Auxin Gradients Determine Reproductive Development in Pea ( Pisum sativum )

ABSTRACT

Auxins are involved in the regulation of fruit set and development; however, the role of IAA is unclear in pea ( Pisum sativum ) since the endogenous auxin 4‐Cl‐IAA appears to be the auxin stimulating ovary (pericarp) growth. To further understand the role of auxins during fruit development, auxin localization, quantitation, transport, and gene expression activity were assessed in this model legume species. IAA levels and auxin activity (DR5::β‐Glucuronidase [GUS] staining and enzyme activity) were substantially reduced in the pericarp vascular tissues, pedicels, and peduncles of fruit upon seed removal, reflecting auxin transport streams derived from the seeds through these tissues. Seed removal modified auxin response factor PsARF7/19, PsARF8, and PsARF5 transcript levels in the pericarp and attachment tissues in a manner suggesting tissue‐specific regulation of their expression by auxin and ethylene. Pericarp application of polar auxin transport inhibitor N‐1‐naphthylphthalamic acid (NPA) increased auxin (DR5::GUS staining/enzyme) activity within pericarps of seeded, but not deseeded fruits, and NPA application to the peduncle modified IAA levels and DR5::GUS staining/enzyme activity, suggesting polar auxin transport from the seeds to surrounding tissues. However, the NPA application did not induce parthenocarpic fruit growth as in other model species. These data support that in pea, auxin is transported from the seeds to adjacent tissues at least partially through NPA‐sensitive pathways, that seed‐derived IAA plays a role in maintaining auxin gradients through the pericarp and attachment tissues likely for establishing the seed as a major sink, and that auxin and ethylene pathways interact to determine the fate of fruit development.

1. Introduction

Auxin mediates fruit development through an integrated process involving its biosynthesis, transport, and signaling, as well as its interaction with other hormonal pathways (Ozga and Reinecke 2003; Sundberg and Østergaard 2009). Various lines of evidence support the hypothesis that auxin acts as a seed‐derived hormonal signal that coordinates the development of seeds with the surrounding ovary (Bal and Østergaard 2021; Gillaspy et al. 1993; Nitsch 1970; Ozga and Reinecke 2003; Ozga et al. 2009; Sundberg and Østergaard 2009). The initial linkage between fruit set and auxin was established by Gustafson (1936) through the application of auxins to emasculated flowers, which resulted in the production of seedless (parthenocarpic) fruits. Application of auxin transport inhibitors such as N‐1‐naphthylphthalamic acid (NPA) and 2,3,5‐triiodobenzoic acid (TIBA) to non‐pollinated flowers/fruits of Arabidopsis ( Arabidopsis thaliana ), tomato ( Solanum lycopersicum ), and cucumber ( Cucumis sativus ; Dorcey et al. 2009; Kim et al. 1992; Serrani et al. 2010) can also induce parthenocarpic fruit development (fruit growth without fertilization), indicating that elevated levels of endogenous auxins can mimic the fertilization process and stimulate subsequent fruit growth.

The TRANSPORT INHIBITOR RESPONSE 1/AUXIN SIGNALING F‐BOX (TIR1/AFB) and AUXIN/INDOLE‐3‐ACETIC ACID (Aux/IAA) family proteins act as co‐receptors in auxin signaling (Calderón Villalobos et al. 2012). In the absence of auxin, Aux/IAAs repress AUXIN RESPONSE FACTORs (ARFs), the transcription factors regulating auxin‐responsive gene expression. Auxin facilitates TIR1/AFB and Aux/IAA protein interaction, resulting in ubiquitination and degradation of Aux/IAAs by the 26S proteasome complex, releasing ARFs to activate transcriptional auxin signaling response (Wang and Estelle 2014). Parthenocarpy is induced by the loss‐of‐function mutations of AtARF8 in Arabidopsis (Goetz et al. 2007) and the Aux/IAA gene SlIAA9 in tomato (Wang et al. 2005), and by elevated auxin levels via the transgenic overexpression of the tryptophan‐dependent IAA‐biosynthetic gene iaaM in tomato (Ficcadenti et al. 1999) and cucumber (Yin et al. 2006). Altogether, these observations establish that auxin can activate ovary growth and fruit set in the absence of ovule fertilization in these species.

Auxin is generally transported by two distinct pathways in plant tissues: an unregulated bulk flow in the mature phloem away from the source tissues, and a second, regulated, carrier‐mediated cell‐to‐cell directional transport (polar auxin transport) required for the generation and maintenance of local auxin maxima and gradients. Polar auxin transport is mediated by the AUXIN RESISTANT 1 (AUX1) and LIKE AUX1 (LAX) family of auxin influx carriers, the PIN‐FORMED (PIN) family of auxin efflux carriers, and the ATP‐binding cassette subfamily B (ABCB)‐type auxin efflux transporters (Hammes et al. 2022; Petrášek and Friml 2009). The PIN auxin proteins are the primary determinants of transport directionality; AUX1/LAX and ABCB proteins mainly generate auxin sinks and control auxin levels in the auxin channels. The gene transcription of these auxin transport proteins is also influenced by auxin levels (Petrášek and Friml 2009). PIN1 can bind with a range of naturally occurring auxins in addition to IAA, including 4‐chloroindole‐3‐acetic acid (4‐Cl‐IAA), indole‐3‐butyric acid (IBA), and indole‐3‐pyruvic acid (IPA), but whether PIN1 can transport these auxins is not known to date (Yang et al. 2022). NPA has been used in many studies to disrupt polar auxin flux in plant tissues, and recently, PIN proteins were confirmed to be direct targets of NPA (Abas et al. 2021; Teale et al. 2021).

Upon ovule fertilization, the developing fruit enters a stage in which ovary growth and development are tightly coordinated with that of the seeds. The removal or destruction of developing seeds results in reduced pericarp growth and subsequent abscission (pea, Ozga and Reinecke 1999; Ozga et al. 1992, 2003). Seeds contain higher auxin levels than the surrounding fruit tissues (pea, Jayasinghege et al. 2017; Kaur et al. 2021; Magnus et al. 1997; Ozga and Reinecke 2003; tomato, Pattison and Catalá 2012), and are a rich source of gibberellins (GAs), both of which are involved in stimulating the growth of surrounding tissues and determining the final fruit size (Dorcey et al. 2009; Eeuwens and Schwabe 1975; Ozga and Reinecke 1999; Ozga et al. 1992, 2003). 4‐Cl‐IAA is a naturally occurring auxin limited to specific members of the Fabaceae family within the Fabeae and Trifoleae tribes, including pea (Lam et al. 2015; Reinecke et al. 1999). In young developing pea seeds (3–4 days after anthesis; DAA), expression of PsTAR1 (Tryptophan Aminotransferase‐Related) and PsYUC10 (YUCCA flavin monooxygenase) auxin biosynthesis genes was associated with high auxin levels (IAA and 4‐Cl‐IAA; Kaur et al. 2021). 4‐Cl‐IAA can mimic the role of seeds by stimulating deseeded pea pericarp growth (Reinecke et al. 1995) through induction of GA biosynthesis (Ozga et al. 2009), inhibition of ethylene action (Jayasinghege et al. 2017; Johnstone et al. 2005), and direct auxin‐induced enhancement of pericarp growth processes (Ozga et al. 2002); IAA, the ubiquitous auxin in plants, does not. However, seed‐derived IAA in pea may play a crucial role in maintaining auxin gradients through the vascular and adjacent tissues of the pericarp and attachment tissues to establish the seeds as a major sink tissue.

In young tomato fruit, auxin biosynthetic gene expression was highest in the seeds, with high expression of specific auxin transport, signaling, and response genes, and auxin activity in the funiculus, placenta, septum, and pericarp of young tomato fruits, suggesting auxin transport from the seeds towards the outer part of the fruit tissues to promote cell expansion to fill the fruit locular cavity (Pattison and Catalá 2012; Pattison et al. 2015). Auxin application and gene expression profiling studies during the pedicel abscission process support the model that auxin transport from the seeds via the funiculus and vasculature of the placenta/ovary to the fruit pedicel and parent plant helps prevent premature fruit abscission (Bangerth 2000; Ito and Nakano 2015; Meir et al. 2010). In this model, a decrease in auxin provides the first signal for abscission, leading to down‐regulation of auxin‐induced and up‐regulation of auxin‐repressed genes that enhance ethylene sensitivity and pedicel abscission zone formation (Ito and Nakano 2015; Meir et al. 2010). Along with reduced auxin levels, insufficient assimilate production and/or allocation to the fruit is associated with abscission (Bangerth 2000).

