N-fixation defect innlp2. Nonetheless, the N-
fixation potential of bothnlp2-1andnlp2-2
was also compromised, albeit to a lesser degree,
when grown in substrate (fig. S1D) and could be
partially complemented by expression ofNLP2
from its native promoter, but not by a mutant
version ofNLP2(fig.S1,EandF).Furthermore,
(^15) N
2 assimilation was compromised innlp2-1
plants nodulated in substrate (Fig. 2A), which
corresponded to a decrease in N content in nod-
ules and shoot, while %N was unchanged (Fig. 2A
and fig. S2D). Evaluation ofnlp2mutants under
inhibitory nitrate regimes indicated that loss
ofNLP2doesnotaffectNsuppressionofnod-
ulation (fig. S2, A to C). Together these results
suggest a positive role for NLP2 in N 2 fixation.
In addition to reduced N fixation,nlp2
nodules appeared less pink (fig. S2E); this was
reflected in lower LgHb expression and heme
levels (Fig. 2, B and C, and fig. S2F). We then
profiledthegeneexpressionofnlp2-1nodules
under different N regimes using substrate
where the mutant forms normal-sized nodules,
rather than using aeroponics where the smaller
nodules on the mutant would confound the
analysis. The results revealed nitrate-dependent
effects on gene expression (Fig. 2D and data
S1). Genes encoding LgHbs, heme biosynthesis
enzymes, and nitrate uptake transporters were
strongly down-regulated innlp2-1(Fig. 2, E and
F, and table S1), whereas jasmonate-related
genes were up-regulated (table S2), as previ-
ously reported in root hairs ofM. truncatula
nin( 5 ). Expression ofNLP2from its own pro-
moter partially rescued gene expression in
nlp2, but a mutant version did not (Fig. 2G).
In silico promoter analysis revealed that
28 of the 55 genes with the largest overall
decrease in expression innlp2nodules con-
tained nitrate-responsive elements (NREs),
which are required for transcriptional activa-
tion by NLPs ( 6 ). This includedSymCRKand
SEN1, which are required for N fixation, and
many genes encoding nodule cysteine-rich
peptides (NCRs) ( 7 , 8 ) (table S3). In addition, a
larger motif consisting of two partly overlap-
ping NREs was detected in theLgHbgenes.
We designated these as“double-NREs”(dNREs),
in which the second half of the first semi-
palindrome (NRE1) overlaps with the first half
of the second element (NRE2) (Fig. 3A). The
dNREs all have the same orientation and are
positioned between–100 and–300bpupstream
of the TSS (Fig. 3B). dNREs were also detected
at the same position and orientation in pro-
moters ofLgHbs of soybean and common bean,
indicating their conservation across papilionoid
legumes (fig. S3A). Examination of dNREs in
promoters of soybean and medicagoLgHbs
suggests that those strongly differing from
the consensus motif have lower expression in
nodules (fig. S3, B and C), implying the direct
involvement of dNREs in the regulation of
LgHbexpression.
Our transcriptomic data verified thatNLP2
andNINare the primary NLPs expressed in
medicago nodules (figs. S1A and S4); because
NIN binding sites strongly resemble NREs
( 9 , 10 ) (fig. S5), we speculated that both NLP2
and NIN may directly promoteLgHbexpres-
sion. Using an electrophoretic mobility shift
assay (EMSA), we found that truncated NLP2
626 29 OCTOBER 2021•VOL 374 ISSUE 6567 science.orgSCIENCE
Fig. 1. Expression ofNLP2inM.truncatulanodules
andnlp2nodulation phenotypes.(A) Nodule expression
ofNLP2. Shown is a histochemically stained nodule cross
section [15 days post-inoculation (dpi) withS. meliloti
Rm2011] expressingpMtNLP2:GUS.Scale bar, 100mm.
(B) Expression [as measured by quantitative reverse
transcription polymerase chain reaction (qRT-PCR)] of
NLP2,NIN,LgHb2,LgHb7, andLgHb8relative to reference
genes in nodules of wild-type (WT) plants inoculated
withS. melilotiRm2011 grown in attapulgite substrate
with no nitrogen.n= 3. (C) Representative DAPI-stained
70-mm nodule sections ofNLP2knockout (NLP2-KO)
and control (EV) roots grown using aeroponics with no
nitrogen. Scale bar, 100mm. Nodules were evaluated 11 dpi
withS. melilotiRm2011 carrying apNifH:GFPreporter.
(D) Acetylene reduction activity of nodules ofNLP2CRISPR-
edited roots grown aeroponically without nitrogen. WT
M. truncatulaseedlings (A17) were transformed with a gRNA
targetingNLP2and roots were genotyped. Mean numbers
of nodules from edited and non-edited roots were compared
(n= 19) with a Kruskal nonparametric test,a= 0.05.
(E) Nodules ofnlp2-1and WT nodules grown aeroponically
without nitrogen, 18 dpi withS. melilotiRm2011. (F) Left:
Biomass of nodulated (nod) (18 dpi withS. melilotiRm2011)
and uninoculated plants. Right: Nodule number. Plants
were grown aeroponically without nitrogen;n= 10 for WT,
11 fornlp2-1. In (B), (D), and (F), error bars denote SD.
In (B) and (F), means were compared with StudentÕsttest,
a= 0.05. In (B), (D), and (F), P< 0.05.
(^1) CAS-JIC Centre of Excellence for Plant and Microbial Science
(CEPAMS), Centre for Excellence in Molecular Plant Sciences
(CEMPS), Shanghai Institute of Plant Physiology and Ecology
(SIPPE), Chinese Academy of Sciences, Shanghai, China.
(^2) LIPME, Université de Toulouse, INRAE, CNRS, Castanet-
Tolosan, France.^3 CIRAD, UMR PVBMT, Pôle de Protection des
Plantes, Saint-Pierre 97410, France.^4 Noble Research Institute,
Ardmore, OK 73401, USA.^5 Shanghai Engineering Research
Center of Plant Germplasm Resource, College of Life Sciences,
Shanghai Normal University, Shanghai 200234, China.
(^6) Department of Plant Sciences, University of Oxford, Oxford
OX1 3RB, UK.^7 John Innes Centre, Norwich Research Park,
Norwich NR4 7UH, UK.
Corresponding author. Email: [email protected]
RESEARCH | REPORTS