csermely P, von Gabain A, Subjeck JR. 1999.
Mammalian Hsp70 and Hsp110 proteins bind to RNA motifs involved in
mRNA stability. The Journal of Biological Chemistry 274: 17318–17324.
Hentze MW, Castello A, Schwarzl T, Preiss T. 2018. A brave new world of
RNA-binding proteins. Nature Reviews. Molecular Cell Biology 19: 327–341.
Kehr J, Kragler F. 2018. Long distance RNA movement. New Phytologist 218:
29–40.
Kishor A, Tandukar B, Ly YV, Toth EA, Suarez Y, Brewer G, Wilson GM.
2013. Hsp70 is a novel posttranscriptional regulator of gene expression that
binds and stabilizes selected mRNAs containing AU-rich elements. Molecular
and Cellular Biology 33:71–84.
Kishor A, White EJF, Matsangos AE, Yan Z, Tandukar B, Wilson GM. 2017.
Hsp70’s RNA-binding and mRNA-stabilizing activities are independent of its
protein chaperone functions. The Journal of Biological Chemistry 292:
14122–14133.
Lazarowitz SG, Beachy RN. 1999. Viral movement proteins as probes for
intracellular and intercellular trafficking in plants. Plant Cell 11: 535–548.
Lee DW, Kim SJ, Oh YJ, Choi B, Lee J, Hwang I. 2016. Arabidopsis BAG1
functions as a cofactor in Hsc70-mediated proteasomal degradation of
unimported plastid proteins. Molecular Plant 9: 1428–1431.
Leng L, Liang Q, Jiang J, Zhang C, Hao Y, Wang X, Su W. 2017. A subclass of
HSP70s regulate development and abiotic stress responses in Arabidopsis
thaliana.Journal of Plant Research 130: 349–363.
Li S, Wang X, Xu W, Liu T, Cai C, Chen L, Clark CB, Ma J. 2021.
Unidirectional movement of small RNAs from shoots to roots in interspecific
heterografts. Nature Plants 7:50–59.
Lorenz R, Bernhart SH, Honer Zu Siederdissen C, Tafer H, Flamm C, Stadler
PF, Hofacker IL. 2011. VIENNARNA package 2.0. Algorithms for Molecular
Biology 6: 26.
Malter JS. 1989. Identification of an AUUUA-specific messenger RNA binding
protein. Science 246: 664–666.
Maxwell JC. 1868. On governors. Proceedings of the Royal Society of London 16:
270–283.
Mayer MP, Bukau B. 2005. Hsp70 chaperones: cellular functions and molecular
mechanism. Cellular and Molecular Life Sciences 62: 670–684.
Medina V, Peremyslov VV, Hagiwara Y, Dolja VV. 1999. Subcellular
localization of the HSP70-homolog encoded by beet yellows closterovirus.
Virology 260: 173–181.
Merret R, Nagarajan VK, Carpentier MC, Park S, Favory JJ, Descombin J,
Picart C, Charng YY, Green PJ, Deragon JM et al. 2015. Heat-induced
ribosome pausing triggers mRNA co-translational decay in Arabidopsis
thaliana.Nucleic Acids Research 43: 4121–4132.
No€el LD, Cagna G, Stuttmann J, Wirthm€uller L, Betsuyaku S, Witte C-P, Bhat
R, Pochon N, Colby T, Parker JE. 2007. Interaction between SGT1 and
cytosolic/nuclear HSC70 chaperones regulates Arabidopsis immune responses.
Plant Cell 19: 4061–4076.
Notaguchi M, Higashiyama T, Suzuki T. 2015. Identification of mRNAs that
move over long distances using an RNA-Seq analysis of Arabidopsis/Nicotiana
benthamiana heterografts. Plant Cell Physiology 56: 311–321.
Olas JJ, Apelt F, Annunziata MG, John S, Richard SI, Gupta S, Kragler F,
Balazadeh S, Mueller-Roeber B. 2021. Primary carbohydrate metabolism
genes participate in heat-stress memory at the shoot apical meristem of
Arabidopsis thaliana.Molecular Plant 14: 1508–1524.