As auxin response in pea fruit differs from that in tomato and Arabidopsis, a detailed study of auxin localization, transport, and activity in young developing pea fruit and attachment tissues was completed to further understand the role of auxin in the coordination of fruit development in this model legume species. IAA levels, auxin activity, selected ARF gene expression profiles, and effects of NPA on auxin transport and fruit growth were determined in pea ovary and attachment tissues in the presence and absence of developing seeds. The content of the ethylene precursor 1‐aminocyclopropane‐1‐carboxylic acid (ACC) and the effect of ethylene on selected ARF gene expression patterns were also determined to identify specific auxin/ethylene pathway interactions during fruit set and development.

2. Materials and Methods

2.1. Plant Materials and Growth Conditions

The DR5::GUS construct in the pRD400 vector (DeMason and Polowick 2009) was transformed into P. sativum cv. I₃ (Alaska‐type) using the EHA105 Agrobacterium strain as described by Reinecke et al. (2013). T3 or T4 generation plants homozygous for the transgene were used as DR5::GUS transgenic lines for histochemical and fluorometric analysis of GUS staining and activity as a marker for auxin activity. All other experiments were carried out using P. sativum cv. I₃ (Alaska‐type) non‐transgenic plants.

Pea seeds (four per pot) were planted in 3‐L plastic pots containing a 4:1 (v/v) mixture of Sunshine #4 potting mix (Sun Gro Horticulture) and sand. Seedlings were thinned to three per pot about 2 weeks after planting. Plants were grown in a chamber with a 16 h photoperiod (16/8 h, 19°C/17°C light/dark) under cool‐white fluorescent lights with an average photon flux density of 350 μmol m² s⁻¹. Slow‐release fertilizer (13‐13‐13, N‐P‐K) was added to the potting medium about 3 weeks after planting. The terminal apical meristem remained intact throughout plant development, the expanding lateral shoots were removed as they developed, and the flowers/fruits on the 2nd to 5th flowering nodes of the main stem were used for all experiments. One flower/fruit per plant was used for all NPA experiments, and all other flower buds were removed as they emerged. The fruit (pericarps) remained attached to the plant throughout all experiments until harvested.

2.2. Plant Treatments and Tissue Harvest

Pericarps from 2 DAA self‐pollinated ovaries (length, 15–20 mm) were either left intact or split down the dorsal suture, and seeds were either left intact (split pericarp; SP treatment) or removed (split pericarp no seed; SPNS treatment) as described by Ozga et al. (1992). All treatments to the pericarp were applied to the inner pericarp wall 12 h after ovary splitting or splitting and deseeding to reduce the effects of wounding and residual seed effects on pericarp tissues. All surgically manipulated ovaries were covered with plastic bags until harvest to maintain high humidity. SP and SPNS fruits and attachment tissues were harvested onto ice 48 h after splitting or splitting and deseeding of 2 DAA pericarps.

NPA was applied to peduncles or pedicels in a lanolin paste (25 μL of 1.5 mg NPA per g lanolin); the control treatment was lanolin (25 μL). NPA was applied to the inner pericarp wall as a solution of 10 μM NPA in 0.1% v/v aqueous Tween 80; the control treatment was 0.1% v/v aqueous Tween 80. NPA was applied to the peduncles of 2 DAA intact, SP or SPNS fruits, or to the pericarps of SP or SPNS fruits, or simultaneously to both the peduncles and pericarps of SP or SPNS fruits. For the peduncle treatment, NPA in lanolin paste or lanolin paste only in a foil strip (10 × 5 mm) was applied to the peduncle 20 mm distal to the pedicel‐peduncle junction (foil covered 5 mm of peduncle; see Figure 1A,B) 12 h after pericarp splitting or splitting and deseeding, or the equivalent time for intact fruit. For the pericarp treatment, the inner wall of SP and SPNS pericarps was treated with 30 μL of NPA or control solution 12 h after pericarp splitting or splitting and deseeding of 2 DAA fruits. For the control SP and SPNS treatments, lanolin was applied to the peduncle and Tween 80 solution was applied to the pericarp. For the intact control treatment, lanolin was applied to the peduncle. Fruits with their associated pedicels and peduncles were harvested onto ice 48 h after NPA or control treatments.

FIGURE 1.

Illustrations of experimental procedures. Lanolin or lanolin + NPA application to the peduncle of a pea plant (A, B). Aluminum foil (Al) strip containing 25 μL of NPA in lanolin or lanolin only (A), and the position of the Al strip on the peduncle (B). For GUS enzyme activity experiments using NPA, a 20 mm section of the peduncle above (section proximal to the pedicel, Ppe) and below (section distal to the pedicel, Dpe) the lanolin application site (tissue at application site was discarded), and the entire pedicel (Pl) were harvested (C). Ovaries were dissected into the pericarp ventral suture (Pvs), pericarp dorsal suture (Pds), central pericarp wall (Cpw; ~3–4 mm wide section), the pedicel (Pl; entire tissue) and 20 mm long sections of the proximal (Ppe) and distal (Dpe) peduncle. For the SP treatments, the seeds with the attached funiculi (Sf) were collected (D). For tissue GUS staining, fresh tissue cross sections were taken from the mid‐length region of the ovary (Intact, SP and SPNS; E) and pedicels (F). The peduncles were partitioned into two regions, proximal (Ppe) and distal (Dpe) to the pedicel, and lanolin or lanolin plus NPA was applied to the peduncle between the two regions (G). Peduncle tissue cross sections were collected from the mid region of Ppe and Dpe after the region with lanolin or lanolin plus NPA was removed (H). The initial cut position is indicated with blue lines and the area where cross sections were taken is noted in red (E–H).

To determine the effect of NPA on fruit set, flowers were either emasculated at −2 DAA or allowed to self‐pollinate. At −2 DAA, pedicels (at the middle) or peduncles (20 mm away from the pedicle‐peduncle junction) of emasculated or pollinated ovaries were treated with NPA in lanolin or lanolin only. Fruits were harvested 7 days after treatment (5 DAA) and growth parameters were assessed. For GUS tissue staining, fresh tissue sections of pedicels attached to −2, 0, 2, and 3 DAA fruits and tissues (ovaries, pedicels and peduncles) at 4 DAA were assessed.

For GUS enzyme activity, hormonal quantitation, and gene expression analyses, the inflorescence tissues were dissected immediately after harvesting into pericarp ventral suture (Pvs), pericarp dorsal suture (Pds), central pericarp wall (Cpw, a 3–4 mm wide section approximately the length of the pericarp within the central pericarp region), pedicels (Pl; entire tissue) and 20 mm long sections of the peduncle that were proximal (Ppe) and distal (Dpe) to the pedicel (see Figure 1C,D). In the peduncle NPA treatments, 20 mm sections above and below the lanolin paste application site were harvested, and the tissue at the application site was discarded. In the other treatments, the Dpe was directly below the Ppe. For Intact and SP treatments, the seed with attached funiculus (Sf) was also harvested for some analyses. After dissection, tissues were frozen in liquid nitrogen and stored at −80°C until analysis. Pedicel tissue was limiting and only sufficient for hormone quantitation, but not gene expression analyses for the SP/SPNS experiment.

For GUS staining, fresh‐tissue cross sections were hand‐cut using a razor blade at the mid‐length of each tissue. Then, two cross sections were taken adjacent to the initial cut for GUS staining. When the pericarp contained seeds (intact and SP), cross sections were taken to include the seed and funiculus. When lanolin with or without NPA was applied to the peduncle, the section containing the lanolin was removed, and 20 mm of the peduncle proximal (Ppe) and distal (Dpe) to the pedicel were excised and used for cross sectioning (see Figure 1E–H).

For ethephon treatments, 2 DAA pollinated fruits were split and deseeded (SPNS), and 12 h later, an ethephon solution (1000 mg L⁻¹ in 0.1% v/v aqueous Tween 80; 30 μL) or 0.1% v/v aqueous Tween 80 (30 μL; SPNS control) was applied to the inner pericarp wall. The ethylene action inhibitor silver thiosulfate (STS; 1 mM in 0.1% v/v aqueous Tween 80; 30 μL) was applied once to the inner pericarp wall as a pretreatment immediately after splitting and deseeding of pericarps, followed by the ethephon solution treatment (30 μL) 12 h later. The pericarps were harvested 8 and 12 h after the ethephon treatment, frozen in liquid nitrogen, and stored at −80°C until analysis.

2.2.1. Fruit Growth Experiments

Treatments were applied to the peduncle (NPA in lanolin or lanolin only for the controls), inner pericarp wall (solutions of NPA or Tween 80 for the controls), or to both peduncle and pericarp, of 2 DAA intact, SP or SPNS fruits 12 h after splitting or splitting and deseeding or at equivalent times for intact fruit. Control treatments for SP and SPNS received both peduncle lanolin and pericarp Tween 80 solution. For the Intact treatment, lanolin was applied to the peduncle. Fruit growth and seed number were assessed 2 days after treatment. In a longer‐term growth study, treatments (NPA or control solution; 30 μL) were applied to the inner pericarp wall of SP or SPNS fruits 12 h after splitting or splitting and deseeding 2 DAA fruits. Subsequently, the Tween 80 control solution was applied to the inner pericarp wall of all SP/SPNS fruits approximately 24, 48, 72, and 96 h after the first solution application (30, 40, 40, and 40 μL, respectively). Pericarps were harvested 2 days after the final solution application (9 DAA), and fruit growth and abscission were recorded.

To determine the effect of NPA on fruit‐set, flowers were either emasculated at −2 DAA or allowed to self‐pollinate, and pedicels or peduncles (as described in Figure 1B) were treated with NPA in lanolin or lanolin only. Fruits were harvested 7 days after treatment (5 DAA) and growth parameters were assessed.

2.3. GUS‐Staining Localization

Tissue cross sections were placed in cell culture plates containing GUS staining solution (1 mM 5‐bromo‐4‐chloro‐3‐indolyl β‐D‐glucuronide (X‐Gluc), 100 mM sodium‐phosphate buffer (pH 7), 0.5 mM potassium ferrocyanide, 0.1% Triton X‐100 (v/v), and 10 mM EDTA; Hull and Devic 1995). Tissues were vacuum‐infiltrated for 15 min in the buffer, followed by overnight incubation in the dark at 37°C. Subsequently, tissues were washed three times with 70% aqueous ethanol (v/v) to remove chlorophylls, and micrographs were taken using an Olympus SZ61 stereo microscope (Olympus Corporation, Japan) fitted with a digital camera (MDC320; LeadzOptics Ltd.) controlled by ScopePhoto 3.0 (Scope Technology Inc.) software. A stage micrometer was used to calibrate the scale of the micrograph images. Five to six biological replicates of each tissue type were assessed for each treatment type per experiment, and each experiment was repeated twice over time. GUS staining was only observed in tissues of DR5::GUS plants when compared to non‐transformed plants.

2.4. Quantification of GUS Enzyme Activity

GUS enzyme activity was determined using the 4‐methyl umbelliferyl glucuronide (MUG) assay as described by Jefferson et al. (1987), with minor modifications (see Protocol S1). For each tissue type, tissues from six inflorescences were pooled as one biological replicate, with 4 biological replicates per treatment. GUS enzyme activity was normalized to total protein content by using the Bio‐Rad Protein Assay kit (Bio‐Rad Laboratories) according to the Bradford (1976) method. No GUS enzyme activity was detected in pericarp, seed plus funiculus, pedicel, or peduncle tissues of non‐transgenic pea plants.

2.5. RNA Extraction and qRT‐PCR Assays

The frozen tissues were ground to a fine powder with a combination of manual grinding with a mortar and pestle in liquid nitrogen and bead beating using a Mini‐Bead Beater (Biospec Products). Total RNA was extracted using a modified Trizol‐based method as described by Ozga et al. (2003), and subsequently DNase treated (DNA‐free kit, Ambion) (Kaur et al. 2021). The total RNA concentration was quantified using a NanoDrop (ND‐1000) spectrophotometer.

The TaqMan One‐Step RT‐PCR Master Mix Reagents Kit (Applied Biosystems) was used to quantify the relative transcript abundance of target genes on a StepOnePlus Real‐Time PCR System (Applied Biosystems). Gene‐specific qRT‐PCR primers and ZEN double‐quenched probes were designed with the PrimerQuest tool (Integrated DNA Technologies; Table S1). Each qPCR reaction consisted of a total volume of 20 μL containing 5 μL of total RNA at a concentration of 40 ng μL⁻¹, 1.2 μL each of 5 μM forward and reverse primers, 0.5 μL of 5 μM probe, 0.5 μL of 40× TaqMan arrayScript UP Reverse Transcriptase and RNase inhibitor, 10 μL of 2× TaqMan qRT‐PCR mix, and 1.6 μL of nuclease‐free water. The qRT‐PCR assay plate design and the thermal profile of amplification were used as described in Ozga et al. (2022).

The efficiency (E) of the qRT‐PCR reaction for each gene was calculated with the formula: E = (10^[−1/slope] − 1) × 100 from linear regression curves obtained by plotting the C _t values against logarithmic values of total RNA concentrations of a five‐point 10‐fold serially diluted series (0.434–434 ng μL⁻¹) of a pooled total RNA sample (made by aliquoting equal volumes of total RNA from all the sample tissues), and correlation coefficients (r ²) were calculated (Table S1; Pfaffl 2001). Relative transcript abundance of the target genes was calculated using the ΔC _t method (Livak and Schmittgen 2001), using the formula: transcript abundance = (1 + E) ^(X−Ct), where X is an arbitrary value equal to or greater than the highest assayed C _t value and E is the efficiency of the amplicon assayed. The arbitrary C _t value was set at 33, and the C _t threshold was set at 0.09 for all the target genes. The pea 18S ribosomal RNA (Ps18SrRNA) gene was used as a loading control for qRT‐PCR assays for all samples. The coefficient of variation of the C _t value of all the samples for Ps18SrRNA gene was less than 1.5%; therefore, the target amplicon mRNA values were not normalized to the 18S signal (Nadeau et al. 2011).

2.6. Quantification of Hormones

Frozen tissues (approximately 1 g fwt per replicate; three biological replicates per treatment) were lyophilized and ground to a fine powder. Quantitation was performed at the National Research Council, Saskatoon, Canada, where the dry powder (approximately 50 mg) was extracted, and auxins 4‐Cl‐IAA and IAA, auxin conjugates IAA‐aspartate (IAA‐Asp), IAA‐glutamate (IAA‐Glu), and ACC were quantified using UPLC‐tandem mass spectrometry with internal standards for quantitation (d ₅ ‐IAA, d ₃ ‐IAA‐Asp, d ₃ ‐IAA‐Glu, and d ₄ ‐ACC). For 4‐Cl‐IAA quantitation, d ₃ ‐IAA‐Leu was used as an external standard (Slater et al. 2013). Quantification details are given in Supporting Information Protocol S2 and Table S2.

2.7. Statistical Analyses

To determine the effects of seeds (SP vs. SPNS), NPA application to the peduncle (Intact versus Intact+NPA), or NPA application to the SP pericarp, peduncle, or to both sites simultaneously on GUS enzyme activity, a two‐way analysis of variance (ANOVA; treatment and tissue‐type) was carried out using the PROC MIXED procedure of SAS 9.4 software (SAS Institute Inc.). Differences between treatment and control within tissue type were tested using the Least Square Means (LSMEANS) statement (LSD test).

To determine the effect of NPA application to the pericarp, peduncle, or simultaneously to the pericarp and peduncle on GUS enzyme activity of deseeded fruit (SPNS) and attachment tissues, a one‐way ANOVA was performed, followed by pair‐wise comparisons among treatment means within tissues using the Holm‐Sidak test (SigmaPlot 13; Systat Software Inc.).

For growth experiments, the effect of pericarp SP and SPNS treatments, and NPA application to the pericarp, peduncle, or both sites simultaneously on growth parameters was evaluated using a one‐way ANOVA using the PROC MIXED procedure of SAS 9.4. Differences between treatments were tested using the LSMEANS statement (LSD test). To determine the effect of NPA on fruit set, pair‐wise mean comparisons were completed using a two‐tailed Student's t‐test, taking variance into account (Microsoft Office Excel).