Oparka KJ, Roberts AG, Boevink P, Santa Cruz S, Roberts I, Pradel KS, Imlau
A, Kotlizky G, Sauer N, Epel B. 1999. Simple, but not branched,
plasmodesmata allow the nonspecific trafficking of proteins in developing
tobacco leaves. Cell 97: 743–754.
Paultre DSG, Gustin MP, Molnar A, Oparka KJ. 2016. Lost in transit: long-
distance trafficking and phloem unloading of protein signals in Arabidopsis
homografts. Plant Cell 28: 2016–2025.
Proctor JR, Meyer IM. 2013. COFOLD: an RNA secondary structure prediction
method that takes co-transcriptional folding into account. Nucleic Acids
Research 41: e102.
Queitsch C, Sangster TA, Lindquist S. 2002. Hsp90 as a capacitor of phenotypic
variation. Nature 417: 618–624.
Rosenfeld N, Elowitz MB, Alon U. 2002. Negative autoregulation speeds the
response times of transcription networks. Journal of Molecular Biology 323:
785–793.
Saplaoura E, Kragler F. 2016. Chapter one –Mobile transcripts and intercellular
communication in plants. In: Lin C, Luan S, eds. The enzymes,vol. 40.
Cambridge, MA, USA: Academic Press, 1–29. [WWW document] URL
https://www.sciencedirect.com/science/article/abs/pii/S1874604716300221
[accessed 12 January 2023].
Schlecht R, Scholz SR, Dahmen H, Wegener A, Sirrenberg C, Musil D, Bomke
J, Eggenweiler HM, Mayer MP, Bukau B. 2013. Functional analysis of Hsp70
inhibitors. PLoS ONE 8: e78443.
Soetaert K, Petzoldt T, Setzer RW. 2010. Solving differential equations in R. R
Journal 2:5–15.
Sung DY, Vierling E, Guy CL. 2001. Comprehensive expression profile analysis
of the Arabidopsis Hsp70 gene family. Plant Physiology 126: 789–800.
Taipale M, Tucker G, Peng J, Krykbaeva I, Lin Z-Y, Larsen B, Choi H, Berger
B, Gingras A-C, Lindquist S. 2014. A quantitative chaperone interaction
network reveals the architecture of cellular protein homeostasis pathways. Cell
158: 434–448.
Takeuchi T, Suzuki M, Fujikake N, Popiel HA, Kikuchi H, Futaki S, Wada K,
Nagai Y. 2015. Intercellular chaperone transmission via exosomes contributes
to maintenance of protein homeostasis at the organismal level. Proceedings of the
National Academy of Sciences, USA 112: E2497–E2506.
ThiemeCJ,Rojas-TrianaM,StecykE,SchudomaC,ZhangW,YangL,
Minambres M, Walther D, Schulze WX, Paz-Ares J et al.2015.Endogenous
Arabidopsis messenger RNAs transported to distant tissues. Nature Plants1: 15025.
Vierling E. 1991. The roles of heat shock proteins in plants. Annual Review of
Plant Physiology 42: 579–620.
Walther D, Kragler F. 2016. Limited phosphate: mobile RNAs convey the
message. Nature Plants 2: 16040.
Williamson DS, Borgognoni J, Clay A, Daniels Z, Dokurno P, Drysdale MJ,
Foloppe N, Francis GL, Graham CJ, Howes R et al. 2009. Novel adenosine-
derived inhibitors of 70 kDa heat shock protein, discovered through structure-
based design. Journal of Medicinal Chemistry 52: 1510–1513.
Winter N, Kollwig G, Zhang S, Kragler F. 2007. MPB2C, a microtubule-
associated protein, regulates non-cell-autonomy of the homeodomain protein
KNOTTED1. Plant Cell 19: 3001–3018.
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
Research
New
Phytologist
2420

Xia C, Zheng Y, Huang J, Zhou X, Li R, Zha M, Wang S, Huang Z, Lan H,
Turgeon R et al. 2018. Elucidation of the mechanisms of long-distance mRNA
movement in a Nicotiana benthamiana/tomato heterograft system. Plant
Physiology 177: 745–758.