For hormone quantitation experiments, to determine the effect of seeds (SP vs. SPNS) or NPA treatment to the peduncle (Intact vs. Intact+NPA) on auxin and ACC concentrations, pair‐wise mean comparisons were completed using a two‐tailed Student's t‐test, taking variance into account (Microsoft Office Excel). For gene expression experiments, the relative transcript abundance of each gene was transformed by log₂, then the effect of seed removal or NPA treatment to the peduncle of intact fruits with treatment and tissues as factors, or the effect of ethephon with treatment and time as factors, on gene expression was assessed using a two‐way ANOVA with R version 4.3.1. Mean separation was performed using Fisher's LSD post hoc test.

For all experimental analyses, statistical significance wasdeclared at p ≤ 0.05.

3. Results and Discussion

3.1. Effect of Developing Seeds and NPA Application on Fruit Set and Pericarp Growth

Pericarp splitting at 2 DAA resulted in a small reduction in pericarp growth (13%–14%; Intact vs. SP) by 4 DAA (Table 1), which may be due to disturbance of the normal xylem/phloem flow dynamics at the dorsal suture and/or wound‐related ethylene evolution. However, seed removal markedly decreased pericarp growth (63%–76%; SPNS vs. SP; Table 1), consistent with previous studies (Ozga et al. 1992).

TABLE 1.

The effect of seed removal and NPA application to the pea pericarp, peduncle, or pericarp and peduncle at 2 DAA on ovary growth and seed number at 4 DAA. Increase in pericarp length calculated as length at 4 DAA minus length at 2 DAA. Data are means ± SE, n = 23 for pericarp data and n = 11 for seed number per pericarp. Different letters (a–f for pericarp data; a–c for seed data) indicate differences between treatments within the parameter (One‐way‐ANOVA, LSD post hoc test, p ≤ 0.05).

Ovary‐type and treatment	Pericarp		Seed number per pericarp
	Increase in length (mm)	Fresh weight (g)
Intact	27.46 ± 0.82 a	0.612 ± 0.03 a	7.9 ± 0.3 a
Intact‐NPA peduncle	29.63 ± 0.91 a	0.667 ± 0.03 a	7.6 ± 0.4 ac
SP	23.96 ± 0.91 b	0.528 ± 0.02 b	6.1 ± 0.2 b
SP‐NPA peduncle	22.09 ± 0.95 b	0.451 ± 0.03 c	6.6 ± 0.2 b
SP‐NPA pericarp	18.58 ± 1.05 c	0.406 ± 0.03 c	6.6 ± 0.4 bc
SP‐NPA peduncle and pericarp	15.26 ± 0.81 d	0.302 ± 0.02 d	6.5 ± 0.3 b
SPNS	5.80 ± 0.51 e	0.193 ± 0.01 e
SPNS‐NPA peduncle	7.83 ± 0.65 f	0.216 ± 0.01 ef
SPNS‐NPA pericarp	7.21 ± 0.45 f	0.245 ± 0.02 f
SPNS‐NPA peduncle and pericarp	7.68 ± 0.69 f	0.243 ± 0.02 f

Two days after application, NPA peduncle treatment had minimal to no effect on pericarp growth or seed number in intact or SP fruits; however, NPA applied to the inner pericarp wall of SP fruits inhibited pericarp growth (22%–23%), and simultaneous application to the peduncle and pericarp was more inhibitory (36%–43%; compared to SP; Table 1). When assessed 7 days after treatment, NPA pericarp treatment had no effect on SP pericarp growth, seed number, or seed fwt (Table 2), indicating that the disruption of polar auxin transport is minimized over time. In deseeded pericarps (SPNS), all NPA treatments slightly increased (~26%) short‐term (2 day) pericarp growth (Table 1); however, for NPA‐pericarp treatments, further pericarp growth was minimal (7 days after treatment), and pericarps abscised similarly to controls (Table 2).

TABLE 2.

Effect of NPA applied to the inner pericarp wall of 2 DAA split (SP) or split and deseeded pea pericarps (SPNS) on pericarp growth, seed number, and fresh weight, and pericarp abscission 7 days after application (9 DAA). Increase in pericarp length calculated as length at 9 DAA minus length at 2 DAA. Day of abscission is the number of days after treatment until pericarp abscission; all SPNS and SPNS‐NPA pericarps abscised by the end of the experiment. Data are mean ± SE, n = 6 to 9 independent replicates. Different letters (a, b) indicate differences between treatments within parameter (One‐way‐ANOVA, LSD post hoc test, p ≤ 0.05).

Ovary‐type and treatment	Pericarp		Seed		Day of abscission
	Increase in length (mm)	Fresh weight (g)	Number	Fresh weight per seed (g)
SP	44.83 ± 2.39 a	1.982 ± 0.17 a	5.67 ± 0.33 a	0.230 ± 0.05 a	—
SP‐NPA	48.00 ± 1.72 a	1.923 ± 0.19 a	6.43 ± 0.35 a	0.298 ± 0.04 a	—
SPNS	1.89 ± 0.56 b	0.078 ± 0.01 b	—	—	6.33 ± 0.24 a
SPNS‐NPA	2.33 ± 0.33 b	0.063 ± 0.01 b	—	—	5.50 ± 0.34 a

NPA applied to the peduncle or pedicel of pre‐pollinated flowers at −2 DAA had minimal to no effect on pericarp growth following pollination (Figure 2A,B). NPA application to either the peduncle or pedicel of emasculated flowers resulted in minor short‐term pericarp growth, and pericarp tissue integrity was generally maintained several days longer than the control (Figure 2A,C). However, NPA treatments did not trigger parthenocarpic fruit growth, and all non‐pollinated fruits subsequently abscised.

FIGURE 2.

Effect of NPA on fruit growth when applied to pedicels or peduncles of pollinated or non‐pollinated (emasculated) pea fruits. At −2 DAA, flowers were either emasculated or allowed to self‐pollinate, and the pedicels or peduncles were treated with NPA in lanolin paste or lanolin paste alone (control). Fruits were harvested 7 days after treatments (5 DAA). Representative pollinated and non‐pollinated fruits 7 days after treatment (A). Effect of NPA treatments on pericarp length and fwt of pollinated (B) and non‐pollinated (C) fruits. Data are means ± SE, n = 8–10. Asterisks denote significantly different means among the control and the NPA treatment within NPA tissue application site and pollination status at p ≤ 0.05.

In our study, all NPA treatments had no long‐term effect on pea fruit growth in the presence or absence of developing seeds. In contrast, in tomato, NPA application to the pedicels of pollinated ovaries completely blocked fruit set while application to pedicels attached to non‐pollinated ovaries induced parthenocarpic fruit set with fruits similar in size to those of pollinated fruits (Serrani et al. 2010). Similarly, in Arabidopsis, disruption of auxin flow in the male sterile cer6‐2 flowers by NPA produced parthenocarpic fruits (Dorcey et al. 2009). In addition, application of IAA to non‐pollinated ovaries also stimulated parthenocarpic fruit growth in tomato and Arabidopsis (Gustafson 1936; Vivian‐Smith and Koltunow 1999).

The markedly different fruit growth responses to the disruption of polar auxin transport by NPA in pea compared to tomato and Arabidopsis suggest that the auxin transport pathways are different and/or the fruit tissue response to auxin varies in these species. In pea, it is the naturally occurring auxin 4‐Cl‐IAA (but not IAA) that can mimic the role of seeds by stimulating deseeded pea pericarp growth (Reinecke et al. 1995). 4‐Cl‐IAA acts through the induction of GA biosynthesis (Ozga et al. 2009), inhibition of ethylene action (Jayasinghege et al. 2017; Johnstone et al. 2005), and direct auxin‐induced enhancement of pericarp growth processes (Ozga et al. 2002). Ethyl and methyl esters of IAA, reportedly more metabolically stable auxins than IAA, were 100 times more active in stimulating fruit growth than IAA when applied to unfertilized tomato fruit (Sell et al. 1953). However, IAA, the ethyl ester of IAA, and IBA had minimal to no pericarp growth‐promotive activity in the absence of seeds in pea (Reinecke et al. 1995). These results, along with daily applications of fresh IAA to SPNS pea pericarps, suggest that the lack of pericarp growth‐promoting activity of IAA in pea is not due to its biochemical or chemical instability (Reinecke et al. 1995). It is possible that NPA affects the transport of 4‐Cl‐IAA differently from that of IAA. While recent data show that PIN1 can bind 4‐Cl‐IAA in vitro (Yang et al. 2022), experimental confirmation of active transport of 4‐Cl‐IAA by PINs and sensitivity to NPA inhibition in planta is lacking. Future studies examining the transport dynamics of 4‐Cl‐IAA, including carrier specificity and NPA sensitivity, are warranted to test this hypothesis.

Although the disruption of polar auxin transport by NPA does not substitute for seed‐derived signaling to induce pea fruit set and development, the cooperative action of auxin (4‐Cl‐IAA in pea) and gibberellins is part of a signal transduction chain that leads to cell division, elongation, and differentiation for further pea ovary (pericarp) development (Ozga et al. 2009). This auxin (IAA)‐gibberellin interaction also appears to be the paradigm for fruit development in tomato (Serrani et al. 2008), Arabidopsis (Dorcey et al. 2009), and other species such as the accessory fruit strawberry ( Fragaria vesca ; Zhou et al. 2021).

3.2. Auxin Activity Patterns Suggest That Seeds Are a Source of Auxin for the Fruit and Attachment Tissues

3.2.1. Fertilization‐Dependent Auxin Activity Patterns in Pea Fruit and Attachment Tissues

In DR5::GUS plants, GUS staining (blue color) indicating auxin activity was mainly observed in the vascular tissues of the pericarp wall (Pwv), and ventral (Pvs) and dorsal (Pds) sutures (Figure 3A,A′,B) of intact pollinated pea ovaries at 4 DAA. Pea ovary and attachment tissue anatomy is detailed in Figures S1–S3. Intense GUS staining was observed in the funiculus and the seed coat tissue adjacent to the seed‐funiculus attachment (Figure 3B), as previously reported (Jayasinghege et al. 2019). High levels of 4‐Cl‐IAA and IAA occur in young developing pea seeds, with lower levels in the surrounding pericarp tissue (3–6 DAA; Magnus et al. 1997). Both 4‐Cl‐IAA and IAA stimulate DR5::GUS expression in pea pericarps, with 4‐Cl‐IAA producing a relatively stronger auxin response as indicated by a higher GUS enzyme activity (Jayasinghege et al. 2019). Therefore, the observed GUS staining and enzyme activities are likely stimulated by both 4‐Cl‐IAA and IAA. Auxin‐sensitive promoter‐reporter systems also localized auxin activity to the developing seed (by 2 DAA) and the funiculus (by 6 DAA) in tomato (Pattison and Catalá 2012), and in Arabidopsis shortly after anthesis (Fuentes and Vivian‐Smith 2009; Larsson et al. 2017). Following ovule fertilization, high expression of auxin efflux carriers SlPIN4 and SlPIN9 was observed in tomato funiculus tissue (Pattison et al. 2015), and AtPIN3 was expressed and basally localized in a single cell file in the funiculus of Arabidopsis (Larsson et al. 2017). These data support the hypothesis that following fertilization, auxin transport streams are established from seeds to fruit tissues.

FIGURE 3.

GUS staining patterns in cross sections of 4 DAA pollinated intact ovaries and attachment tissues and pedicels attached to pollinated and non‐pollinated ovaries at −2 to 3 DAA from DR5::GUS‐expressing pea plants. Representative GUS staining patterns in pollinated intact ovaries at 4 DAA (A and B; A′, Pds end of pericarp), the attached pedicels (C and D; red arrowheads denote GUS staining in cortex cells proximal to the pericarp dorsal suture attachment), and peduncle tissues proximal and distal to the pedicel (E–G). In (H), ovaries were either emasculated at −2 DAA (non‐pollinated) or allowed to self‐pollinate, and GUS staining of pedicels attached to pollinated and non‐pollinated ovaries at −2, 0, 2, and 3 DAA are shown (the pedicels are oriented with cortical tissue proximal to the pericarp ventral suture attachment at the top of the micrographs). Scale bars: A and B = 1000 μm; C–H = 200 μm. C, cortical cells staining for GUS activity; C/PP, cambium/protophloem/protoxylem; F, funiculus; P, pericarp; Pds, pericarp dorsal suture; Ph, phloem; Pvs, pericarp ventral suture; Pwv, pericarp wall vasculature; S, seed; X, xylem.

In pedicels attached to 4 DAA pollinated fruits, GUS staining was most intense in the vascular bundle cambium/protophloem/protoxylem region (C/PP), with GUS staining also visible in some cortical cells (Figure 3C,D). In peduncle tissue, GUS staining was localized mainly in the vascular bundle C/PP region (Figure 3E–G). Prior to pollination (−2 DAA), intense GUS staining was observed in the pedicel cortical tissue proximal to the pericarp ventral suture attachment (Figure 3H). By 2 DAA, the GUS staining ratio of cortical to vascular tissue in the pedicel decreased, and cortical tissue GUS staining was localized proximal to the pericarp dorsal suture attachment of both pollinated and non‐pollinated ovaries (Figure 3H). Therefore, the change in auxin activity patterns in the pedicel cortical tissue from −2 DAA to 2 DAA appears to be developmentally regulated and not directly linked to the fruit pollination status. GUS staining intensity was higher in the pedicels attached to pollinated ovaries than non‐pollinated ovaries, which will abscise (2–3 DAA; Figure 3H), indicating seeds are an auxin source for the pedicel. High GUS staining was also observed in the pedicel vascular tissue of seed‐bearing tomato fruit (Dong et al. 2021, Pattison and Catalá 2012). Additionally, auxin efflux carrier SlPIN1 silencing was associated with increased auxin activity and IAA content in the tomato ovary and reduced content at the pedicel abscission zone, and accelerated pedicel abscission (Shi et al. 2017). These data support the premise that the establishment and maintenance of auxin gradients from the ovary to the pedicel inhibit pedicel abscission zone formation.

3.3. Tissue‐Specific Distribution of Seed‐Derived Auxin Depends on Polar Auxin Transport

3.3.1. Effect of NPA Application to the Peduncle of Intact and SP Fruits

NPA application to the peduncle of 2 DAA intact fruits did not alter the GUS staining pattern but intensified staining in pericarp and attachment tissues at 4 DAA (Figure 4A,B). GUS enzyme activity increased above the NPA application site in the proximal peduncle tissue (4‐fold), the pedicel and pericarp ventral suture (approximately 3‐fold), and the central pericarp wall (2.1‐fold; Figure 4C,D). Splitting the pericarp dorsal suture without seed removal (SP) increased GUS enzyme activity in the pedicel and peduncle of SP fruits compared to intact fruits (Figure S4A–D), suggesting that more seed‐derived auxin was directed to these tissues when the pericarp dorsal suture was surgically split.

FIGURE 4.

The effect of NPA applied to the peduncles of 2 DAA intact ovaries on GUS staining patterns and enzyme activity in fruit and attachment tissues of DR5::GUS‐expressing plants at 4 DAA. Representative micrographs of GUS staining in fresh tissue cross sections from the mid region of the intact ovary, pedicel, and peduncle (sections proximal and distal to the pedicel) of no NPA (lanolin control) (A) or NPA in lanolin (B; B′, seed and funiculus) peduncle treatments. Scale bars: Ovary = 1000 μm; pedicel and peduncle = 200 μm. GUS enzyme activity in tissues of control (no NPA) or NPA‐treated peduncle treatments (C). Data are means ± SE; n = 4 biological replicates, each biological replicate contains tissues from six inflorescences. Asterisks denote significantly different treatment means within tissues at p ≤ 0.05. Illustration showing the effect of NPA application to the peduncle of intact ovaries on GUS enzyme activity in fruit and attachment tissues (D). C/PP, cambium/protophloem/protoxylem; Cpw, central pericarp wall; Dpe, peduncle section distal to pedicel; F, funiculus; P, pericarp; Pds, pericarp dorsal suture; Pl, pedicel; Ppe, peduncle section proximal to pedicel; Pvs, pericarp ventral suture; Pwv, pericarp wall vasculature; S, seed; SCv, seed coat vasculature; Sf, seed plus funiculus.