Xoconostle-Cazares B, Xiang Y, Ruiz-Medrano R, Wang H-L, Monzer J, Yoo
B-C, McFarland K, Franceschi VR, Lucas W. 1999. Plant paralog to viral
movement protein that potentiates transport of mRNA into the phloem.
Science 283:94–98.
Yang L, Perrera V, Saplaoura E, Apelt F, Bahin M, Kramdi A, Olas J, Mueller-
Roeber B, Sokolowska E, Zhang W et al. 2019. m
5
C methylation guides
systemic transport of messenger RNA over graft junctions in plants. Current
Biology 29: 2465–2476.
Yang Y, Mao L, Jittayasothorn Y, Kang Y, Jiao C, Fei Z, Zhong G-Y. 2015.
Messenger RNA exchange between scions and rootstocks in grafted grapevines.
BMC Plant Biology 15: 251.
Zhang W, Thieme CJ, Kollwig G, Apelt F, Yang L, Winter N, Andresen N,
Walther D, Kragler F. 2016. tRNA-related sequences trigger systemic mRNA
transport in plants. Plant Cell 28: 1237–1249.
Zhang Z, Zheng Y, Ham BK, Chen J, Yoshida A, Kochian LV, Fei Z, Lucas WJ.
2016. Vascular-mediated signalling involved in early phosphate stress response
in plants. Nature Plants 2: 16033.
Zimmer C, von Gabain A, Henics T. 2001. Analysis of sequence-specific binding
of RNA to Hsp70 and its various homologs indicates the involvement of
N- and C-terminal interactions. RNA 7: 1628–1637.
Zuker M, Stiegler P. 1981. Optimal computer folding of large RNA sequences
using thermodynamics and auxiliary information. Nucleic Acids Research 9:
133–148.
Supporting Information
Additional Supporting Information may be found online in the
Supporting Information section at the end of the article.
Dataset S1 Sequences of the used YFP-HSC70.1 fusion con-
structs.
Dataset S2 RT and PCR primers used in this study.
Fig. S1 Schematic predicted folding structures of used HSC70.1
coding sequences.
Fig. S2 Comparison of the hydrogen bonds connection between
HSC70.1 and cloned short variable region coding sequences.
Fig. S3 Confocal laser scanning microscopy images of epidermal
cells expressing the fusion constructs in Arabidopsis.
Fig. S4 YFP(s)::HSC70.1 localization and expression in Col-0
wild-type and hsc70.1 hsp70.4.
Fig. S5 RNA immunoprecipitation (RIP) sample controls and
identified transcripts found enriched in the RIP samples.
Fig. S6 HSC70.1 in vitro translation assay and reaction pathway
used for feedback calculation.
Fig. S7 Representative images of YFP-HSC70.1/Col-0 grafted
plants grown in liquid culture.
Fig. S8 Presence and integrity of YFP-fusion constructs in grafted
wild-type roots.
Fig. S9 Expression of fusion constructs in hsc70.1 hsc70.4
mutants.
Fig. S10 Germination assays and analysis of root growth.
Fig. S11 YFP(s)::HSC70.1 fusion does not complement hsc70.1
hsp70.4 root growth.
Fig. S12 YFP(s)::HSC70.1 fusion does not complement hsc70.1
hsp70.4 early flowering.
Table S1 Western blot bands density relative to mock control.
Table S2 Arabidopsis thaliana HSC70 orthologous transcripts
annotated as mobile.
Table S3 Chemical reactions, system of ordinary differential
equations and parameters.
Please note: Wiley is not responsible for the content or function-
ality of any Supporting Information supplied by the authors. Any
queries (other than missing material) should be directed to the
New Phytologist Central Office.
Ó2022 The Authors
New Phytologist Ó2022 New Phytologist Foundation
New Phytologist (2023) 237: 2404–2421
www.newphytologist.com
New
Phytologist Research 2421