GUS staining patterns were similar in SP and intact fruits after NPA application to the peduncle (compare Figures 4A,B and S5A,B), with one exception. GUS staining was observed in xylem parenchyma cells of pedicels attached to SP fruit (50% of pedicels), but not in the control no‐NPA treatment (compare Figures S5A and S5D). This may be due to the NPA‐induced elevated pericarp auxin levels entering the transpiration xylem stream of SP fruit and subsequently being unloaded into the pedicel xylem parenchyma cells. Peduncle NPA treatment increased GUS enzyme activities in the SP pericarp central wall (2‐fold) and ventral suture (3‐fold), similar to that in the intact fruits (Figure 4C,D, intact fruit; Figure 5A,C, SP fruit). As pericarp splitting alone increased GUS enzyme activity in the pedicel and peduncle tissues (Figure S4C,D), peduncle NPA application had less (proximal peduncle) or no effect (pedicel and distal peduncle) on GUS enzyme activity in SP attachment tissues (Figure 5A,C) compared to those attached to NPA‐treated intact fruits (Figure 4C,D).

FIGURE 5.

The effect of NPA application to the peduncle or inner pericarp wall of 2 DAA split pericarps (SP) on GUS enzyme activity in fruit and attachment tissues of DR5:: GUS‐expressing pea plants at 4 DAA. At 2 DAA, NPA in lanolin was applied to the peduncle of SP fruit (A), or NPA in 0.1% aqueous Tween 80 was applied to the inner pericarp wall of SP fruit (B). For the control SP treatment, lanolin was applied to the peduncle and 0.1% aqueous Tween 80 was applied to the pericarp. GUS enzyme activity was assessed in tissues of the SP control and NPA treatments at 4 DAA. Data are means ± SE; n = 4 biological replicates, each biological replicate contains tissues from six inflorescences. Asterisks denote significantly different treatment means within tissues at p < 0.05. Illustrations showing the effect of NPA application to the peduncle (C) or pericarp (D) on GUS enzyme activity in SP fruit and attachment tissues. Cpw, central pericarp wall; Dpe, peduncle section distal to pedicel; Pl, pedicel; Ppe, peduncle section proximal to pedicel; Pvs, pericarp ventral suture; Sf, seed plus funiculus.

NPA application to the inner pericarp wall of 2 DAA SP fruits did not affect GUS staining patterns in ovary tissues at 4 DAA but intensified GUS staining (Figure S5A,C). GUS staining was also observed in the xylem parenchyma cells of both the pedicels and peduncles attached to NPA‐treated SP fruits (60% of pedicels and 50%–53% of distal and proximal peduncles, Figure S5C, red arrowheads). Pericarp NPA treatment increased GUS enzyme activities in the SP pericarp ventral suture by 3‐fold and the central pericarp wall by 1.6‐fold, with minor or no effects on the attachment tissues (Figure 5B,D). Data from the NPA‐pericarp application experiments are consistent with NPA inhibition of polar auxin transport within the ovary with developing seeds, resulting in auxin accumulation and higher auxin activity in the pericarp tissues and reduced or minimal changes in auxin activity in the attachment tissues.

Additional disruption of polar auxin transport by simultaneous application of NPA to the peduncle and pericarp of 2 DAA SP fruits resulted in GUS staining patterns similar to that in the pericarp‐NPA treatment, but a higher frequency of GUS staining was observed in the xylem parenchyma of pedicels (86%) and peduncles (33%–44%) when compared to NPA treatments to the peduncle or pericarp alone (Figure S6A, red arrowheads). GUS enzyme activities further increased in the pericarp ventral suture and pedicel tissue with simultaneous NPA application to the pericarp and peduncle compared to single‐site application, and a trend or significant increase in GUS enzyme activity in the pericarp wall tissue was also observed (Figure S6B,C). In tomato, NPA also increased GUS staining in placental tissues of pollinated DR5::GUS‐expressing fruits (Pattison and Catalá 2012), and pedicel NPA application increased the amount of ovary‐applied ³H‐IAA remaining in the ovaries while reducing the amount in the pedicel below the application site (Serrani et al. 2010). Overall, these data indicate that polar auxin transport is involved in the establishment of an auxin gradient from developing pea seeds through the ventral pericarp suture to the pedicel and peduncle tissues.

3.3.2. Effect of Seed Removal and NPA Application to Deseeded Pericarps

Seed removal at 2 DAA (SPNS) did not affect GUS staining patterns; however, staining intensities were lower in all tissue types compared to SP by 4 DAA (Figure 6A,B). Consistently, GUS enzyme activity decreased approximately 2‐fold with seed removal in the ovary (Cpw and Pvs), pedicel, and proximal peduncle tissues. In the distal peduncles (the most distant tissue from the seeds), GUS enzyme activity was reduced approximately 4‐fold in the absence of seeds (Figure 6C,D). Additionally, NPA application to SPNS pericarps had no effect on GUS enzyme activity in any of the tissues assessed (compare SPNS to SPNS NPA‐pericarp treatments; Figure 7A,B). These data further support the premise that seeds are a major auxin source for pea fruit and attachment tissues.

FIGURE 6.

The effect of seed removal at 2 DAA on GUS staining patterns and enzyme activity in fruit and attachment tissues of DR5::GUS‐expressing pea plants at 4 DAA. Representative micrographs of fresh tissue cross sections from the mid region of the ovary, pedicel, and peduncle (proximal and distal to the pedicel) when seeds were present (A; SP), or 2 days after seeds were removed from the pericarp (B; SPNS). Scale bars: Ovary = 1000 μm; pedicel and peduncle = 200 μm. GUS enzyme activity was assessed in tissues of SP and SPNS treatments (C). Data are means ± SE; n = 4 biological replicates, each biological replicate is composed of tissues from six inflorescences. Asterisks denote significantly different treatment means within tissues at p ≤ 0.05. Illustration showing the effect of deseeding on the GUS enzyme activity in fruit and attachment tissues (D). C/PP, cambium/protophloem/protoxylem; Cpw, central pericarp wall; Dpe, peduncle section distal to pedicel; F, funiculus; P, pericarp; Pds, pericarp dorsal suture; Pl, pedicel; Ppe, peduncle section proximal to pedicel; Pvs, pericarp ventral suture; Pwv, pericarp wall vasculature; S, seed.

FIGURE 7.

The effect of NPA application at 2 DAA to the pericarp, peduncle, or simultaneously to the pericarp and peduncle on GUS enzyme activity in deseeded ovaries (SPNS) and attachment tissues of DR5::GUS expressing pea plants at 4 DAA. At 2 DAA, NPA in lanolin was applied to the peduncle, or NPA in 0.1% aqueous Tween 80 was applied to the inner pericarp wall, or NPA was applied to both the peduncle and pericarp of SPNS fruit. For the control SP and SPNS treatments, lanolin was applied to the peduncle and 0.1% aqueous Tween 80 was applied to the pericarp. GUS enzyme activity was assessed in tissues at 4 DAA (A). Data are means ± SE; n = 4 biological replicates, each biological replicate contains tissues from six inflorescences. Means with different letters (a, b, c) are significantly different within tissue at p ≤ 0.05. Illustration showing the effect of NPA application to the pericarp (B) on GUS enzyme activity in SPNS fruit and attachment tissues (blue box = no change). Cpw, central pericarp wall; Dpe, peduncle section distal to pedicel; Pl, pedicel; Ppe, peduncle section proximal to pedicel; Pvs, pericarp ventral suture.

GUS enzyme activity was lower in the SPNS pedicel and pericarp wall tissue compared to SP regardless of NPA application (to peduncle, pericarp, or both sites simultaneously; Figure 7A), and this was associated with minimal pericarp growth and subsequent fruit abscission. In contrast, non‐pollinated fruits from NPA‐treated DR5::GUS‐expressing tomato plants showed intense GUS staining in the pedicels and placental tissues, and NPA treatment resulted in parthenocarpy and thickening of the fruit pedicel distal to the abscission zone (Pattison and Catalá 2012). A continuous polar flow of auxin through the abscission zone has been suggested to be vital to prevent the premature abscission of the attached organ (leaves, flowers, or fruits), and in some species, this is suggested to involve an auxin‐induced decrease in ethylene sensitivity (Brown 1997; Ma et al. 2015; Meir et al. 2010, 2015; Oberholster et al. 1991). The reduction of auxin activity in pedicels attached to deseeded pea ovaries that will subsequently abscise is consistent with the concept that pedicel auxin depletion facilitates the formation of an abscission zone at the pedicel‐peduncle junction.

3.4. Seed Removal Decreases and NPA‐Application Increases IAA Levels in Specific Pericarp and Attachment Tissues

In the intact ovary, IAA levels were substantially higher in the seed/funiculus (Sf) tissue than in the pericarp and attachment tissues (Figure 8C). 4‐Cl‐IAA was only detected in the seed/funiculus (1.4 ± 0.3 ng g⁻¹ fwt; Sf); however, we found that the method for extraction and quantification of auxins used in this study was sensitive for low levels of IAA and IAA‐conjugates, but not sensitive enough for quantification of low levels of 4‐Cl‐IAA (higher levels occur in young pea seeds and lower levels in pericarp tissues; Magnus et al. 1997). In the intact fruit, which is at the seed stage prior to rapid development of the embryo, IAA levels in the seed/funiculus tissue are higher than in the other fruit tissues, but the GUS activity is similar to that of the other fruit and attachment tissues (Figures 4C and 8C). A similar scenario was observed in DR5‐GUS‐transformed tomato fruit, where minimal GUS staining in 5 DAA fruit/seeds was in contrast with the direct measurement of IAA (~8 ng g⁻¹ fwt) at this stage (Pattison and Catalá 2012). Using a second auxin‐responsive reporter (DR5rev‐mRFPer) consisting of a DR5rev promoter (which is known to be more auxin sensitive than DR5 during embryo development; Liao et al. 2015) coupled to a monomeric red fluorescent protein (mRFPer), fluorescence was detected in the ovules after fertilization, and the signal was strong and widespread in the developing seeds at 2 DAA, and in the funiculus and the outer layers of the placenta that surround the seeds at 6 DAA (Pattison and Catalá 2012). Therefore, it is likely that the DR5‐GUS auxin reporter and the associated MUG assay that quantifies the GUS enzyme activity are not very sensitive to auxin levels in the seed/funiculus tissue at this stage of development.

FIGURE 8.

IAA and ACC content in 4 DAA fruit and attachment tissues of SP and SPNS 2 days after seed removal (A and B), or 2 days after NPA application to the peduncle attached to intact fruits (lanolin control or NPA in lanolin; C and D). Data are means ± SE; n = 3 biological replicates. Asterisks denote significantly different treatment means within tissues using the two‐tailed Student's t‐test (p ≤ 0.05). Cpw, central pericarp wall; Dpe, peduncle section distal to pedicel; Pds, pericarp dorsal suture; Pl, pedicel; Ppe, peduncle section proximal to pedicel; Pvs, pericarp ventral suture; Sf, seed‐funiculus. Y‐axis for A, 0–65:100–160, for B, 0–65:100–6000 for C, 0–400:4000–10,000.

Seed removal markedly reduced IAA levels (8.8‐fold) in the pericarp ventral suture, to which seeds connect via the funiculus. IAA levels decreased further as the tissue distance from the seeds increased (pedicel, 10.7‐fold; proximal peduncle, 18.6‐fold; distal peduncle, 24.5‐fold; compare SP and SPNS treatments; Figure 8A). The reduction of IAA levels is consistent with decreased GUS enzyme activity observed with seed removal in these tissues (Figure 6C,D). Lower IAA levels were also observed in pea pericarps of non‐pollinated ovaries compared to pollinated ovaries with developing seeds (Jayasinghege et al. 2019). In the absence of seeds, the levels of IAA‐Asp, a major IAA amino acid conjugate produced in pea that is considered to be a metabolic end product, were lower in the above tissues, as well as the central pericarp wall, indicating a reduction in IAA metabolism with seed removal (Figure S7A).

NPA application to the peduncle of intact fruit increased IAA levels in tissues above the application site 1.2 to 1.3‐fold in the Ppe and pericarp (Cpw, Pds, Pvs; Figure 8C). Although a trend toward higher IAA levels in the pedicels was observed with NPA treatment, the difference was not statistically significant; however, the pedicel IAA‐Asp and IAA‐Glu levels were significantly higher, indicating that the flux through IAA was higher (Figure S7B). The increase in tissue IAA levels with peduncle NPA application is consistent with increased GUS enzyme activity observed for these tissues (Figure 4C,D).

3.5. Seed Removal Reduces Pericarp ACC Levels

During early pea fruit development, ethylene evolution increases in the pericarp in the absence of pollination or due to seed removal (Johnstone et al. 2005; Orzáez et al. 1999; Savada et al. 2017), likely facilitating fruit senescence. The levels of ACC, the immediate precursor to ethylene, were markedly reduced 2 days after seed removal in pericarp tissues (Cpw, 10‐fold; Pds, 5.3‐fold; Pvs, 7.8‐fold; SP vs. SPNS; Figure 8B). These data suggest that conversion of ACC to ethylene was higher in SPNS than SP pericarp tissues, which is consistent with higher ethylene evolution levels in SPNS pericarps 8, 12, and 24 h after seed removal compared to pericarps with seeds (SP; Johnstone et al. 2005). Consistently, an increase in PsACO1 and PsACO2 expression was correlated with a 6‐fold increase in ethylene evolution by 3 DAA and subsequent non‐pollinated pea fruit senescence (Savada et al. 2017).

ACC levels were not significantly affected in the pedicel or peduncle tissues 2 days after seed removal (Pl, Ppe and Dpe; Figure 8B). Consistently, ethylene evolution was minimal in the peduncle tissues attached to pollinated and non‐pollinated pea ovaries from 0 to 3 DAA (Savada et al. 2017). However, an increase in PsACO1 and PsACO2 expression and ethylene evolution was observed in pedicels attached to non‐pollinated fruit by 3 DAA (Savada et al. 2017), which likely facilitates abscission as fruit senescence progresses. NPA treatment to the peduncle of intact fruit had minor to no effect on ACC levels in the fruit or attachment tissues (Figure 8D), suggesting that the NPA‐induced increases in auxin levels (Figure 8C) and auxin activity (Figure 4C,D) had minimal impact on ACC levels.

3.6. Auxin and Ethylene Pathways Interact to Coordinate Signaling Between the Ovary and Attachment Tissues

In Arabidopsis, ARF5, 6, 7, 8, and 19 are considered to function as transcriptional activators in auxin response (Guilfoyle and Hagen 2007; Tiwari et al. 2003). A phylogenetic analysis of pea transcriptional activator ARFs within the P. sativum genomedatabase (Kreplak et al. 2019) was performed based on the evolutionary similarities with the 23 ARFs from Arabidopsis (Figure S8). PsARF5, two PsARF8 genes (PsARF8‐1 and PsARF8‐2), and two PsARF7/19 genes were identified. PsARF7/19–1 and PsARF7/19–2 have a dual nomenclature, as these highly homologous genes did not differentiate into discrete ARF7 or ARF19 classes.

Seed removal (SPNS) was associated with reduced peduncle auxin activity (Figure 6C,D) and reduced expression of PsARF8‐1 (Dpe, 2.8‐fold; Ppe, 4.4‐fold), PsARF8‐2 (Dpe and Ppe, 3‐fold), and PsARF7/19–1 (Dpe, 3‐fold; Ppe, 3.4‐fold; Figure 9C–E). However, seed removal had little to no effect on PsARF8‐1, PsARF8‐2, and PsARF7/19‐1 expression in pericarp tissues, even though auxin activity was reduced (Figures 6C and 9C–E). In tomato, arf8 mutants induce parthenocarpy, and additionally, arf8b‐1 was shorter than WT, and the arf8a8b double mutant was shorter and produced smaller leaves than arf8b‐1, indicating ARF8A and ARF8B additively promote stem growth and leaf expansion (Hu et al. 2023). The responsiveness of PsARF8‐1, PsARF8‐2, and PsARF7/19‐1 expression to auxin changes in the peduncle may similarly indicate a role for these genes in auxin‐related peduncle growth and development in pea.

FIGURE 9.

The effect of seed removal and ethylene on the relative transcript abundance of the auxin receptor PsAFB6 and PsARFs in fruit and attachment tissues. The transcript abundance of PsAFB6 (A), PsARF5 (B), PsARF8‐1 (C), PsARF8‐2 (D), PsARF7/19–1 (E), and PsARF7/19–2 (F) in fruit and attachment tissues of SP and SPNS fruit (2 days after seed removal) at 4 DAA. The effect of ethephon (Eth) and ethephon plus STS (STS + Eth) applied to 2 DAA SPNS pericarps on PsARF5, PsARF8‐1, PsARF7/19‐1, and PsARF7/19‐2 pericarp transcript abundance 8 and 12 h after treatment (G–J). Data are means ± SE (n = 3–4 biological replicates for A–F and n = 2–3 for G–J). Means with different letters (a–f) are significantly different across treatments and tissues within genes at p ≤ 0.05 (A–F) and across treatments and time within genes (G–J) using a two‐way ANOVA with means separation by LSD (p ≤ 0.05). Cpw, central pericarp wall; Dpe, peduncle section distal to pedicel; Pds, pericarp dorsal suture; Ppe, peduncle section proximal to pedicel; Pvs, pericarp ventral suture.

PsARF5 transcript abundance was higher in SPNS pericarp and attached peduncle tissues than that in tissues connected to fruits with developing seeds (SP; Figure 9B). Consistently, in tomato, SlARF5 transcript abundance decreased in pollinated fruit and markedly increased in non‐pollinated fruits from 0 to 3 DAA (Liu et al. 2018). Characterization of higher‐order loss‐of‐function tomato ARF mutant combinations (SlARF5, SlARF7, SlARF8A, and SlARF8B) showed that all four SlARFs function as inhibitors of fruit initiation, as mutants showed an increased frequency of parthenocarpic fruit growth. In contrast, these ARFs act as promoters of subsequent fruit growth, indicated by the reduced size of parthenocarpic fruits in higher‐order mutants (Hu et al. 2018, 2023).

Functional studies of ARF2, ARF1, ARF7, and ARF19 suggested that these transcription factors act with partial redundancy to promote senescence and floral abscission (Ellis et al. 2005). Furthermore, ARF19 and ARF7 not only participate in auxin signaling, but also play a critical role in ethylene responses in Arabidopsis roots, and ethylene was shown to modify root ARF19 expression (Li et al. 2006). Tissues from SPNS pericarps, which will senesce and abscise, exhibited higher PsARF7/19‐2 expression (Figure 9F) and greater ethylene evolution (Johnstone et al. 2005) compared to SP pericarps, which will continue to grow. Furthermore, ethylene treatment decreased the transcript abundance of PsARF5, PsARF8‐1, and PsARF7/19‐1 (Figure 9G–I), but increased that of PsARF7/19‐2 in deseeded pericarps (Figure 9J). The ethylene action inhibitor STS reversed the ethylene effect on gene expression in each case. In tomato, SlARF7/SlIAA9 functions as part of a signaling repressor complex inhibiting tomato fruit set and development by repressing the expression of growth‐related genes and activating ethylene biosynthesis gene ACO4 expression (Hu et al. 2018). Our data also suggest that auxin‐ethylene pathways interact through PsARF7/19‐2 to facilitate ovary senescence in ovaries without developing seeds.

Although application of NPA to the peduncle of intact pea fruit increased auxin activity in all tissues except the seed/funiculus (Figure 4), and IAA levels in all pericarp tissues and the proximal peduncle (Figure 8C), it had minimal to no effect on the expression of PsARF5, PsARF8‐1, PsARF8‐2, PsARF7/19‐1, and PsARF7/19‐2, with one exception (Figure S9). Transcript abundance of peduncle PsARF5 increased 1.6‐fold, suggesting a role for PsARF5 in auxin‐related responses in this tissue.

Developing seeds also influence the transcript abundance of TIR1/AFB auxin receptors. Higher PsAFB6 transcript abundance was observed in SPNS pericarps and attachment tissues than that of tissues from fruit with developing seeds (SP; Figure 9A), and this was associated with reduced auxin activity in all tissues and reduced IAA levels in the pericarp ventral suture and the attachment tissues (Figures 6C and 8A). Previous studies observed that pea pericarp PsAFB6 expression increased with seed removal. Application of 4‐Cl‐IAA or IAA suppressed PsAFB6 expression in deseeded pea pericarps 2 h after treatment; however, only 4‐Cl‐IAA was able to suppress expression for an extended period, at least up to 12 h after application (Ozga et al. 2022). Application of IAA to deseeded pericarps had no effect on the expression of other auxin receptors (PsTIR1a, PsTIR1b, PsAFB2, PsAFB4), but 4‐Cl‐IAA also suppressed PsTIR1b and PsAFB4, although to a lesser extent than PsAFB6 (Jayasinghege et al. 2019; Ozga et al. 2022). Ethylene also increased PsAFB6 expression in deseeded pea pericarps, which was inhibited by pretreatment with 4‐Cl‐IAA but not IAA (Ozga et al. 2022). Therefore, in a low auxin environment, the TIR1/AFB receptor transcript pool will be enriched in PsAFB6, indicating that it may play a role in fruit growth as part of a feedback‐regulation loop to increase auxin sensitivity in these tissues.

4. Conclusion

This study has advanced our understanding of how auxins contribute to fruit development in pea. Although the auxin‐responsive DR5 promoter analysis cannot distinguish between auxin types, we can conclude that an auxin gradient is established from the seeds to the fruit and attachment tissues. Our analysis shows that IAA is involved in establishing this gradient, and that auxin distribution from pea seeds to the ovary, pedicel, and peduncle tissues is at least partially mediated by polar auxin transport. Furthermore, the expression of specific auxin response elements is modulated in the fruit and attachment tissues when the seeds (a source of auxin) are removed from the fruit. In the peduncle tissue that will remain viable following senescence of the deseeded fruit, a decrease in IAA levels and auxin activity is associated with decreased expression of PsARF8‐1, 8‐2, 7/19‐1, and 7/19‐2. In the fruit tissue, PsARF7/19‐2 expression is up‐regulated by deseeding and ethylene application, demonstrating a link between the auxin and ethylene pathways likely to facilitate ovary senescence in the absence of seeds. Auxin activity was mainly localized to the vascular tissues of the ovary, pedicel, and peduncle. In a low auxin environment, the TIR1/AFB receptor transcript pool is enriched in PsAFB6, indicating that it may play a role in reproductive development as part of a feedback‐regulation loop to increase auxin sensitivity in the ovary and attachment tissues. The inability of the polar auxin inhibitor NPA to stimulate parthenocarpy or longer‐term ovary growth in pea, as it does in other model fruit species (tomato and Arabidopsis), may reflect differences in the biological activity and/or transport dynamics of 4‐Cl‐IAA compared to IAA during pea fruit development, and/or other factors are involved as some species from plant families that do not contain 4‐Cl‐IAA also do not exhibit auxin‐induced parthenocarpy.

Author Contributions

J.A.O. and D.M.R.: conceptualization, methodology; D.D.A. and H.K.: formal analysis, visualization, validation; D.D.A., H.K., and C.P.A.J.: investigation; J.A.O.: resources, supervision, funding acquisition, writing – original draft; J.A.O., D.D.A., D.M.R., H.K., C.P.A.J.: writing – review and editing.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Illustrations of the inflorescence, and pedicel and peduncle anatomy.

Figure S2: Vascular arrangement in the pea pericarp.

Figure S3: Tissue arrangement at the pea pericarp ventral and dorsal sutures.

Figure S4: Effect of pericarp splitting on GUS staining patterns and enzyme activity.

Figure S5: Effect of NPA applied to the peduncle or pericarp of SP on GUS staining patterns.

Figure S6: Effect of NPA applied simultaneously to the peduncle and pericarp of SP on GUS staining patterns and enzyme activity.

Figure S7: IAA, IAA‐Asp, and IAA‐Glu content in SP and SPNS fruits, or after NPA application to the peduncle of Intact fruits.

Figure S8: Phylogenetic analysis of P. sativum auxin response factor proteins (PsARFs).

Figure S9: Effect of NPA application to the peduncle of Intact fruit on expression of PsARFs.

Table S1: Primers and probes used for qRT‐PCR quantitation.

Table S2: Precursor‐to‐product ion transitions of hormones and standards used for UPLC/ESI‐MS/MS analysis.

Protocol S1: Quantification of GUS enzyme activity.

Protocol S2: Quantification of hormones.

Adihetty, D. D. , Kaur H., Jayasinghege C. P. A., Reinecke D. M., and Ozga J. A.. 2025. “Auxin Gradients Determine Reproductive Development in Pea ( Pisum sativum ).” Physiologia Plantarum 177, no. 5: e70497. 10.1111/ppl.70497.

Data Availability Statement

All data supporting the findings of this study are available within the paper and within its Supporting Information published online.